WO1994009144A1 - Novel plants and processes for obtaining them - Google Patents

Abstract

Plants, particularly cereal plants, which have improved ability to synthesise starch at elevated or lowered temperatures and/or to synthesise starch with an altered fine structure are produced by inserting into the genome of the plant (i) a gene(s) encoding a form of an enzyme of the starch or glycogen biosynthetic pathway, particularly soluble starch synthase and/or branching enzyme and/or glycogen synthase, which display an activity which continues to increase over a temperature range over which the activity would normally be expected to decrease, and/or (ii) a gene(s) encoding sense and anti-sense constructs of enzymes of the starch biosynthetic pathway, particularly soluble starch synthase and/or branching enzyme and/or glycogen synthase, which alters the natural ratios of expression of the said enzymes or inserts enzymes with special structural characteristics which alter the natural branching pattern in starch.

Description

NOVEL PLANTS AND PROCESSES FOR OBTAINING THEM

This invention relates to novel plants having an improved ability to produce starch including an improved ability to produce structurally-altered starch. Such novel plants are capable of producing higher yields than known plants, and/or are capable of producing starch of altered quality. The invention further relates to processes for obtaining such plants.

Temperature is one of the most important ecological factors governing the natural distribution of plants and their satisfactory growth and yield potential in agricultural cultivation. A crop will give its highest yields, and lowest risk of failure, when it is cultivated as close as possible to the specific temperature optima for each of its development stages in the course of the growing season. In many agricultural regions, temperature is not a stable factor for crops with extended growth periods and the plants may suffer stress because of temperatures which are too high or too low, or both, in different intensities and over short or long time intervals. Many of the world's food crops are cultivated in regions where their yield is constrained by what we have called "thermal thresholds" for optimal growth (Keeling and Greaves, 1990). The actual optimum temperature for maximal yield differs amongst different crops: for example, it is well known that certain cereal plants (such as wheat, barley and maize) give a maximum grain yield at around 25° to 30°C. This optimum temperature is the basis for the calculation of Heat Units (HU) which are used to calculate Growing Degree Units (GDU). GDU is a measure of the HUs a plant requires within a particular temperature range to reach a maximum yield. "Grain filling" is dry matter accumulation in the grain, and occurs over the period during which the grain increases in weight. Published information on crops such as wheat, barley, maize, rice and sorghum, shows that as the growth temperature increases the duration of grain filling declines. At temperatures below the optimum temperature, the rate of grain filling increases with increasing temperature. This compensates or even overcompensates for reduced duration of grain filling such that, overall, the yield increases with temperature (Figure 1). However, at higher temperatures the grain filling rate fails to increase further with temperature and, indeed, declines with temperature increases above 30°C. These changes in grain filling rate fail to compensate for further reduction in grain filling duration. The overall effect is reduced yield (and alεo starch quality) because cereal crops spend significant proportions of time during the grain filling period at temperatures which are higher than optimal for grain filling rate. Thus the limitations on grain filling rate imposes a penalty on the total amount of dry matter that can be accumulated in any one growing season. In addition there is a change in starch fine structure, affecting its quality, because the amylose/amylopectin ratio is affected and the starch granule density is reduced. Although these effects are documented in the literature, their cause has hitherto been unknown.

Arrhenius first developed a generalisation for the effects of temperature on the rates of biochemical reactions and proposed a unifying mechanism involving energy of activation (Ea) of chemical reactions. This has become known as the Arrhenius Equation: ln(k₂/k₁) = (E /R)(l/T₁ + 1/T₂) where the k values represent the velocity constants of each temperature (T) and R is the gas constant. The

Arrhenius equation defines a chemical mechanism for the generalisation of a factor "Qι₀" where the rate will double or treble for a 10°C rise in temperature. Q_1Q - Velocity (T + 10)°C/Velocity T°C

Enzymes, which determine the velocity of any chemical reaction in living organisms, act by lowering the free energy of activation values of the chemical reactions to the extent that the thermal energy which is present in the organism is sufficient to activate the reactants. Thus the Q₁₀ values associated with enzyme catalysed reactions are therefore a physical characteristic of the energy of activation achieved by that enzyme. Q-_n values will therefore remain fixed provided (i) the catalytic site of the enzyme-protein remains functional and (ii) the substrate itself is not affected significantly by changed temperature. It is a well known fact that at temperatures which exceed the ther odynamic stability of proteins

(normally considered to be about 40 to 50°C) the protein structure will "unfold" and its catalytic function will become disrupted and hence it will no longer favour conversion of substrates to products. This phenomenon is well characterised in the literature and has been extensively studied in a variety of living systems as well as in thermophilic organisms. The finding that dry matter accumulation in starch-storing crops apparently has an optimum temperature of 25 to 30°C is very unusual because this temperature is lower than the temperature at which proteins are normally considered to become disrupted. The phenomenon thus lacks any explanation in the literature. Our previous studies have led to a new understanding of the metabolic pathway of starch synthesis in developing starch storing tissues (Keeling e_t al, 1988, Plant Physiology, 87:311-319; Keeling, 1989, ed. CD. Boyer, J.C. Shannon and R.C. Harrison; pp.63-78, being a presentation at the 4th Annual Penn State Symposium in Plant Physiology) .

We have established that grain filling rate and duration is governed by factors in the grain itself rather than being due to some component of the source tissues. Since starch comprises up to 75% of the grain dry weight in cereals it is logical that the limitation to plant dry matter accumulation referred to above will be most likely due to an effect on starch deposition. Thiε suggests that cereal specieε or varietieε with higher yield will be characteriεed by either longer duration of εtarch syntheεiε in the grain or greater increaεes in the rate of starch syntheεiε at elevated temperatureε.

Furthermore, this work has given us an understanding of the factors involved in determining the optimum growth temperatures of all starch storing plants, and an understanding of the enzymes responsible for determining the fine structure of starch. In particular, we have investigated the biochemical reactions and interactions between the soluble starch synthase and branching enzymes which contribute to the fine- structure of starch deposition in plants.

An object of the present invention is to provide novel plants having an increased capacity to produce starch and a capacity to produce starch with an altered fine structure.

According to the present invention there is provided a method of producing a plant with altered εtarch εyntheεiεing ability compriεing stably incorporating into the genome of a recipient plant one or more than one donor gene specifying an enzyme involved in a εtarch or glycogen bioεynthetic pathway.

The above method is generally applicable to all plants producing or storing starch. The recipient plant may be: a cereal such as maize (corn), wheat, rice, sorghum or barley; a fruit-producing species εuch as banana, apple, tomato or pear; a root crop such as cassava, potato, yam or turnip; an oilseed crop εuch aε rapeseed, sunflower, oil palm, coconut, linseed or groundnut; a meal crop such as soya, bean or pea; or any other suitable species. Preferably the recipient plant is of the family Gramineae and most preferably of the species Zea mays.

The method according to the invention may be used to produce a plant having an improved capacity to produce εtarch at elevated or lowered temperature. Aε noted previously, yield increases with temperature until a temperature optimum is reached. Above the temperature optimum, yield begins to decrease. Thus plants growing above (or below) their temperature optimum may not be reaching their full yield potential. Improving the plant's capacity to produce starch at temperatures above or below the temperature optimum will result in increased yield. This may be achieved by increasing the amount or type of starch synthesising enzymeε present in the plant. It may also be achieved by altering the actual temperature optimum of starch synthesis to suit the growing conditions of a particular plant. Thus a crop variety may be produced which is adapted to the growth temperature of a particular environment (including particular sites or geographical regions). Normally, it is most useful to improve starch production at temperatures in excess of the normal optimum temperature. The method according to the invention may also be used to produce a plant having the ability to synthesise starch with an altered fine εtructure. Thiε may be due to a shift in the temperature optimum of starch synthesis or due to other her reasons (such as a change in the overall balance of the different enzymes in the biosynthetic pathway). The fine structure of the starch affects itε quality. It is thus possible to generate crops producing starch which is better adapted or targetted to the crops' end-use (such as starch with improved processing properties, improved digestibility, etc).

The method and resulting alterations in starch εyntheεising ability are particularly advantageous for maize. The temperature optimum of starch synthesis pathway enzyme activity may be matched more closely to the higher temperature rangeε encountered in the typical maize growing regions of the world. In addition, the fine structure of the starch may be changed so that novel starches are made in the recipient plant.

The donor gene to be used may be obtained from any biological source which produces starch or glycogen-synthesising enzymes: it may be of plant origin or fungal origin or bacterial origin or animal origin. For example, donor genes may be derived from a plant selected from the following species Zea mays (especially the varieties Lima 38, Guanajuato 13, Lima 45, Doebley 479 or teosinte 154), Zea diploperenniε, Zea luxurianε, Zea perennis, Zea tripsacum, Zea parviglumiε, and Zea mexicana.

Preferably the donor gene εpecifies soluble starch synthase (E.C. 2.4.1.21) and/or branching enzyme (E.C. 2.4.1.18) and/or glycogen synthaεe of bacterial origin (E.C. 2.4.1.21) or animal origin (E.C. 2.4.1.11).

We describe the isolation, purification and characterisation of the enzymes soluble starch synthase and branching enzyme and the production of antibodies which can be used in the identification of soluble starch synthase and branching enzyme cDNA clones. We alεo describe the use of cDNA cloneε of plant soluble starch synthaseε (SSS - εoluble starch synthase), plant and bacterial branching enzymes (BE - branching enzyme) and plant and animal and bacterial glycogen synthases (GS - glycogen synthase). , „, , „ PCI7GB92/01881 09144 o

The donor gene may be an additional copy of the gene specifying the normal enzyme present in the plant (that is, an additional copy of the wild-type gene). Increased gene expression may also be elicited by introducing multiple copies of enhancer sequenceε into the 5'-untranεcribed region of the donor gene.

The donor gene may specify an alternative enzyme with improved properties compared to those 0 of the normal plant enzyme. For example, glycogen synthase (involved in the glycogen biosynthetic pathway) may be seen as equivalent to soluble starch synthaεe (in the εtarch biosynthetic pathway) as it catalyses a similar reaction. 5 However, glycogen synthase (such as glgA found in E coli) haε a higher temperature optimum of activity than soluble starch synthase, and so its expression in the plant εhould improve the plant's capacity to produce starch at elevated temperatures. A second 0 example is bacterial branching enzyme (such as glgB from E coli) which is equivalent to plant branching enzyme, but may have improved propertieε.

The donor gene may also specify a modified allelic form of the enzyme with kinetic or allosteric properties different to those of the normal plant enzyme. In particular, the enzyme encoded by the donor gene may have a temperature optimum of activity higher than that of the recipient plant enzyme, or may show enhanced thermal-stability (which may enhance the duration of grain filling) .

Expresεion of improved allelic formε of an enzyme within a recipient plant will produce a plant with increaεed starch-yielding capacity and/or better starch quality. Examples of such plants include: a plant (especially maize) having a εtarch synthesiεing ability which does not decrease with temperature between 25 to 30°C; a plant (especially maize) containing a soluble starch synthase and/or branching and/or glycogen synthase enzyme with a Q,_Q value greater than 0.8 between 25 and 35°C; a plant (especially maize) containing a soluble starch synthaεe and/or branching and/or glycogen synthase enzyme which is resiεtant to reduction of activity after exposure to a temperature in exceεε of 40°C for two hourε.

Geneε encoding improved allelic forms may be obtained from suitable biological organisms, or endogenous wild-type genes may be manipulated by standard protein or genetic engineering techniques.

Starch-producing and glycogen-producing organismε (including plants, fungi, bacteria, animal cells) may be used as sourceε of improved enzyme geneε. Such organisms may be screened for allelic forms of the enzyme which are more catalytically active than those typically found in the crop. It is also possible to alter the properties of the enzyme in the recipient plant through protein and genetic engineering. Genes encoding variants of the enzymes may be created using molecular techniques or mutagenesiε.

The donor gene may be an antiεense sequence which reduces expression of the enzyme in the recipient plant. For example, this may be used to alter the balance of the different starch-synthesising enzymes present in the cell which may change the amount or type of εtarch produced. In particular, it is poεεible to inεert geneε in both senεe and/or anti-sense orientations in order to effect a change in the natural ratios of variouε isoforms of branching enzyme and soluble starch synthase in the recipient plant. Altering the natural ratios of the soluble starch synthase and branching enzyme activities of the recipient plant results in starch being produced with new and novel branching structures, amylose/amylopectin ratios, and an altered starch fine structure. An antisense donor gene (used to reduce expression of the wild-type gene) may also be incorporated with a sense donor gene encoding a more active replacement enzyme. For example, a senεe gene encoding glycogen εynthase may be incorporated together with an antiεense gene for soluble starch synthaεe to potentiate the effect of increased temperature optimum described above.

It is possible to insert more than one copy of the donor gene into the recipient genome. Each donor gene may be identical, or a combination of different donor genes may be incorporated. For example, the donor genes may have differing sequences which may encode more than one allelic form of the enzyme or may be derived from more than one source.

In summary, the following are examples of genetic manipulation methods which may be uεed to produce plants with an altered ability to synthesise starch:

(1) Insertion of an additional copy of the wild-type enzyme gene;

(2) Inεertion of a more active or more thermally reεiεtant enzyme gene obtained from a _n P

suitable biological organism;

(3) Insertion of multiple copieε of the wild-type or enhanced activity enzyme gene;

(4) Modification of the enzyme gene by techniqueε known in protein engineering to achieve alterations in the kinetics and/or allostericε of the enzyme reaction (eg to achieve thermal stability) ;

(5) Modification of the promoter sequences using techniques known in protein engineering to achieve enzyme over-expression;

(6) Modification of the sequence of the enzyme gene and/or its promoter and/or the transit peptide. The said donor gene may be derived from a sexually compatible donor plant and inserted into the recipient plant by sexual crossing of donor and recipient plants.

Alternatively, the donor gene may be isolated from a suitable biological organism, such as a plant, bacterium, fungus or animal cell. Insertion of the donor gene iε effected by genetic transformation of the recipient plant. If the donor organism is not sexually compatible with the recipient, the gene to be incorporated into the recipient plant genome is excised from the donor organism and the genome of the recipient plant is transformed therewith using known molecular techniqueε. The advantages of the transformation method are that isolated enzyme genes, genes from diverse biological sourceε and/or anti-sense constructs (not just εenεe conεtructε) may be incorporated in the recipient plant genome.

Preferably the recipient plant is of the family Gramineae and most preferably of the species Zea mays. Other recipient plants may include rice, wheat, or barley, and fruit crops such as tomato.

The invention also provides a plant having one or more than one donor gene specifying an enzyme involved in a starch or glycogen biosynthetic pathway stably incorporated into its genome such that itε ability to produce starch is altered. Such plants may have an improved capacity to produce starch at elevated or lowered temperature, and/or an ability to syntheεiεe εtarch with an altered fine εtructure. Hence εuch plantε are capable of producing higher εtarch yieldε at certain temperatures and/or are capable of producing starch with an improved quality.

The variety of poεsible plants, donor genes and resulting effects have been discussed previously.

The invention also provides the εeedε and progeny of εuch plantε, and hybridε whose pedigree includes such plants.

The preεent invention will now be deεcribed, by way of illustration, by the following description and examples with reference to the accompanying drawings of which:

Figure 1 is a graph showing the effect of temperature on final grain weight.

Figure 2 is a graph of εtarch synthesis rate against temperature. Figure 3 is a graph of field temperature against time spent above that temperature.

Figure 4 shows computer simulations of different rate-models of temperature/activity profiles. PCI7GB92/0188Ϊ ₃

Figure 5 is a graph of rate againεt temperature for εeveral εtarch-εynthesising enzymes.

Figure 6 is a graph of Q_1Q values against temperature for εeveral εtarch-εyntheεiεing enzymes.

Figure 7 is a graph of rate against temperature for maize endosperm soluble starch synthaεe and branching enzyme.

Figure 8 iε a graph of Q_1Q valueε against temperature for maize endosperm soluble starch εynthaεe and branching enzyme.

Figure 9 iε a graph of rate against temperature for rice, maize, sorghum and millet soluble starch synthase.

Figure 10 is a graph of Q._Q values against temperature for rice, maize, sorghum and millet soluble starch εynthaεe.

Figure 11 is a graph εhowing soluble starch synthase and ADPG pyrophosphorylase activity against pre-incubation temperature.

Figure 12 is a graph showing soluble starch synthaεe activity and εtarch synthesis against pre-incubation temperature. Figure 13 is a graph showing soluble starch synthaεe activity againεt time after pre-incubation at 40°C.

Figure 14 is a graph showing recovery of soluble starch synthaεe and UDPG pyrophosphorylase activity against time after heat treatment.

Figure 15 is a graph εhowing Q_{* Q} valueε of εoluble εtarch εynthaεe and glycogen synthase againεt temperature.

Figure 16 iε a graph of εoluble εtarch synthase activity against temperature using various primers.

Figure 17 is a graph of soluble starch synthase activity against temperature using various primers.

Figure 18 shows the nucleotide sequence for E coli glycogen synthase.

Figure 19 shows the construction of two glycogen εynthaεe tranεformation vectors. Figure 20 is a graph showing the frequency diεtribution for Q_1Q values (25 to 35°C) for different maize lineε.

Figure 21 iε a graph showing soluble starch synthase activity against temperature for variouε maize genotypeε.

Figure 22 iε a graph showing soluble starch synthaεe activity eluted from the mono-Q column. Figure 23 εhowε activity of variouε εtarch syntheεiεing enzymeε at different gene doses in the sugary maize mutant.

Figure 24 shows activity of various starch synthesiεing enzymeε at different gene doεeε in the dull maize mutant.

Figure 25 εhowε activity of variouε starch synthesiεing enzymes at different gene doses in the shrunken maize mutant.

Figure 26 shows activity of various starch synthesiεing enzymeε at different gene doses in the amylose extender maize mutant.

EXPERIMENTAL BACKGROUND TO THE INVENTION In order to examine in detail the relationεhip between starch εyntheεis rate and temperature in vivo, we have conducted experiments on maize plants ₅

growing under defined temperature regimes in constant environment cabinets. In these experiments the maize ears have been maintained at defined temperatures independently of the rest of the plant by means of thermostatically controlled glass fibre heating mantles placed around the ears. Rates of starch synthesis at precisely regulated temperatures were calculated from dry weight accumulation curves for the maize kernels over a defined period. The analysiε of maize endoεperm εtarch syntheεis at different temperatures showed that the rate increases with temperature between 15 and 25°C. Between 25 and 30°C there was no significant increase in rate. The rate at 35°C was significantly lower than the rate at 30°C (Figure 2).

To examine the relevance of the in vivo starch accumulation rate versus temperature curve in the field, we monitored field temperatures during the starch filling period at various field stations in the mid-west of the United States. Air temperatureε were monitored uεing a thermocouple attached to a recording device which recorded temperatures at five minute intervals. In addition, temperatures within the maize kernels were recorded by inserting thermocouples into the grain. Significant periods of time are spent above the optimal temperature (Figure 3). We used computer modelling techniques to calculate the expected yield benefits of altering the in vivo starch εynthesiε versus temperature curves in several defined ways (Figure 4). For example, increasing the temperature optimum for grain filling rate from 25 to 30°C gives an anticipated yield benefit in excess of 10% over the years which we have recorded (Table 1).

TABLE 1 Computed increases in yield for a 5°C increase in temperature optimum

17

In addition to in vivo information on temperature dependence, we have obtained in vitro evidence on the biochemical baεiε for the temperature dependence. We applied 14C-glucoεe to developing maize grain in vitro at different temperatures and measured the flux of radioactivity through different pathways. Using 1- 14C-glucoεe and 6- 14C-glucose it was possible to determine the in vitro rates of glycolyεis and the pentose phoεphate pathway as well as starch synthesiε. The data in Table 2 below εhow that between 20 and 40°C there waε no εignificant increaεe in the rate of εtarch εyntheεis, though the flux of radioactivity through the pentoεe phoεphate pathway and glycolyεis increased approximately three-fold and two-fold respectively. The failure of the starch synthesiε rate to increase above 25°C is therefore due to the starch synthesiε enzymeε in the amyloplaεt (i.e. any one or any combination of soluble and bound starch εynthase, branching enzyme and ADP-glucose pyrophosphorylaεe) , and not due to a failure in the εupply of εucrose or ATP at high temperatures.

TABLE 2 Effect of temperature on radioactivity released as carbon dioxide or incorporated into starch from 1-14C and 6-14C glucose

To analyse further the biochemical basis for the temperature optimum for grain filling rate in cereals, we meaεured the activities with respect to temperature of most of the maize endosperm enzymes in the pathway for converting sucrose to starch, including soluble starch synthase (SSS) and branching enzyme (BE). Two aspects have been studied:

(1) The temperature dependence of the rates of the reactions catalysed by these enzymeε; and.

(2) The stabilitieε of the individual enzymeε during incubationε at different temperatureε both in vivo and in vitro.

The glycogen synthase enzyme has alεo been studied (see (3) below).

(1) Temperature Dependence

The reaction rateε of moεt of the enzymes studied (alkaline pyrophosphatase,phoεpho- glucomutaεe, UDP glucoεe pyrophoεphorylaεe, hexokinase, phoεphoglucosisomeraεe, εucrose εynthase, ADP glucose pyrophosphorylase and bound εtarch synthase) have activities which increase with temperature at least up to 45°C (Figure 5). This temperature exceeds the highest temperature recorded during grain filling in the field. When replσtted aε Q._Q valueε (Figure 6) it iε clear that these enzymic reactions are stable acrosε this temperature range. This is consistent with the. theoretical expectation from the Arrhenius equation referred to above.

The temperature responses of SSS and BE were, however, a complete contrast (Figure 7). The apparent temperature optima for activity are 25°C for SSS and 27.5°C for BE. When replotted as Q₁₀ values (Figure 8) it is clear that there is an apparently constant decline in enzymic efficiency with increasing temperature. This is indicative of some decrease in catalytic activation brought about by depressed interaction between the enzyme and its subεtrates. As these are the only enzymes in the pathway with temperature optima for activity below 30°C, it is apparent that one or both of SSS and BE are "rate limiting" for starch synthesis between 20 and 30°C. The failure of starch synthesis rate in vivo to increase with temperature between 25 and 35°C must be due to a failure of the activities of one or both of these enzymes to increase in activity over this temperature range.

Grain has been sampled from maize plants grown at 20, 25 and 30°C and SSS and BE activity haε been meaεured over several temperatures. The temperature at which the grain are growing was found to have no effect on the temperature optima for activity of these two enzymes. Temperature activity curves for these two enzymeε from the endoεperms of several commercial U.S. maize hybrids show the same temperature optima. As shown in Figure 8, Q_1Q valueε for in vivo rateε and SSS and BE activities show remarkable similarities. This is in contrast to, for example, ADP-glucose pyrophosphorylase where Q_1Q is fixed across different temperatures as predicted by the Arrheniuε equation.

Compariεonε of the temperature activity curveε for SSS from maize, rice, sorghum and millet show that the temperature optima for activity were consistently around 25°C (Figure 9). There were, however, differences in Q. _Q between 15 and 25°C, the Q,_Q increasing in the order maize, εorghum, rice and millet (Figure 10), that is, in the order of increasing climatic temperatures of the locations where the crops are grown. Further studies of dicot plants has led to the discovery that the Q_1Q for SSS can be considerably higher in some plantε than we have observed in the monocot plantε εo far εtudied. For example in tomato leaf and fruit the Q,_Q valueε were significantly higher than that seen in maize (Figure ) .

High increaseε in Q_1Q between 15 and 25°C mean that SSS may not become rate limiting at 25°C, pathway rate control being paεsed to another enzyme(s). The in vivo rate of flux through the starch pathway may, therefore, continue to increase with temperature above 25°C.

The temperature activity curves of the green leaf tissue forms of SSS and BE are εimilar to those observed for the developing grain. These enzymeε, we conclude, limit photosynthesis at temperatures above 25°C.

The temperature activity curves of the developing tassel forms of SSS and BE are similar to those observed for the developing grain forms, again confirming that these two enzymes limit starch synthesiε at temperatures above about 25^βC.

(2) Temperature Stability

Experiments with both wheat and maize have confirmed that increased temperature not only affects the rate of enzyme catalysed reaction of SSS, but also results in enzyme instability resulting in long term loss of enzyme activity. We have termed this phenomenon "knockdown" in order to distinguish it from the changes in reaction rate described above. "Knockdown" iε defined aε the change in maximum catalytic activity aεsayed at 20°C following a heat treatment of the enzyme either iri vitro or i_n vivo at a higher temperature. Incubation of extracted maize endosperm SSS in vitro at temperatures up to 37°C did not cause any losε of enzyme activity when εubεequently assayed at 20°C. At temperatures higher than 37°C, however, progresεive long term loεs of enzyme activity occurred, such that by 40°C, 80% of the enzyme had become inactivated. When peeled maize grains were preincubated for three hours at 40°C, there was a εignificant drop in SSS activity when εubεequently extracted and aεεayed at 20°C (Figure 11). In contraεt the extractable activity of ADP-glucoεe pyrophoεphorylaεe waε not altered by this same preincubation. Pre-treatments for three hours at 10, 20 or 30°C did not result in any drop in the activity of SSS. Similar in vitro incubations of wheat grain resulted in substantial loss of SSS activity at temperatures as low as 30°C while ADP-glucose pyrophosphorylase activity was unaffected. The drop in SSS activity at 40°C coincided with a drop in starch synthesis measured as 14C incorporation into starch assayed in-vitro

(Figure 12) . In experiments in which maize ears were heated in situ on the plant, temperatures in excess of 37°C caused loss of SSS activity when subεequently extracted and assayed at 20°C. The activity of ADP-glucose pyrophosphorylase was, again, unaffected by thiε treatment. Following heat treatmentε at 40°C for a range of timeε, it waε found that the extractable SSS activity declined dramatically over a two hour period (Figure 13). When the cobε were returned to 20°C following a period at 40°C, the extractable SSS activity only returned to itε initial activity after a period of 24 hourε at the lower temperature (Figure 14). Similar experimentε with wheat earε alεo showed that knockdown of SSS occurred in vivo ^ru

at temperatures of 30°C and above. Recovery of SSS activity in wheat alεo took over 24 hours.

(3) Temperature Dependence of Glycogen Synthase The biochemical reaction catalysed by soluble starch synthase (ADPGlucose:1,4-α-D-glucan4-α-D- glucosyltranεferaεe (E.C. 2.4.1.21)) involves the sequential addition of glucose donated from the sugar nucleotide ADPGlucose to a glucan chain uεing an α-1,4 linkage. Other εourceε of thiε chemical reaction exist in nature as the enzymes glycogen synthaεe notable from bacteria (ADPGlucose: 1,4-α-D-glucan4-α-D-glucosyltransferase (EC2. .1.21) ) and animals (UDPGlucose:glycogen 4-α-D-glucoεyltransferaεe (EC2.4.1.11) ) .

When we studied the temperature dependence of the glycogen synthase reaction we found that there was a remarkable increase in activity with increasing temperature. This waε in dramatic contraεt to the temperature dependence of the reaction catalysed by soluble starch εynthase (Figure 15). The biochemical cause of this new temperature response of glycogen synthase indicates that thiε enzyme can maintain itε normal activation energy for the reaction even at higher temperatureε that reduce SSS activity. The glycogen synthase enzymes catalytic site is either able to tolerate the higher temperature or else iε able to compensate for possible changes in the conformation of the substrate itself. Thus becauεe glycogen εynthase catalyseε the εame chemical reaction aε SSS, it haε very wide application as a means of increasing plant starch depoεition at higher temperatureε. The nucleotide and amino-acid sequences of glycogen synthaεe are known (i) from E.Coli GenBank/EMBL #J02616 (Kumar et al, J Biol Chem _34 16256-16259 (1986)), (ii) from rabbit skeletal muscle (Hang et al, FASEB J .3 3532-3536 (1989)), and (iii) from human muscle (Browner et al, Proc Nat Acad Sci QS_ 1443-1447 (1989)). The bacterial (ADPGlucose: 1,4-α-D-glucan-4-α-D-glucoεyl- tranεferaεe (EC 2.4.1.21)) and animal (UDPGlucose: glycogen 4-α-D-glucosyltraεferaεe (EC 2.4.1.11)) sequences are NOT homologous. Furthermore the bacterial forms are not phosphorylated and also are not allosterically affected by glucose 6-phosphate. Finally, the bacterial enzyme useε ADPG (ie, like plants) and only the animal forms uεe UDPG. Thiε makes the bacterial enzyme the ideal choice for using on plants. The εtructural geneε for the bacterial glycogen εynthaεe are mapped to pOP12 in E.Coli and glycogen synthase map to glgA. Nucleotide sequencing further refines the position of glgA. The translation start of glgA is known to be immediately after glgC and the nucleotide sequence determined. The NH_ εequence was known so that the actual start of the glgA gene was unambiguously determined as well as confirming the direction of transcription. The deduced amino acid sequence show complete homology with the known NH~ sequence and with the known amino acid sequence. Different bacterial enzymes show 90% sequence homology. There iε complete agreement between the reported and deduced amino acid sequences for the enzyme. Cells transformed with the gene produce a polypeptide that has sequence homology with the known amino acid sequenceε. THE BIOCHEMICAL CAUSE OF THE TEMPERATURE EFFECT

As was εtated above, the Q, ₀ characteriεticε of an enzyme catalysed reaction is determined by the energy of activation of that enzyme. The atypical behaviour of SSS and BE to increasing temperature shows that the normal activation energy of this reaction which appears to be optimal at lower temperatureε (for example, around 10°C) iε not carried through to the higher temperatureε. The decay in enzyme rate with increasing temperature is indicative of some change in the interaction between the enzyme and its substrate. Thiε could be due to a change in the enzyme catalytic site or else a change in the conformation of the substrate itself. Studies of the effects of different primers on the temperature-dependence of SSS activity shows that the precise temperature optimum is influenced by the nature of the substrate (Figures 16 and 17). Furthermore, when very low molecular weight primers were used in the enzyme reaction the temperature responεe curve was dramatically altered such that the Q_1Q was 1.6 between 25 and 35°C for maltotriose, maltotetroεe and maltoheptoεe primers. Collectively these data indicate that the biochemical cause of the temperature optimum of SSS activity is related in some way to the molecular structure of starch and its interaction with SSS and BE.

STARCH FINE STRUCTURE ALTERATIONS

Detailed studieε of the εtarch mutants in corn (amylose extender, sugary, dull and waxy) has shown-uε that the expression of enzyme activities is altered in these mutants such that there are new ratios of several enzymes involved in the pathway of starch synthesis. Figures 23 to 25 show the considerable over-expression of several enzymes in the pathway of starch synthesis, particularly where there is a specific lesion in the normal pattern of starch deposition. Furthermore we have found that not only is there a change in overall enzyme expression of some enzymes, but also there is an over-expression of some isoformε of SSS and BE enzymeε. Theεe mutationε therefore not only cauεe a reduction in expreεεion of SSS and BE enzymeε, as has been reported in the literature, but there is apparently an over-expresεion of the other isoforms of theεe εame enzymeε in the pathway. Thiε change in enzyme isoform expression occurs in response to or as a conεequence of a change in flux of sugars to starch. There is then a resulting change in starch fine structure which has been documented in the literature, which is now known to be due to a combination of a reduction in expression of SSS and BE, together with an over-expression of other isoforms of SSS and BE which have different roleε in the starch asεembly process. These new findings from our ongoing work has led to our efforts to alter gene-expression levels of SSS and BE in cereal endosperm in order to effect a new starch εtructure in corn. Furthermore, since we now know that different isoforms have different roles in starch asεembly there iε the opportunity of transforming corn with forms of SSS and BE that have other different properties which will effect a change in the fine structure of the starch (eg using SSS and BE from other diverse biological

) sources of plants, animals, bacteria and fungi).

CONCLUSIONS FROM THE EXPERIMENTS

Our resultε have demonεtrated that the failure of the starch εyntheεis rate in cereal endoεperm to increase with temperature above 20-25°C is due to the fact that in the temperature range of 25-30°C the rate of starch syntheεiε iε controlled by the activitieε of the SSS and BE enzymeε. The rateε of the reaction catalysed by SSS and BE fail to increase with temperature above 25°C. In addition, knockdown of SSS activity may further reduce yields in maize above 37°C and in wheat above 30°C. Crops εuch aε wheat, barley and maize frequently experience temperatures above 25°C during their grain filling period. This invention increases the temperature optimum for the grain filling rate above 25°C by relieving the limitations imposed by the reaction properties of SSS, BE or both. The enzyme glycogen synthase has a radically different response to temperature and iε an ideal source of enzyme for increasing starch synthesis at temperatures that exceed the temperature optima of SSS and or BE. Changing the temperature optimum for starch synthesis in plants increases plant yield as well aε changing other important propertieε εuch as fruit texture and sweetneεε. Furthermore the alterationε in expression levels of SSS and BE using both sense and antisense constructs resultε in an alterations in the fine- εtrucure of the εtarch produced in the recipient plantε. In addition, effecting this change in ratioε of enzyme expreεsion using temperature stable enzymes results in a more stable and _{2 8} ^{rt l}

defined type of starch quality that is unique to the recipient plants.

USE OF GLYCOGEN SYNTHASE

A gene encoding a glycogen synthaεe (GS) enzyme may be extracted from bacteria or animalε and introduced into a donor plant by tranεformation. The temperature εenεitivity of εtarch εyntheεiε may be improved by transforming plant genomes with a gene encoding glycogen synthase. Referring to Figure 15 herewith, it can be seen that the activity of the glycogen synthase enzyme continues to increase with temperature at least to around 40°C, well in excess of the temperature maxima of the other plant enzymes, soluble starch synthase and branching enzyme, associated with starch syntheεiε. This data confirms the εtatus of glycogen synthase as an enzyme with high Q10. It is also mentioned that the glycogen synthase temperature stability is better than any of the corn-derived enzymes from even the best of the germplasm which haε been εcreened for thiε property.

Glycogen εynthaεe catalyεes the same reaction aε εoluble εtarch εynthase [ADPglucose:1, -α-D- glucan-4-α-D-glucoεyltransferase (E.C. 2.4.1.21)] which catalyses the sequential addition of glucose donated from the sugar nucleotide ADP-glucose to a glucan chain using an α-1,4 linkage. Other sources of this α-1,4 linkage reaction exist in nature as the enzyme glycogen synthaεe, notably from bacteria [ADPglucose: 1,4-a-D-glucan 4-α-D-glucosyl- tranεferaεe (E.C. 2.4.1.21)] and animalε [UDP-glucoεe: glycogen 4-α-D-glucoεyltranεferaεe (E.C. 2.4.1.11)]. Again the reaction involved is the sequential addition of glucose donated from a nucleotide sugar to a glucan chain via an α-1,4 linkage.

While not wishing to be bound by this particular explanation, it is believed that the different stability to increased temperature between soluble starch synthaεe and glycogen synthase is connected with changes in the structure of glucan chains at higher temperatureε. It iε generally agreed that the structure of εtarch (at leaεt in εolution) iε a "εtatistical helix", that is, it existε as a dynamic mixture of structured helices and random coils. At temperatures much in excess of 20-25°C, the hydrogen bonding between the glucan double helices beginε to break down, favouring the less organised random coil conformation. It is hypothesised that the active site of soluble starch synthase lies on the glucan chain when present as a helix whereas the glycogen εynthase site of action iε found on the random coil εtructure. By thiε reaεoning, their enzymatic reaction may be the εame but their temperature responεeε dramatically different.

The most favoured sourceε of the glycogen synthase gene for use in this invention are bacterial rather than animal sources for the following reasons:

(1) the bacterial glycogen synthase and plant soluble starch εynthase both use ADPG, whereas the animal GS enzyme uses UDPG;

(2) the bacterial GS and plant SSS enzymes do not ₃₀

have any phosphorylation sites for activation, whereas the animal enzyme does; and, (3) the animal GS enzyme requireε glucose-6-phoεphate aε a co-factor and is allosterically activated, whereaε the plant SSS and bacterial GS enzymeε are not.

For these reasons the bacterial GS gene is preferred. The nucleotide and amino acid sequenceε are known from the literature, for example, E.coli GenBank/EMBL #J02616 (Kumar et.al., J.Biol.Chem. 34, 16256-16259 (1986); rabbit εkeletal muεcle (Zhang et.al., FASEB J. 3, 2532-2536 (1989); and, human muscle (Browner et.al., Proc,Natl.Acad.Sc. 86, 1443-1447 (1989). The bacterial and animal GS εequenceε are not homologouε. The εtructural genes for the bacterial GS are mapped to pOP12 in E.coli and glycogen synthase maps to glgA. Nucleotide εequencing further refined the poεition of glgA. The tranεlation εtart point of glgA is known to be immediately following glgC and the nucleotide sequence determined. The NH₂ sequence was known εo that the actual start of the glgA gene waε unambiguouεly determined as well as confirming the direction of transcription. The deduced amino acid sequence εhowε complete homology with the known NH- εequence and with the known amino acid εequence. Different bacterial enzymes show 90% homology. There is complete agreement between the reported and deduced amino acid sequences for the enzyme. Cells transformed with the gene produce a polypeptide that has sequence homology with the known amino acid sequences.

Figure 18 shows the nucleotide sequence for E.coli glycogen synthase as retrieved from EMBL #J02616. It is not a large protein: the structural gene is 1431 base pairs in length, specifying a protein of 477 amino acids with an estimated molecular weight of 49,000. It is known that problems of codon usage can occur with bacterial geneε inεerted into plant genomeε but thiε is generally not so great with E.coli geneε aε with those from other bacteria such aε thoεe from Bacilluε. Glycogen synthase from E.coli has a codon usage profile much in common with maize genes but it is preferred to alter, by known procedures, the sequence at the translation start point to be more compatible with a plant conεensus sequence: glgA G A T A A T G C A G cons A A C A A T G G C T

The GS gene construct requires the presence of an amyloplast transit peptide to ensure its correct localisation in the amyloplast. It is believed that chloroplast transit peptides have similar sequences but other potential sourceε are available such aε that attached to ADPG pyrophosphorylase (Plant Mol. Biol. Reporter (1991) 9, 104-126). Other potential transit peptides are those of small subunit RUBISCO, acetolactate synthase, glyceraldehyde-3P- dehydrogenase and nitrite reductaεe. For example,

Conεenεus sequence of the transit peptide of small subunit RUBISCO from many genotypes has the sequence:

MASSMLSSAAV—ATRTNPAQAS MVAPFTGLKSAAFPVSRK QNLDITSIA SNGGRVQC and the corn εmall subunit RUBISCO has the sequence:

MAPTVMMASSAT-ATRTNPAQAS AVAPFQGLKSTASLPVARR

SSRSLGNVA SNGGRIRC The transit peptide of leaf starch synthase from corn has the sequence:

MA ALATSQLVAT RAGLGVPDAS TFRRGAAQGL RGARASAAAD

TLSMRTASARA APRHQQQARR GGRFPSLWC The transit peptide of leaf glyceraldehyde-3P- dehydrogenase from corn has the sequence:

MAQILAPS TQWQMRITKT SPCATPITSK MWSSLVMKQT

KKVAHSAKFR VMAVNSENGT The putative tranεit peptide from ADPG pyrophoεphorylaεe from wheat has the sequence: RASPPSESRA PLRAPQRSAT RQHQARQGPR RMC

It is poεεible however to express the glycogen synthase constitutively using one of the well-known constitutive promoters such as CaMV35S but there may be biochemical penalties in the plant resulting from increased starch deposition throughout the entire plant. Deposition in the endosperm iε much preferred.

Poεεible promoterε for uεe in the invention include the promoterε of the εtarch synthase gene, ADPG pyrophosphorylaεe gene, and the εucrose εynthaεe gene.

Figure 19 herewith illuεtrateε the conεtruction of two glycogen εynthase vectors for uεe in thiε invention to transform tomato. Vectorε are deεcribed with either conεtitutive or fruit-εpecific promoterε. The tranεit peptide from tomato ribuloεe-biε-phoεphate carboxylaεe iε incorporated to direct the product to the plaεtid. USE OF BACTERIAL BRANCHING ENZYME Another embodiment of thiε invention iε to uεe a temperature εtable form of branching enzyme which can be obtained from bacteria. Branching enzyme [1, -α-D-glucan: 1,4-α-D-glucan

6-α-D-(1,4-α-D-glucano) tranεferaεe (E.C. 2.4.1.18)] converts a ylose to amylopectin, (a εegment of a 1,4-α-D-glucan chain iε tranεferred to a primary hydroxyl group in a εimilar glucan chain) εometimeε called Q-enzyme. Like soluble starch synthaεe, thiε reaction alεo haε temperature-dependent propertieε in plants, presumably because of the same molecular mechanisms of helix-to-chain transitionε. It is reasonable to believe that the bacterial BE enzyme will behave εimilarly.

The most favoured sources of the branching enzyme gene for use in this invention are bacterial although plant enzymes can also be used (rice endosperm, Nakamura etal., Physiologia Plantarum

84, 329-335 (1992); pea embryo, Smith, Planta 175, 270-279 (1988); maize endoεperm, Singh and Preiss, Plant Physiology 79, 34-40 (1985); Vos- Scherperkeuter etal.. Plant Physiology 90, 75-84 (1989)). The nucleotide and amino acid sequences for bacteria are known from the literature (Kiel JAKW et. al, 1991, Mol Gen Genet, 230(1-2) :136-144) . The structural genes for the bacterial BE are mapped to pOP12 in E.coli and branching enzyme maps to glgB.

The BE gene construct may require the presence of an amyloplast transit peptide to ensure its correct localisation in the amyloplast, as discussed previously for the glycogen εynthaεe gene construct.

It is posεible to expreεε the branching enzyme constitutively using one of the well-known constitutive promoters such as CaMV35S but there may be biochemical penalties in the plant reεulting from increaεed εtarch depoεition throughout the entire plant. Deposition in the endosperm is much preferred.

Posεible promoterε for uεe in the invention include the promoterε of the starch synthase gene,

ADPG pyrophosphorylaεe gene, and the sucrose synthase gene.

Branching enzyme vectors may be used to transform corn, with either constitutive or endosperm-specific promoters. The transit peptides from corn amyloplaεt-εpecific enzymeε may be incorporated to direct the product to the plastid.

EFFECTS ON STARCH FINE STRUCTURE

Although this invention is directed primarily to improvement of the deposition of starch at elevated climatic temperatures, alteration of starch deposition inevitably leadε to alteration of εtarch fine εtructure.

In cereal cropε, changing the ratioε and activitieε of SSS and BE and/or the εource of the enzymes (eg replacing maize SSS with pea SSS) alters the fine-branching εtructure of the starch. For example, the fine branching structure of starch is determined by the overall activitieε of the various isoformε of the SSS and BE enzymeε being expressed during εtarch deposition in the developing endosperm. Altering the ratios of these isoformε may be achieved by tranεformation techniqueε in which εome of the natural enzyme activitieε are repreεεed whilst others are over-expressed in a manner analogous to the changes reported herein for the starch mutants of corn.

EFFECTS ON TEXTURE

Improved starch deposition also leads to alteration of the texture of cropε such as tomatoes because increased amounts of starch in the fruit would increase the total solids content.

This effect would be significant in tomatoes which are grown for processing into paste. The quality of paste produced from processed tomatoes is in part related to the viscoεity of the product which iε uεually determined by the Bostwick flow rate, reduced flow rate being deεirable. The factors that interact to give a thicker product with reduced flow rate are complex, involving interactions between insoluble and εoluble components. It is important to note that the characteristics of components in whole fruit will change during processing because of enzyme action and chemical changes brought about by heating which is involved in tomato procesεing by the εo-called "hot-break" method.

The conεistency of hot break paste is likely to be improved by increasing the level of insoluble solidε in the whole fruit uεed in processing.

Increased levels of soluble and inεoluble solids in procesεing tomatoes has been an object of plant breeders for many years.

Soluble solidε are the solutes in the tomato PC17GB92/01881

36

εerum and conεiεtε primarily of carbohydrateε. Paste is normally sold on the basis of its natural tomato soluble solids (NTSS) content. Because the sugars are the major contributors to NTSS, a higher sugar content contributes to a higher yield of paste per tonne of tomatoes. The correlation between NTSS and total solidε (TS) iε very high, although the relationεhip varieε amongεt tomato cultivars. High NTSS levels in ripe fruit may be an indirect measure of the starch component of the insoluble εolids during fruit development. Sugar content is a critical component of the flavour of tomatoes.

Insoluble solidε (IS) consist mainly of the polysaccharideε in the cell wall. Residual starch will alεo contribute to the IS although, in normal ripening, thiε formε a small component. The IS/TS ratio partially determines the conεistency of tomato productε. Where high conεistency iε required, a greater quantity of IS improves the product quality. IS are measured as both water-insoluble solidε (WIS) and alcohol-insoluble εolidε (AIS). the AIS quantities are greater than those for WIS because smaller polysaccharides are less soluble in 80% ethanol than in water.

It is believed that, by the method of this invention, the increased amounts of starch may very well accumulate in plastid granules and are, therefore, unlikely to contribute much to improved consistency. However, the heating involved in the hot break procesε and in the concentration εtep of hot and cold break productε is likely to burst the granules and partially solubiliεe the εtarch, reεulting in increaεed viεcoεity. Enhanced starch εyntheεis is also likely to elevate the import of carbohydrate into the fruit which may also result in enhanced levels of soluble solids in the ripe fruit. This may also be conεidered aε advantageouε in fruit intended for the freεh fruit market.

In εummary, then, improved insoluble solids leads to improved consiεtency of paste and higher yields of paste per tonne of tomato processed, whereas increased εoluble εolids will result in higher yields of paste per tonne of tomato procesεed and improved sugar component of flavour.

In tomato, the free sugars are almost entirely fructose and glucose. Sucrose is present but rarely exceeds 1% of the dry weight. Starch and structural polysaccharideε are the major forms of storage of imported carbon. Starch levels increase in the early stages of fruit development, followed by a decrease to virtually zero by ripeness. Thus, in ripe fruit, hexoses (glucose and fructose) are the primary component of the soluble solids and account for about 50% of the fruit dry weight.

Thus, from the known facts about the ripening process it is believed that increasing the soluble solidε iε advantageouε and that thiε increase may be obtained by transformation with a gene encoding glycogen synthaεe.

Our invention is thus applicable to plants whose value is due to the texture and/or sweetness and/or taste of the fruit. For example fruits such aε tomato, melon, peach, pear, etc where the accumulation of starch is an early event in fruit development but is later degraded during fruit ripening releasing sugars. In this case the desirable trait is extra starch deposition which is useful in providing enhanced fruit texture and sweetnesε aε well aε increaεed thickening quality during cooking (eg, during ketchup/catsup manufacture from tomato).

OTHER USEFUL EFFECTS

Starch deposition in the developing pollen grainε of cerealε is an esεential prerequiεite for pollen viability. If starch syntheεiε is impaired by high temperatures in the developing pollen cells, reduced pollen viability will result. The insertion, according to the invention, of thermally stable variants of the SSS and/or BE and/or GS enzyme genes into cereal pollen increases pollen viability and seed set is lesε impaired by exposure to high temperatures.

Starch synthesiε in leaf chloroplasts is also limited by the thermal lability of the SSS and BE enzymes. Photosynthetic rates in green leaf tissue are dependent in part on the ability of the cells to convert fixed carbon into starch and εucroεe. When εtarch synthesis becomes limited at high temperatures because of rate limitation by SSS or BE, there is an accumulation of metabolic intermediateε in the chloroplaεt, causing feedback inhibition of ribulose bisphosphate regeneration, reducing the overall carbon fixation by photoεynthesiε. Inεertion of thermally εtable variantε of the enzymeε, or additional copieε of the relevant gene(ε), increaεeε the photosynthetic rate at high temperatures. This will increase the yield of all crops whose yield is dependent on the photosynthetic rate of the source leaves during at leaεt a part of their life cycle. Examples include cereal cropε εuch aε maize, sorghum, wheat, barley and rice; root crops εuch as potato, turnip, yam and cassava; sugar crops such as beet and cane; oilεeed cropε εuch aε rapeεeed, εunflower, oil palm, coconut, linεeed and groundnut; fruitε εuch aε appleε, pears and bananas; and meal crops such as εoya, beanε and peas. The yields of all of theεe cropε may be limited by εource activity at particular timeε in their life cycle.

Our invention iε alεo applicable to plantε whoεe yield dependε on groεs biomass accumulation and which are limited by photoεynthetic rate. Examples are forage crops such as grasses, ryegrass, forage maize and alfalfa; treeε grown for wood, pulp or ethanol production and vegetables such as cauliflower, cabbage and sprouts.

CROSS-BREEDING

In one specific aspect, our invention is a method of producing a novel εubεtantially homozygouε maize (corn) line having εuperior εtarch depoεition properties which compriseε, (i) identifying a range of potential donor plants which are sexually compatible with a recipient maize plant and screening producing εoluble εtarch synthaεe (SSS) enzyme or branching enzyme (BE) to determine the heat εtability of the reaction of at leaεt one of εaid enzymeε; (ii) identifying a plant producing an enhanced SSS or BE enzyme that iε εignificantly more heat-stable for the enzyme reaction than the corresponding enzyme in the _{4 Q}

recipient maize plant; (iii) crosεing the identified plant with the recipient maize plant; (iv) selecting from among the progeny thoεe expressing the heat stable enzyme reaction of the donor; and, (v) breeding therefrom εo as to produce a novel substantially homozygous maize line having an enhanced rate of starch depoεition.

Genotypeε with a measurable enhancement in SSS or BE activity or altered characteristicε with reεpect to temperature, are introduced into a back-croεεing programme with a commercial maize inbred. Progeny are εelected on the basis of genetic similarity to the commercial line, using RFLP's (reεtriction fragment length polymorphisms), but with the desired SSS or BE characteristics. Selected progeny are entered into further back-crossing against the commercial line. The end result is a new maize inbred line, genetically vary similar to the parental line, having enhanced SSS or BE activity. The temperature optimum of starch synthesiε iε meaεured before including the new line in hybrid production.

Another embodiment of our invention is a method of producing a novel substantially homozygous maize (corn) line having superior starch deposition properties which comprises, (i) identifying a range of potential donor plants which are sexually compatible with a recipient maize plant and screening for plants with high rates of grain starch εynthesiε at elevated temperatures;

(ii) identifying a plant producing an enhanced rate of starch deposition that is significantly more heat-stable than the corresponding rate in the recipient maize plant; (iii) crosεing the 4/09144 ₄₁

identified plant with the recipient maize plant; (iv) εelecting from among the progeny those expressing the high rate of εtarch depoεition; and, (v) breeding therefrom εo aε to produce a novel

5. εubεtantially homozygouε maize line having an enhanced rate of starch deposition.

Genotypes with a meaεurable enhancement in εtarch depoεition rateε with reεpect to temperature are introduced into a back-crossing programme with

10 a commercial maize inbred. Progeny are selected on the basis of genetic similarity to the commercial line, using RFLP's (restriction fragment length polymorphismε) , but with the desired SSS or BE characteristics. Selected progeny are entered into

15 further back-crossing against the commercial line. The end result is a new maize inbred line, genetically vary similar to the parental line, having enhanced SSS or BE activity. The temperature optimum of starch synthesis is measured 0 before including the new line in hybrid production.

EXAMPLE 1 Sexual Crosεing The plantε selected for screening were maize 5 plants: either commercial maize varieties or varieties from more exotic collections. The material selected for εcreening waε from amongst other Zea germplasm, for example, Zea tripεacum, perennis, diploperennis, luxurians, parviglumis, 0 mexicana and mays. Many thousands of potential donors exiεt throughout the world where maze is grown as a cultivated crop or where it existε in the wild plant population, for example in South and Central America and Africa. Moεt of the Zea family can be inter-bred by traditional plant breeding methodε.

One characteristic sought for use in this invention was an increaεe (or minimal loss) in activity from 25 to 35°C. A Q._Q value for each line was obtained by a method which is hereinafter described. The frequency distribution for all the germplasm assayed in the screen (Figure 20) showε that the range of variation for thiε trait iε quite narrow. However, our screen succeeded in locating a few rare occurrences from the extremely wide selection of Zea germplaεm. Acroεε all the germplaεm the overall average drop in activity between 25 and 35^PC waε around 40% ( Q_{1 Q} - 0.61) with a range of 65% to 0% drop in activity (Q._Q range of 0.35 to 1.00). A εelection of sixteen inbreds used in commercial hybrids showed a narrower range of 41% to 20% drop in activity (Q_1Q of from 0.59 to 0.80 with an average of 0.70) indicating that conventional maize breeding which involves selection of yield-for-moisture may posεibly have been effective in weeding-out the very worεt forms of the enzyme. However, there is clearly room for further improvement in several inbreds and hence it is posεible to gain potential yield benefit in commercial hybrids.

The very best forms of SSS identified (from a screen of nearly 1,000 sourceε of Zea germplaεm) where in 4 exotic lines from Peru (Lima 38 and Lima 45) and Mexico (Guanajuato 13) and teosinte with Q10 as high as 1.0. Three of these lines were obtained from the Plant Introduction Centre, Iowa State University: Numbers P1515021, Ames 8545, PI490879 and one teosinte line obtained from Dr John Deobley Zea mays εubεp. Mexicana, Doebley 479. The temperature/activity profiles (Figure 21) of these forms of the enzyme(s) show a dramatic difference from one commercially valuable inbred line (UE95) which was included for comparison.

These two exotic lineε have been crossed into commercial inbred germplasm to produce F, hybrids.

GENE MANIPULATION

A donor gene may alεo be introduced into a recipient plant by tranεformation.

When genetic manipulation techniques are employed in this invention there are at least four possibilities:

(1) Increase the amount and activities of the enzyme SSS or BE or GS or any combination thereof in a recipient plant, such as a commercial maize line or population by the insertion of extra gene copies of the SSS, BE or GS enzymes. The source of these extra copies may be the recipient line itself as the technique would simply increase the amount of enzyme available in the grain rather than the changing of the properties of the enzyme(ε). The gene promoterε and other regulatory sequences may also be altered to achieve increased amounts of the enzyme in the recipient plant.

(2) The insertion of a gene or genes specifying SSS and/or BE and/or GS enzymes with activities which increase with temperature up to 30°C. Achieving this requires the identification of a source of the thermally stable enzyme gene. However, the use of genetic manipulation techniques for introduction of the new genes places no restriction of the source to sexually compatible species and genetic material may be obtained from any source of the SSS, BE and GS enzymes having the desired temperature/activity characteristics. Enzymes from plant species other than cereals, particularly plants which grow well in tropical or sub-tropical regions will most likely exhibit greater temperature εtability in the SSS and BE enzymes than maize. The exogenous genetic material may altered by genetic engineering to give the desired characteristicε to the enzymeε. Techniqueε are known in εo-called "protein engineering" which can alter the characteriεticε of an enzyme. (3) Change the ratioε of activitieε of the isoforms of enzymes SSS or BE or any combination thereof in a recipient plant, such as a commercial maize line or population by the insertion of extra gene copies of the SSS, BE enzymeε and/or by insertion of anti-sense gene constructs. The source of these extra copies or antisense constructs may be the recipient line itself as the technique would simply increase or decrease the amount of enzyme available in the grain rather than the changing of the properties of the enzyme(s). The gene promoters and other regulatory sequences may also be altered to achieve increased amounts of the enzyme in the recipient plant.

(4) The insertion of a gene or genes specifying SSS and/or BE and/or GS enzymes with activitieε which effect a change in the fine εtructure of the εtarch. Achieving this requires the identification of a source of the enzyme gene. However, the uεe of genetic manipulation techniqueε for introduction of the new genes places no restriction of the εource to εexually compatible species and genetic material may be obtained from any source of the SSS, BE and GS enzymes having the desired characteristicε. Enzymeε from plant εpecies other than cereals, particularly plants which grow well in tropical or sub-tropical regions will most likely exhibit altered starch specificitieε in the SSS and BE enzymeε than maize. The exogenouε genetic material may altered by genetic engineering to give the desired characteristics to the enzymes. Techniques are known in so-called "protein engineering" which can alter the characteristicε of an enzyme.

ISOLATION OF A DONOR GENE

The donor gene may be isolated from a suitable biological organism (including plantε, fungi, bacteria or animal cellε) after εcreening a range of potential donor organisms producing soluble starch synthase (SSS) enzyme, glycogen synthase (GS) or branching enzyme (BE) to determine the heat stability of the reaction of at least one of said enzymes. (In bacteria or animalε, glycogen iε depoεited rather than εtarch and εo theεe organisms contain glycogen synthases and branching enzymeε).

EXAMPLE 2 Screening for source material

We have devised a means (an enzyme-activity based screen) of identifying heat-stable forms of the enzymes. When the natural Zea mays temperature senεitive enzymes are replaced with temperature stable enzymes from another organism, provided the replacement enzymes are not disrupted by low temperatures in the environment, maize plants with higher rates of εtarch synthesiε at temperatureε above 25°C will reεult.

The εcreen uεed in this invention identified enzyme forms with increased (or minimal losε in ) activity from 25 to 35°C. However, the characteristics sought for use in the invention can be any one, or several in combination, of the following characteriεticε:

(1) a high increaεe in activity between 20 and 25°C;

(2) increaεed (or minimal losε in ) activity from 25 to 30°C;

(3) increaεed (or minimal loss in ) activity from 30 to 35°C;

(4) increased stability to "knockdown", where knockdown is defined as an irreversible loss in activity caused by elevated temperatures; and,

(5) instrinεically high activity at 20°C.

EXAMPLE 3 Enzyme Activity Assay In order to assess the suitability of the native enzymes of potential gene donor plants, it is necesεary to be able to aεεay for the thermally stable enzymes. We have devised the following procedure. 1. Homogeniεe grain or endoεperm tiεsue and extract protein into a buffer solution which maintains the SSS and BE enzymes in an active form. 2. Asεay SSS activity by adding 50 l of the enzyme extract to 25μl of primer (glycogen, _{4 ?}

amylopectin, εtarch) and 100/1 of a buffer εolution of the following final concentrationε: lOOmM bicine, 25mM potaεεium chloride, 5mM EDTA, and lOmM reduced glutathione. 3. Begin the asεay by adding 14C-ADP-glucoεe.

Stop the aεεay after a defined time at a defined temperature by adding 1ml of O.lM sodium hydroxide.

Precipitate the primer by bringing the mixture to

75% methanol concentration. Centrifuge and recover the primer pellet and redisεolve in sodium hydroxide and reprecipitate in 75% methanol.

Repeat the repreciptation procedure.

4. Following the second wash, dissolve the primer in 1M hydrochloric acid at 100°C. Cool and add the mixture to εcintillation fluid and measure the

14 C-ADP-glucoεe transferred to the primer.

5. BE activity is assayed by adding 50μl of the enzyme extract to a mixture containing citrate buffer, AMP, and phosphorylase A. 6. Start the reaction by adding an aliquot of

14 C-glucose-1-ρhosphate. Stop the reaction after a defined time at a selected temperature by adding

O.lM sodium hydroxide.

7. Add a polysaccharide (amylose, glycogen, amylopectin, starch) as a carrier and precipitate the polysaccharide with 75% methanol. Centrifuge and wash the polysaccharide precipitate and dissolve and count the scintillation as described in step 4 above. BE activity iε defined aε the εtimulation achieved by the enzyme in the incorporation of 14C-glucoεe from the glucose-1-phosphate into the polysaccharide by the phosphorylaεe A. EXAMPLE 4 Iεolation of soluble starch synthase (SSS) and branching enzyme (BE) genes

Using standard cloning techniques, the SSS and BE genes may be isolated.

The source of the genes was a US sweet-corn line of Zea mays, from which the enzyme protein was purified. Endospermε from the maize line were hσmogeniεed in a buffer which maintainε the SSS and BE in active form. The enzymeε were partially purified by ammonium εulphate fractionation, followed by DEAE Sepharoεe chromatography, followed by FPLC using a Superose gel filtration column, followed by FPLC using a Mono Q anion exchange column. The Superose column allows separation of SSS from BE activity. Further purification of individual isoforms was achieved by hydroxyapatite, cation exchange FPLC or isochromatofocussing.

Purification of the SSS from maize (Silver Queen) has been achieved (Wasserman etal. Plant Phyεiology, (in preεs)) by a combination of ammonium sulphate precipitation, ion exchange chromatography, affinity chromatography (Affi-Gel Blue) and two FPLC εtepε, Mono-Q and Superose-12. This reεults in up to 5,000-fold purification with yields up to 5%. The SSS polypeptide was a single subunit of molecular weight 86kD. Other SSS polypeptides were present in a US dent inbred line at around 70kD and 105kD molecular weight. Ammonium εulphate precipitation of SSS I iε best achieved using 40% ammonium εulphate which produces a translucent SSS-enriched pellet which is next dialysed and further fractionated using DEAE-cellulose ion-exchange chromatography. Theεe εtepε increaεe εpecific activitieε by up to 50-fold. The affinity chromatography εtepε rely on the ADP attached to a Sepharoεe matrix either through the N-amino group or through The ribose hydroxyl group. SSS I is readily eluted with 2mM ADP at pH 8.5. Specific activitieε are increased by up to 10-fold uεing the affinity chromatography step with a yield of around 15%.. SSS iε next purified by Mono-Q FPLC εtepε with elution of activity at low potaεεium chloride (not more than 200mM KC1) .

Purification of the SSS and BE enzymes from the US inbred line identified three SSS and three BE isoformε. Figure 22 εhowε the data for SSS; BE behaveε εimilarly. Preliminary investigations have suggested that these isoformε have εlightly different temperature optima of activity and also slightly different temperature thresholds for knockdown. In the final purification step the SSS or BE preparations were loaded on to SDS PAGE gels. The bands corresponding to the SSS or BE polypeptides were cut out and eluted. The pure polypeptide was then used as an antigen to generate polyclonal antibodies in a rabbit. The antibodies were then tested for specificity to the SSS or BE polypeptides. N-terminal amino acid sequences were also obtained from the polypeptides.

Full sequencing of the maize polypeptides is continuing. Amino acid sequencing of the maize SSS polypeptide has provisionally yielded the following partial sequence: Ala-Ala-xxx-Arg-Lyε-Ala-Val-Met- Val-Pro-xxx-Gly-xxx-Aεn-Arg-Glu-Phe-Val-Lyε-Tyr-Leu -Phe-xxx-Met/Phe-Ala-Gln. The final εequence may be compared to the amino acid sequence of pea SSS I and SSS II published by Dry e_t ajL (1991, Plant Journal, 2:193-202).

The antibodies may be used to screen a maize endosperm cDNA library for clones derived from the mRNAs for SSS or BE in an in vitro transcription/ translation syεtem.

The cDNAε thuε derived may be uεed to probe a maize genomic library and the maize SSS and BE genomic DNAε may be iεolated. In addition

N-terminal amino acid sequence information for SSS and BE may be used to generate oligonucleotide probes. These probes may be used to screen the maize genomic library and the maize SSS and BE genomic DNAs may be isolated.

TRANSFORMATION

(i) Insertion of extra copies of the gene Maize genomic DNAs isolated as above may subsequently be transformed into either protoplasts or other tissues of a maize inbred line or population. The existing gene promoters ensure that the extra genes are expreεsed only in the developing endosperm at the correct developmental time. The protein sequenceε likewiεe ensure that the enzymes are inserted into the amyloplast.

Transgenic maize plantε are regenerated and the endoεperms of these plants are tested for increaεed SSS and BE enzyme activity. The kernelε are alεo tested for enhanced rate of starch syntheεiε at different temperatureε. The plantε are then included in a breeding programme to produce new maize hybrids with higher rates of _{ιnM λ λ} PCT/GB92/018 1 09144

51

εtarch syntheεiε at temperatureε above the normal optimum.

(ii) Inεertion of geneε εpecifying SSS and/or BE with higher temperature optima for activity. Thiε iε alεo achieved by standard cloning techniques. The εource of the temperature-εtable formε of the SSS or BE or GS geneε iε any organism that can make starch or glycogen. Potential donor organisms are εcreened and identified aε deεcribed above. Thereafter there are two approaches:

(a) via enzyme purification and antibody/sequence generation using the protocol described above.

(b) using SSS and BE and GS cDNAs as heterologous probes to identify the genomic DNAs for SSS and BE and GS in libraries from the organism concerned. The gene transformation, plant regeneration and testing protocols are as described above. In this instance it is necessary to make gene constructs for transformation which contain the regulatory sequences from maize endosperm SSS or BE or another maize endosperm starch synthesis pathway enzyme to ensure expression in endosperm at the correct developmental time (eg, ADPG pyrophosphorylase). One specific example of thiε is with the bacterial glycogen synthase enzyme which we have found to be essentially tolerant of temperatures up to 40°C. The nucleotide and amino-acid sequences of glycogen synthase are known (i) from E Coli GenBank/EMBL #J02616 (Kumar et al, J Biol Chem 34 16256-16259 (1986)). (ii) from rabbit skeletal muscle (Zhang et al, FASEB J 3 2532-2536 1989)), and (iii) from human muscle (Browner et al, Proc Nat Acad 5cl 86 1443-1447 (1989)). Gene constructs used to transform plants requires the regulatory sequences from maize endosperm SSS or BE or another maize endosperm starch syntheεiε pathway enzyme to ensure expression in endosperm at the correct development time (eg, ADPG pyrophosphorylaεe) . Furthermore the gene conεtructε alεo requires a suitable amyloplaεt tranεit-peptide εequence εuch aε from maize endoεperm SSS or BE or another maize endoεperm starch syntheεis pathway enzyme to censure expresεion of the amyloplaεt at the correct developmental time (eg, ADPG pyrophosphorylase) . Genetic protein engineering techniques may also be used to alter the amino acid sequence of the SSS or BE or GS enzymes to impart higher temperature optima for activity. The genes for SSS and/or BE and/or GS may be cloned into a bacteria which relies on these enzymes for survival. Selection for bacteria surviving at evaluated temperatures enables the isolation of mutated thermostable enzyme forms. Transformation of maize with the altered genes is carried out as described above.

Genetic protein engineering techniques may alεo be used to alter the amino acid sequence of the maize SSS or BE enzymes to impart higher temperature optima for activity. The geneε for SSS and/or BE may be cloned into bacteria relies on the these enzymes for survival. Selection for bacteria surviving at elevated temperatures enables the isolation of mutated thermostable enzymes forms. Transformation of maize with the altered genes is carried out as described above. (iii) Changing the ratios of activitieε of the isoforms of enzymes SSS or BE.

This is also achieved by standard cloning techniques. The source of the SSS or BE genes is maize using the protocol described above. Plants are then transformed by insertion of extra gene copies of the isoforms of SSS, BE enzymes and/or by inεertion of anti-sense gene constructs. The gene promoters and other regulatory sequenceε may alεo be altered to achieve increaεed amountε of the enzyme in the recipient plant.

(iv) insertion of a gene or geneε εpecifying SI. and/or BE and/or GS enzymes with activities which effect a change in the fine structure of the starch.

Thiε is also achieved by standard cloning techniques. The source of the special forms of the SSS or BE or GS genes is any organism that can make starch or- glycogen. Potential donor organisms are screened and identified aε deεcribed above. Thereafter there are two approaches: (a) via enzyme purification and antibody/sequence generation using the protocol described above. (b) using SSS and BE and GS cDNAs as heterologouε probeε to identify the genomic DNAε for SSS and BE and GS in librarieε from the organiεm concerned. The gene transformation, plant regeneration and testing protocols are as deεcribed above. In thiε instance it is necessary to make gene constructε for transformation which contain the regulatory sequences from maize endosperm SSS or BE or another maize endosperm starch synthesiε pathway enzyme to enεure expreεεion in endoεperm at the correct developmental time (eg, ADPG pyrophoεphorylaεe) .

Claims

1. A method of producing a plant with altered starch synthesising ability comprising stably incorporating into the genome of a recipient plant one or more than one donor gene specifying an enzyme involved in a εtarch or glycogen bioεynthetic pathway.

2. A method aε claimed in claim 1 in which the plant haε an improved capacity to produce starch at elevated or lowered temperature.

3. A method as claimed in claim 1 or claim 2 in which the plant has an ability to synthesise starch with an altered fine structure.

4. A method as claimed in any of claims 1 to 3 in which the temperature optimum of εtarch synthesis is increased or decreased.

5. A method as claimed in any of claims 1 to 4 in which at least one of the donor genes is derived from a plant.

6. A method as claimed in claim 5 in which the donor gene is derived from a plant of the specieε Zea mayε, Zea diploperennis, Zea luxurianε, Zea perennis, Zea tripsacum, Zea parviglumiε, Zea mexicana or teoεinte.

7. A method aε claimed in claim 6 in which the donor gene iε derived from a plant of the Zea mayε varietieε Lima 38, Guanajuato 13, Lima 45, Doebley 479 or teoεinte 154. 55

8. A method aε claimed in any of claimε 1 to 7 in which at leaεt one of the donor genes is derived from a bacterium.

9. A method as claimed in any of claims 1 to 8 in which at leaεt one of the donor geneε iε derived from a funguε.

10. A method aε claimed in any of claimε 1 to 9 in which at leaεt one of the donor genes is derived from an animal cell.

11. A method aε claimed in any of claimε 1 to 10 in which the donor gene εpecifies εoluble starch synthase (E.C. 2.4.1.21) and/or branching enzyme (E.C. 2.4.1.18) and/or glycogen synthase of bacterial origin (E.C. 2.4.1.11) or of animal origin (E.C. 2.4.1.21).

12. A method as claimed in any of claims 1 to 11 in which at least one of the donor genes specifieε a modified allelic form of the enzyme.

13. A method aε claimed in any of claims 1 to 12 in which at least one of the donor genes specifieε a mRNA antiεenεe to the mRNA encoded by the wild-type gene.

14. A method aε claimed in any of claimε 1 to 13 in which at least one of the donor genes is derived from a sexually compatible donor plant and is inεerted into the recipient plant by sexual croεεing of donor and recipient plantε.

15. A method as claimed in any of claim 1 to 14 in which at leaεt one of the donor geneε iε incorporated into the recipient genome by genetic tranεformation.

16. A method aε claimed in any of claimε 1 to 15 in which the recipient plant iε of the family Gramineae.

17. A method as claimed in claim 16 in which the recipient plant is of the specieε Zea mayε.

18. A method as claimed in any of claimε 1 to 15 in which the recipient plant iε a tomato.

19. A plant having one or more than one donor gene εpecifying an enzyme involved in a εtarch or glycogen biosynthetic pathway stably incorporated into its genome such that its ability to synthesise starch is altered.

20. A plant as claimed in claim 19 which has an improved capacity to produce starch at elevated or lowered temperature.

21. A plant as claimed in claim 19 or claim 20 which haε an ability to εyntheεise εtarch with an altered fine structure.

22. A plant as claimed in any of claims 19 to 21 in which starch syntheεiε haε an increaεed or decreaεed temperature optimum.

23. A plant as claimed in any of claims 19 to 22 in which at least one of the donor genes is derived from a plant.

24. A plant aε claimed in claim 23 in which the donor gene is derived from a plant of the species Zea mays, Zea diploperennis, Zea luxurians, Zea perennis, Zea tripεacum, Zea parviglumis, Zea mexicana or teosinte.

25. A plant aε claimed in claim 24 in which the donor gene is derived from a plant of the Zea mays varieties Lima 38, Guanajuato 13, Lima 45, Doebley 479 or teosinte 154.

26. A plant as claimed in any of claims 19 to 25 in which at least one of the donor genes is derived from a bacterium.

27. A plant as claimed in any of claims 19 to 26 in which at least one of the donor genes is derived from a fungus.

28. A plant as claimed in any of claims 19 to 27 in which at least one of the donor genes is derived from an animal cell.

29. A plant as claimed in any of claims 1 to 10 in which the donor gene εpecifieε εoluble εtarch synthaεe (E.C. 2.4.1.21) and/or branching enzyme (E.C. 2.4.1.18) and/or glycogen εynthaεe (E.C. 2.4.1.11 or E.C. 2.4.1.21).

30. A plant as claimed in any of claims 19 to 29 in which at least one of the donor genes specifies a modified allelic form of the enzyme.

31. A plant as claimed in any of claims 19 to 30 in which at least one of the donor genes specifies a mRNA antisenεe to the mRNA encoded by the wild-type gene.

32. A plant aε claimed in any of claimε 19 to 31 which is of the family Gramineae.

33. A plant as claimed in claim 32 which is of the εpecieε Zea mayε.

34. A plant aε claimed in any of claimε 19 to 31 which iε a tomato.

35. A plant aε claimed in any of claimε 19 to 34 having a εtarch εyntheεiεing ability which doeε not decreaεe with temperature between 25 to 30°C.

36. A plant as claimed in any of claims 19 to 35 having a donor gene encoding soluble starch synthase and/or branching and/or glycogen synthaεe enzyme with a Q,_Q value greater than 0.8 between 25 and 35°C. PCI7G

59

37. A plant as claimed in any of claims 19 to 36 having a donor gene encoding soluble starch synthase and/or branching and/or glycogen synthase enzyme which is resiεtant to reduction of activity after expoεure to a temperature in exceεε of 40°C for two hourε.

38. Seeds of a plant as claimed in any of claims 19 to 37.

39. A plant which iε derived from a plant aε claimed in any of claimε 19 to 37.

40. A plant aε claimed in claim 39 which iε a hybrid.