AU4704499A

AU4704499A - Dna encoding oat acetyl coa carboxylase

Info

Publication number: AU4704499A
Application number: AU47044/99A
Authority: AU
Inventors: Margaret A. Egli; R. Gary Fulcher; Susanne Groh; Shahryar F. Kianian; Ronald L. Phillips; David A. Somers
Original assignee: University of Minnesota
Current assignee: University of Minnesota
Priority date: 1998-06-22
Filing date: 1999-06-22
Publication date: 2000-01-10
Also published as: WO1999067367A1; WO1999067367A9; CA2332643A1; EP1092011A1

Description

WO 99/67367 PCT/US99/14022 DNA ENCODING OAT ACETYL CoA CARBOXYLASE Background of the Invention Acetyl CoA carboxylase (ACCase) is an enzyme involved in many important metabolic pathways in plant, animal and bacterial cells. The enzyme is especially important in fatty acid synthesis in plants and is sensitive to 5 inhibition by some types of herbicides. Structurally, ACCases are biotinylated and are quite large enzymes consisting of one or more subunits. For example, most ACCases of animals, higher plants, and yeast are dimers of 420 to 700 kD native MW and contain subunits of 200 to 280 kD. Diatom and algal ACCases are 700 to 740 kD tetramers of 160 to 180 kD subunits. Bacterial ACCase 10 consists of three dissociable proteins; biotin carboxylase (51 kD), biotin carboxyl carrier protein (22.5 kD), and biotin transcarboxylase (130 kD). Acetyl CoA Carboxylase (ACCase) catalyzes the ATP-dependent formation of malonyl-CoA from acetyl-CoA and bicarbonate in animal, plant, and bacterial cells. Malonyl-CoA is an essential substrate for (i) de novo fatty 15 acid (FA) synthesis, (ii) fatty acid elongation, (iii) synthesis of secondary metabolites such as flavonoids and anthocyanins, and (iv) malonylation of some amino acids and secondary metabolites. Synthesis of malonyl-CoA is the first committed step of flavonoid and fatty acid synthesis and current evidence suggests that ACCase catalyzes the rate-limiting step of fatty acid synthesis. 20 Formation of malonyl-CoA by ACCase occurs via two partial reactions and requires a biotin prosthetic group: (i)E-biotin + ATP + HCO 3 --> E-biotin-C0 2 + ADP + Pi (ii)E-biotin-C0 2 + Acetyl-CoA --> E-biotin + malonyl-CoA (NET)Acetyl-CoA + ATP + HCO 3 --> malonyl-CoA + ADP + Pi 25 In E. ali, these reactions are catalyzed by three distinct components; biotin carboxylase, biotin transcarboxylase, and biotin carboxyl carrier protein, which can be separated and yet retain partial activity. Plant and animal ACCases contain all three activities on a single polypeptide. In plants, most ACCase activity is located in plastids of green and 30 non-green plant tissues including leaves and oil seeds. Leaf ACCase activity is primarily located in mesophyll cells, but lesser amounts have been found in C-4 WO 99/67367 PCT/US99/14022 2 bundle sheath cells and in epidermal cells. The subcellular location of ACCase activity in epidermal cells is unknown, but since synthesis of very long-chain fatty acids (VLCFA) for formation of waxes, cutin, and suberin occurs on the endoplasmic reticulum (ER), malonyl-CoA might also be derived from a 5 cytosolic ACCase. In contrast, rat ACCase is primarily cytosolic or associated with the outer mitochondrial membrane. De novo fatty acid synthesis in chloroplasts involves successive 2-carbon additions to acetate, using malonate as the 2-C donor. All intermediates are attached to acyl carrier protein (ACP). Synthesis in plastids 10 resembles that in E. coli in that the fatty acid synthesis complex can be dissociated into separate enzymes: P-ketoacyl-ACP synthase (KAS), P-ketoacyl ACP reductase, p-hydroxyl-ACP dehydratase, and enoyl-ACP reductase, acetyl-CoA:ACP transacylase, and malonyl-CoA:ACP transacylase. A highly active KASIII isozyme catalyzes the condensation of acetyl-CoA and 15 malonyl-ACP. Successive additions of malonyl-CoA to acyl-ACPs catalyzed by KAS I form C16 acyl-ACP, some of which is converted to C18 acyl-ACP by KAS II and then to C18:1-ACP. Fatty acid metabolism then diverges; de esterification allows movement to the cytoplasm (eukaryotic path) where fatty acids may be further unsaturated and/or elongated by additions of malonyl-CoA 20 in the ER. Alternatively, fatty acids are linked to glycerol-3-phosphate (prokaryotic path), further unsaturated, and used for synthesis of chloroplast lipids. A portion of cytoplasmic lipids returns to the chloroplast. The relative contributions of these two paths are species-specific but appear to be relatively flexible in mutants blocked in either path. In oil-storing organs such as 25 cotyledons and monocot embryos, the triacylglycerides are stored in cytoplasmic oil bodies surrounded by a single unit membrane. Condensation of malonyl-CoA with phenylpropionyl-CoAs or acetyl-CoA leads to synthesis of flavonoids, anthocyanins, or to polyacetates. Condensation is increased by light, elicitors, or pathogens and may be the rate 30 limiting step in synthesis of some phytoalexins. In addition to the secondary metabolites derived by de novo synthesis, malonyl conjugates of flavonoid glycosides, formed by malonyl-CoA:flavonoid glycoside malonyltransferase, D-amino acids and 1 -amino-carboxyl-cyclopropane (ethylene precursor) are WO 99/67367 PCTIUS99/14022 3 found in plants. Malonylated compounds accumulate in vacuoles, probably after synthesis in the cytoplasm. An important property of ACCase is the central role it plays in fatty acid synthesis and accumulation in plants and seeds. Available evidence 5 supports the idea that ACCase activity is the rate-limiting step for de novo fatty acid synthesis in plants. High rates of ACCase activity in vitro parallel or slightly precede high rates of lipid deposition or [ 1 4 C]acetate incorporation into lipids in developing leaves and oil seeds. Significant changes in plant ACCase activity occur during chloroplast development and increase in ACCase activity 10 correlates with lipid deposition in developing oil seeds. Turnham et al., Biochem. J., 212:223 (1883); and Beittenmiller et al., Plant PhysioL, 10_0:923 (1992). Among other properties, ACCase in most monocots is also inhibited by several herbicides. [14C]acetate incorporation into maize lipids is 15 strongly inhibited by fluazifop and sethoxydim due to inhibition of plastid ACCase. In barley, however, fluazifop had little effect on [1 4 C]acetate incorporation into very long-chain fatty acids. Since synthesis of very long chain fatty acids occurs in the cytosol on the ER, and de novo fatty acid synthesis occurs in the plastids, cytosolic malonyl-CoA might be supplied by a 20 herbicide insensitive ACCase isozyme. There are three general mechanisms by which plants may be resistant to, or tolerant of, herbicides. These mechanisms include insensitivity at the site of action of the herbicide (usually an enzyme), rapid metabolism (conjugation or degradation) of the herbicide, or poor uptake and translocation of 25 the herbicide. Altering the herbicide site of action from a sensitive to an insensitive form is the preferred method of conferring tolerance on a sensitive plant species. This is because tolerance of this nature is likely to be a dominant trait encoded by a single gene, and is likely to encompass whole families of compounds that share a single site of action, not just individual chemicals. 30 Therefore, detailed information concerning the biochemical site and mechanism of herbicide action is of great importance and can be applied in two ways. First, the information can be used to develop cell selection strategies for the efficient identification and isolation of appropriate herbicide-tolerant variants. Second, it WO 99/67367 PCT/US99/14022 4 can be used to characterize the variant cell lines and regenerated plants that result from the selections. Tissue culture methods have been used to select for resistance (or tolerance) using a variety of herbicides and plant species (see review by 5 Meredith and Carlson, 1982, in Herbicide Resistance in Plants, eds. Lebaron and Gressel, pp. 275-291, John Wiley and Sons, NY). For example, P. C. Anderson et al., in U.S. Patent No. 4,761,373, disclose the use of tissue culture methods to produce maize plants resistant to herbicidal imidazolidones and sulfonamides. The resistance is due to the presence of altered acetohydroxy acid synthase 10 which is resistant to deactivation by these herbicides. Certain 1,3-cyclohexanediones exhibit general and selective herbicidal activity against plants. One such cyclohexanedione is sethoxydim {2 [1-(ethoxyimino)-butyl]-5-[2-(ethylthio)propyl]-3-hydroxy-2-cyclohexen-1 one}. Sethoxydim is commercially available from BASF (Parsippany, New 15 Jersey) under the designation POASTTM. Other herbicidal cyclohexanediones include clethodim, (E,E)-(+) 2-[1-[[(3-chloro-2-propenyl)oxy] imino]propyl]-5-[2-(ethylthio)propyl]-3 hydroxy-2-cyclohexen-1-one; available as SELECTTM from Chevron Chemical (Valent) (Fresno, California); cloproxydim, (E,E)-2-[1-[[(3-chloro-2 20 propenyl)oxy]imino] butyl]-5-[2-(ethylthio)propyl]-3-hydroxy-2-cyclohexen-1 one; available as SELECTONETM from Chevron Chemical (Valent) (Fresno, California); and tralkoxydim, 2-[1-(ethoxyimino)propyl]-3-hydroxy-5 mesitylcyclohex-2-enone, available as GRASPTM from Dow Chemical USA (Midland, Michigan). 25 For purposes of reference in the present specification, the herbicides described in the two preceding paragraphs and other structurally related herbicidal compounds, are collectively referred to as the cyclohexanedione family of herbicides. Certain aryloxyphenoxypropanoic acids exhibit general and 30 selective herbicidal activity against plants. In these compounds, the aryloxy group may be phenoxy, pyridinyloxy or quinoxalinyl. One such herbicidal aryloxyphenoxypropanoic acid is haloxyfop, {2-[4-[[3-chloro-5 (trifluoromethyl)-2-pyridinyl]oxy]phenoxy]- propanoic acid}, which is available WO 99/67367 PCT/US99/14022 5 as VERDICT TM from Dow Chemical USA (Midland, Michigan). Another is diclofop, {(±)-2-[4-(2,4-dichlorophenoxy)-phenoxy]propanoic acid}, available as

HOELON

T M from Hoechst-Roussel Agri-Vet Company (Somerville, New Jersey). 5 Other members of this family of herbicides include fenoxyaprop, (±)-2-[4-[(6-chloro-2-benzoxazolyl)oxy] phenoxy]propanoic acid; available as WHIPTM from Hoechst-Roussel Agri-Vet Company (Somerville, New Jersey); fluazifop, (±)-2-[4-[[5-(trifluoromethyl)-2-pyridinyl]oxy]phenoxy]propanoic acid; available as FUSILADETM from ICI Americas (Wilmington, Delaware); 10 fluazifop-P, (R)-2-[4-[[5-(trifluoromethyl)-2-pyridinyl]oxy]phenoxy]propanoic acid; available as FUSILADE 2000TM from ICI Americas (Wilmington, Delaware); and quizalofop, (±)-2-[4[(6-chloro-2-quinoxalinyl) oxy]phenoxy]propanoic acid; available as ASSURETM from E. I. DuPont de Nemours (Wilmington, Delaware). 15 For purposes of reference in the present specification, the herbicides referred to in the two preceding paragraphs and other structurally related herbicidal compounds, are collectively referred to as herbicidal aryloxyphenoxypropanoic acids. Thus, there is a need for methods to develop plants that are 20 resistant or tolerant to herbicides. There is also a need to increase the oil and/or fatty acid content of the plants and seeds, as well as for methods to increase their resistance to herbicides. Moreover, there is a need to identify and clone genes important in conferring herbicide tolerance and in increasing the oil content of plants. 25 Summary of the Invention The present invention provides an isolated and purified DNA molecule comprising a DNA segment encoding a plant acetyl CoA carboxylase such as a monocot acetyl CoA carboxylase, e.g., a maize or oat acetyl CoA 30 carboxylase, and methods for conferring herbicide tolerance and/or altering the oil content of plants by introducing and expressing a plant acetyl CoA carboxylase gene in the plant cells. The DNA molecule encoding a plant acetyl CoA carboxylase can encode an unaltered plant acetyl CoA carboxylase or an WO 99/67367 PCT/US99/14022 6 altered plant acetyl CoA carboxylase substantially tolerant to inhibition by cyclohexanedione or aryloxyphenoxypropanoic acid herbicides as well as encoding an antisense DNA sequence that is substantially complementary to a plant acetyl CoA carboxylase gene or to a portion thereof. A DNA molecule of 5 the invention can also further comprise an amino terminal plant chloroplast transit peptide sequence operably linked to the plant acetyl CoA carboxylase gene. The DNA molecule of the invention preferably encodes the plastidic form of plant acetyl CoA carboxylase. The invention also provides for probes, e.g., a probe which is detectably labeled or which binds to a detectable label, or 10 primers, e.g., a single-strand DNA molecule having at least seven nucleotide bases, that hybridize to a DNA encoding a plant acetyl CoA carboxylase, e.g., a DNA molecule comprising SEQ ID NO:23 or its complement. For example, the present invention provides an isolated and purified DNA molecule of at least seven nucleotide bases which hybridizes under high stringency conditions to a 15 DNA molecule comprising a DNA segment encoding a plant acetyl CoA carboxylase and provides a hybridization probe comprising an isolated and purified DNA molecule of at least seven nucleotide bases, which is detectably labeled or which binds to a detectable label, which DNA molecule hybridizes under high stringency conditions to the non-coding strand of a DNA molecule 20 comprising a DNA segment encoding a plant acetyl CoA carboxylase. High stringency conditions are defined as: hybridization at 65 C for at least 16 hours in 5xSSC, lx Denhardt's solution, 50 mM Tris-HCl, pH 8, 0.2% SDS, 10 mM EDTA, 0.1 mg/ml salmon sperm DNA, followed by washing twice for 5 minutes in 2xSSC, 0.5% SDS at 25'C, once for 10 minutes in 0.2xSSC, 0.1% SDS at 25 25'C and twice for 30 minutes in 0.2xSSC, 0.1% SDS at 65'C. The method of imparting cyclohexanedione or aryloxyphenoxypropanoic acid herbicide tolerance to a plant includes the steps of introducing a chimeric or recombinant DNA molecule comprising a gene coding for a plant acetyl CoA carboxylase or an altered or a functional mutant 30 thereof operably linked to a promoter functional in a plant cell into cells of a susceptible plant, and regenerating the transformed plant cells to provide a differentiated plant. The promoter can be an inducible or tissue specific promoter or provide for overexpression of at least about a 2-fold amount of a WO 99/67367 PCT/US99/14022 7 native plant acetyl CoA carboxylase. The functional linkage of a promoter to the chimeric DNA molecule results in an expression cassette. Expression of the chimeric DNA molecule is in an amount effective to render the acetyl CoA carboxylase and/or the plant tissue substantially tolerant to the herbicides relative 5 to the native acetyl CoA carboxylase present in said plant. Herbicide tolerance can be achieved in the plants by at least two methods, including increasing the level of gene expression of a native or unaltered acetyl CoA carboxylase, or by introducing an altered gene coding for an acetyl CoA carboxylase that is less sensitive to herbicide inhibition. The level of gene expression can be increased 10 by either combining a plant acetyl CoA carboxylase gene with a promoter that provides for a high level of gene expression, such as a 35S cauliflower mosaic virus promoter (CaMV), or by introducing multiple copies of the gene into the cell so that the multiple copies of the gene are integrated into the genome of transformed plant cells. The preferred plant cells into which to introduce the 15 expression cassette of the invention, to achieve herbicide tolerance, are monocot plant cells. Preferably, the cells are not oat cells. Once transformed cells exhibiting herbicide tolerance are obtained, transgenic plants and seeds can then be regenerated therefrom, and evaluated for stability of the inheritance of the herbicide tolerance trait. 20 The invention also provides a method for altering, e.g., lowering or raising, the oil content in a plant. As described hereinbelow, a locus having a major effect on oat (Avena sativa L.) groat oil content was located on linkage group 11 by quantitative trait locus (QTL) dissection methods. A partial oat acetyl-CoA carboxylase (ACCase) cDNA with >84% identity to plastidic 25 ACCase sequences from wheat or maize was linked to this locus. Similar QTL and ACCase placements were obtained with two recombinant inbred populations (n ; 137) with one common parent. The locus on linkage group 11 linked to the ACCase locus accounted for up to 48% of the phenotypic variance for groat oil content. Other loci were identified which accounted for an additional 14% of the 30 phenotypic variance. The major locus plus additional QTLs represent three potentially homoeologous genomic segments in oat that can be followed by as few as three RFLP markers to produce a large shift in groat oil content. Oat acetyl CoA carboxylase DNA is also useful in molecular marker-assisted WO 99/67367 PCTIUS99/14022 8 selection to facilitate breeding high- or low-oil oat, genetic dissection of the groat oil trait, and to identify specific genes with large effects whose activity may be manipulated via genetic transformation. Thus, the invention provides a method which includes the steps of 5 introducing a chimeric or recombinant DNA molecule comprising a gene coding for a plant, e.g., oat, acetyl CoA carboxylase or an altered or a functional mutant thereof operably linked to a promoter functional in a plant cell into the cells of plant tissue and expressing the gene in an amount effective to alter the oil content of the plant cell. An alteration in oil content can include a change in 10 total oil content over that normally present in that type of plant cell, a change in the location of the oil in the plant, plant tissue, seed or plant part, and/or a change in the type of oil present in the cell. For example, low-oil oat includes, but is not limited to, an oat plant, plant part, plant tissue or groat having about 8 to about 40, preferably about 8 to about 30, and more preferably about 20 to 15 about 30, percent dry weight of oil relative to the corresponding parental, e.g., nontransformed or inbred, oat plant, plant part, plant tissue or groat. High-oil oat includes, but is not limited to, an oat plant, plant part, plant tissue or groat having about 2 to about 8, preferably 1 to about 4, and more preferably about 2 to about 4, percent dry weight of oil relative to the corresponding parental oat plant, plant 20 part, plant tissue or groat. An alteration in oil content in the plant cell can be achieved by at least two methods including: (1) an increase or decrease in expression of an unaltered or altered plant acetyl CoA carboxylase gene; or 25 (2) by introducing an unaltered, altered or functional mutant plant acetyl CoA carboxylase gene. The level of gene expression of an unaltered plant acetyl CoA carboxylase gene can be increased by either combining an unaltered plant acetyl CoA carboxylase with a promoter that provides for a high level of gene expression, or by 30 introducing multiple copies of an expression cassette into cells so that multiple copies of the gene are integrated into the genome. When an altered or a functional mutant plant acetyl CoA carboxylase gene codes for an enzyme that exhibits an increase in specific activity, it can lead to an increase in total oil WO 99/67367 PCT/US99/14022 9 content of the plant cell. When an altered or a functional mutant acetyl CoA carboxylase gene codes for an enzyme having a decrease in specific activity, it may lead to a decrease in the total oil content of the plant cell. Preferably, the expression cassette is introduced into dicot plants such as soybeans, canola, and 5 sunflower. In an especially preferred version, transformed cells exhibiting about a 1.2- to 5-fold increase in total oil content and/or expression or specific activity of acetyl CoA carboxylase are selected for and used to generate transgenic plants and seeds exhibiting a substantial increase in oil content. A substantial increase in oil content depends on the oil content normally present in the plant or seed 10 and can range from about a 1.2 to a 20-fold increase. The invention also provides for a method of producing plant acetyl CoA carboxylase in a host cell. The method includes the steps of introducing an expression cassette comprising a chimeric gene encoding a plant acetyl CoA carboxylase or an altered or a functional plant acetyl CoA 15 carboxylase operably linked to a promoter into a host cell and expressing the gene in an amount sufficient to permit crystallization of the plant acetyl CoA carboxylase. An expression cassette can include a promoter that is functional in either a eukaryotic or a prokaryotic cell. Preferably, the expression cassette is introduced into a prokaryotic cell, such as E. cli, that is routinely used for 20 production of recombinantly produced proteins. Recombinantly produced and crystallized plant acetyl CoA carboxylase can then be used to identify other herbicides and that bind to and inhibit acetyl CoA carboxylase in plants. In addition, the availability of large amounts of purified enzyme can permit the screening of the efficacy of such herbicides in terms of their ability to bind to, or 25 otherwise inhibit, the activity of the enzyme. The present invention also provides a method of introducing an exogenous plant acetyl CoA carboxylase gene into a host cell comprising transforming host cells in vitro with an expression cassette comprising a chimeric or recombinant DNA molecule encoding a plant acetyl CoA 30 carboxylase gene operably linked to a promoter functional in the host cell, expanding the transformed host cells in vitro, and identifying a transformed host cell which expresses the chimeric DNA molecule.

WO 99/67367 PCT/US99/14022 10 Further provided is a method to identify a plant, e.g., an oat plant, having altered characteristics, e.g., altered groat oil content. The method comprises contacting a probe comprising at least a portion of a nucleic acid sequence encoding oat acetyl CoA carboxylase with a sample from an oat plant 5 in an amount sufficient to form complexes, wherein the sample comprises plant genomic nucleic acid. The amount of complex formation is then detected or determined. Preferably, the plant cells in the sample are monocot plant cells, e.g., wheat, rice, maize, barley, millet (e.g., foxtail or finger), sorghum, sugarcane, rye, oat or timothy cells. Also preferably, the DNA molecule encodes 10 a polypeptide comprising SEQ ID NO:24, e.g., a DNA molecule comprising SEQ ID NO:23. Thus, the invention provides a nucleic acid fragment useful in marker-assisted breeding programs such as those which identify plants having altered groat oil content or composition. Also provided is an expression cassette comprising an 15 endosperm-specific transcriptional control region, e.g., the DylO or zein transcriptional control region, operably linked to a DNA encoding a plant acetyl CoA carboxylase, or subunit thereof. Other endosperm-specific promoters are also useful in the compositions and methods of the invention. These promoters include, but are not limited to, ltrl promoter, a synthetic promoter with 20 endosperm box of HOD (Vincentecarbajosa et al., Plant J., 13, 629 (1998)); rice glutelin gene promoter (Sindhu et al., Plant Sci., 13_0, 189 (1997); Yoshihara et al., Plant Sci., 121, 63 (1996)); barley alpha-amylase inhibitor subunit promoter (Gross et al., Plant Mol. Biol., 34, 331 (1997)); gamma zein promoter (Towent et al., Plant Mol. Biol., 34, 139 (1997)); maize opaque-2 locus promoter (Rossi et 25 al., J. Biol. Chem., 272, 13758 (1997); Gallusci et al., Mol & Gen. Genet., 244, 391 (1994)); the zmGBS, zmZ27, osAGP or osGT1 promoter (Russell et al., Transgenic Res., f6, 157 (1997)); alpha globulin promoter (Nakase et al., Plant Mol. Biol., 33., 513 (1997)); hordein promoter (Sorenson et al., Mol. & Gen. Genet., 25_, 750 (1996)); Viciafaba legumin and vicilin storage protein gene 30 promoter (Wobus et al., I Plant Physiol., 145, 592 (1995)); rice waxy gene promoter (Hirano et al., Plant & Cell Physiol., 3-6, 37 (1995)); oat globulin gene osglo5 promoter (Schubert et al., Plant Mol. Biol., 26, 203 (1994)); napA WO 99/67367 PCT/US99/14022 11 promoter (Stalberg et al., Plant Mol. Biol, 23, 671 (1993)); and rice prolamin gene promtoer (Zhou et al., Science in China, 3_6, 1307 (1993)). The term "consists essentially of' as used with respect to the present DNA molecules is defined to mean that a major portion of the nucleotide 5 sequence encodes an ACCase, and that nucleotide sequences are not present which encode proteins other than ACCase or functional equivalents thereof. Brief Description of the Figures Figure 1: A schematic depiction of the fatty acid biosynthesis 10 pathway in plants. Figure 2: A graph depicting the effect of sethoxydim on the growth of mutant maize callus. Figure 3: A graph depicting the shoot length growth of maize seedlings seven days after treatment with sethoxydim. 15 Figure 4: A graph depicting the shoot length growth of maize seedlings fourteen days after treatment with sethoxydim. Figure 5: Total soluble and biotinylated polypeptides in ACCase purification fractions from seedling leaves of maize inbred A619. Proteins were separated by SDS-PAGE in 7.5% gels and then silver-stained (Panel A). An 20 identical gel was Western-blotted and a longitudinal section of each lane was probed with avidin (Panel B). Lanes were 1: crude (10 gg); 2: (NH 4

)

2 SO4 (10 pg); 3: S-300 (5 jg); 4: Blue Sepharose (2 pg); 5: Mono-Q ACCase II (5 gg); and 6: Mono-Q ACCase I (5 pg). Diagonal lines between lanes indicate position of molecular weight markers shown on the left. 25 Figure 6: Immunoprecipitation of ACCase activity from B73 leaf, embryo, endosperm, and BMS suspension cultured cells. Equal activities (0.58 nmol min-') were incubated with 16 gL serum (immune plus preimmune), immune complexes were precipitated with Protein A-agarose, and ACCase activity remaining in the resulting supernatant fraction was measured relative to 30 the preimmune serum control. Figure 7: Herbicide inhibition of acetyl-(AcCoA) or propionyl-CoA (Prop-CoA)-dependent H 1 4

CO

3 -incorporation into acid-stable WO 99/67367 PCT/US99/14022 12 product by ACCase I and II Mono-Q fractions. Activities in the presence of haloxyfop (1 pM) are expressed relative to the minus herbicide control. Figure 8: Comparison of the peptide sequence of maize cDNA clones #15-14 and #18-5 with chicken ACCase. The approximate locations of 5 the biotin carboxylase, biotin binding site, and biotin transcarboxylase functional domains are indicated for the chicken sequence. The percentages of amino acid identity are indicated by cross-hatched boxes for the maize coding sequence. Regions of genomic DNA Type I and Type II clone sequences that align with cDNA #18-5 are indicated by solid heavy lines. The approximate locations of 10 subclone #28 and #16 from genomic Type I and subclone #34 from genomic Type II clones are indicated. Figure 9: Northern blot of total RNA from leaf, immature embryo and endosperm tissue (16 days after pollination), and Black Mexican Sweet corn (BMS) cells. Lanes contain 10 jg total RNA and were probed with 15 the 2 kb EcoRI fragment of lambda clone #15-14. Figure 10: DNA sequence (SEQ ID NO. 1) of a 2 kb EcoRI fragment of lambda clone #15-14 including a portion of a maize ACCase gene located at bases 2883 to 83 from the 3' stop codon. Figure 11: Graph of ACCase activity during seed development in 20 two high oil soybean cell lines and one low oil soybean cell line. Figure 12: Cloning strategy to obtain the complete coding sequence of the maize ACCase gene. Figure 13: DNA sequence (SEQ ID NO:5) of a 7470 base pair cDNA of a maize ACCase gene. (Genbank Accession No. U19183). 25 Figure 14: Predicted amino acid sequence of the complete ACCase gene of maize (SEQ ID NO:6). Figure 15: Restriction map of ACCase genomic clones. Figure 16: Partial nucleotide sequence of a Type A, ACCase genomic clone (SEQ ID NO:12). 30 Figure 17: Partial nucleotide sequence of clone 5A, a Type A ACCase genomic clone (SEQ ID NO:13). Figure 18: Partial nucleotide sequence of six Type A 2 ACCase genomic clones (A-F) (SEQ ID NOs 14, 15, 16, 17, 18, and 19).

WO 99/67367 PCTIUS99/14022 13 Figure 19: Partial nucleotide sequence of three Type B ACCase clones (SEQ ID NOs 20, 21 and 22). Figure 20: SDS/PAGE analysis of chloroplast importation of "S labeled ACCase polypeptides. (A) Pea and maize chloroplasts incubated with 5 "S-labeled ACCase polypeptides for 30 minutes. (B) A time course analysis of the importation of "S-labeled ACCase polypeptides into maize chloroplasts. Figure 21: Partial DNA sequence encoding oat ACCase (SEQ ID NO:23, Genbank Accession No. AF072737). Figure 22: Partial amino acid sequence of an oat ACCase (SEQ 10 ID NO:24). Figure 23: Scan of test statistics generated by MQTL for the two recombinant inbred populations. A and B are the main effect and QTL x E interaction graphs for the Kanota X Ogle (KO) population, respectively. C and D are the main effect and QTL x E interaction graphs for the Kanota X Marion 15 (KM) population, respectively. Thresholds for type I error rate of 5% after 1000 replications are reported above the graphs. Figure 24: Histogram of % groat oil concentration for the mean of all environments relative to the number of recombinant inbred lines (RILs) with each parental genotype for the XaccaseA locus. A) The distribution for the 20 Kanota X Ogle population, and B) distribution for the Kanota X Marion population. Figure 25: Phylogram of ACCases. Detailed Description of the Invention 25 The present invention provides a DNA molecule encoding a plant acetyl CoA carboxylase gene and methods for conferring herbicide tolerance and/or altering the oil content of plants by introducing and expressing a plant acetyl CoA carboxylase gene in the plant cells. In plants, acetyl CoA carboxylase plays a central role in regulating fatty acid synthesis and in the 30 sensitivity of monocots to cyclohexanedione or aryloxyphenoxypropanoic acid herbicides. In accord with the present invention, a plant acetyl CoA carboxylase gene is identified, isolated and combined with a promoter functional WO 99/67367 PCTIUS99/14022 14 in a plant cell to provide a recombinant expression cassette. A plant acetyl CoA carboxylase gene can be introduced and expressed in a plant cell. Depending on the type of plant cell, the level of gene expression, and the activity of the enzyme encoded by the gene, introduction of a plant acetyl CoA carboxylase gene into 5 the plant cell can confer herbicide tolerance and/or alteration of the oil of the plant cell. In monocots, an exogenously introduced plant acetyl CoA carboxylase gene can be expressed at a level effective to render the cells of the plant tissue substantially tolerant to cyclohexanedione or 10 aryloxyphenoxypropanoic acid herbicide levels which normally inhibit a native or endogenous acetyl CoA carboxylase. A native acetyl CoA carboxylase is an enzyme that is normally encoded and expressed in the plant cell prior to transformation. An exogenously introduced plant acetyl CoA carboxylase gene is a gene which has been isolated and amplified from either the same or different 15 type of cell. Exogenous introduction and expression of a plant acetyl CoA carboxylase gene in both monocots and dicots can result in alteration of the oil content and quality of plant tissue and seeds. Exogenous introduction and expression in a host cell, such as a bacteria, can provide for sufficient amounts of plant acetyl CoA carboxylase to allow for crystallization and isolation of the 20 enzyme. Crystallized plant acetyl CoA carboxylase is useful to identify other herbicides that bind to and can inhibit plant acetyl CoA carboxylases. The enzyme could also be used to screen potential herbicidal compounds for efficacy. A. Formation of an Expression Cassette An expression cassette of the invention can comprise a chimeric 25 DNA molecule encoding a plant acetyl CoA carboxylase gene or an altered or functional mutant thereof operably linked to a promoter functional in a plant cell. The gene can code for a plant acetyl CoA carboxylase that is substantially tolerant to herbicides, preferably cyclohexanedione and/or aryloxyphenoxypropanoic acid herbicides. An expression cassette of the 30 invention can also include an antisense DNA sequence that is substantially complementary to an acetyl CoA carboxylase gene or a portion thereof operably linked to a promoter functional in a plant cell.

WO 99/67367 PCT/US99/14022 15 1. Isolation and Identification of a Gene Coding for a Plant Acetyl CoA Carboxylase A gene encoding a plant acetyl CoA carboxylase can be identified and isolated by standard methods, as described by Sambrook et al., Guide to 5 Molecular Cloning: A Laboratory Manual, Cold Spring Harbor, NY (1989). The gene can be obtained either from monocot or dicot plant cells. When the gene encoding a plant acetyl CoA carboxylase is obtained from a dicot plant, the enzyme encoded by the gene exhibits tolerance to cyclohexanedione or aryloxyphenoxypropanoic acid herbicides. The gene can also be obtained from 10 herbicide-tolerant maize cell lines, prepared as described in U.S. Patent No. 5,162,602. A gene encoding a plant acetyl CoA carboxylase can be identified by screening of a DNA or cDNA library generated from plant cells. Screening for DNA fragments that encode all or a portion of the gene encoding a plant 15 acetyl CoA carboxylase can be accomplished by complementation of an auxotrophic mutant of acetyl CoA carboxylase in E. coli (fabE) (Bachman, Microbiological Reviews, 47:180 (1983)) or yeast (acc) (Michionada, Eur. .I Biochem., 111:79 (1980)) or by screening of plaques for binding to antibodies that specifically recognize a plant acetyl CoA carboxylase. DNA fragments that 20 can restore ACCase activity in E. cli or yeast and/or plaques carrying DNA fragments that are immunoreactive with antibodies to a plant ACCase can be subcloned into a vector and sequenced and/or used as probes to identify other cDNA or genomic sequences encoding all or a portion of a plant acetyl CoA carboxylase gene. 25 Specific examples of cDNA sequences encoding a portion of a plant acetyl CoA carboxylase gene include DNA fragments that include a DNA sequence that substantially corresponds to the coding sequence for the transcarboxylase active site of a plant acetyl CoA carboxylase, DNA fragments that include a DNA sequence that substantially corresponds to a coding sequence 30 for the biotin binding site of a plant acetyl CoA carboxylase, a DNA fragment encoding the 5' transcriptional start sequence of a plant acetyl CoA carboxylase gene, and a DNA fragment encoding the 3' transcriptional stop sequence for the acetyl CoA carboxylase gene. Substantially corresponding DNA sequences WO 99/67367 PCT/US99/14022 16 share about 90% to about 100% DNA sequence homology. Especially preferred cDNA probes can be obtained from lambda clone #18-5 which include DNA sequences corresponding to the transcarboxylase active site domain and the biotin binding site domain. Lambda clone #18-5 includes EcoRI subclones of 5 3.9 kb, 1.2 kb, or 0.23 kb. Lambda subclone #18-5I is an 3.9 kb EcoRI subclone. The lambda subclone #18-51 has been deposited with the American Type Culture Collection, Rockville, MD, and given Accession No. 69236. In a preferred version, a plant acetyl CoA carboxylase gene is identified and isolated from an herbicide tolerant maize cell line prepared as 10 described in Example II. A cDNA library can be prepared by oligo dT priming. Plaques containing DNA fragments can be screened with antibodies specific for maize acetyl CoA carboxylase. DNA fragments encoding a portion of an acetyl CoA carboxylase gene can be subcloned and sequenced and used as probes to identify a genomic acetyl CoA carboxylase gene. DNA fragments encoding a 15 portion of a maize acetyl CoA carboxylase can be verified by determining sequence homology with other known acetyl CoA carboxylases, such as chicken or yeast acetyl CoA carboxylase, or by hybridization to acetyl CoA carboxylase specific messenger RNA. Once DNA fragments encoding portions of the 5', middle and 3' ends as well as the transcarboxylase active site or biotin binding 20 site of a plant acetyl CoA carboxylase are obtained, they can be used to identify and clone a complete genomic copy of a maize acetyl CoA carboxylase gene. To isolate a complete copy of a maize acetyl CoA carboxylase gene, a maize genomic library can then be probed with cDNA probes prepared as described above. Portions of the genomic copy or copies of a plant acetyl CoA 25 carboxylase gene can be sequenced and the 5' end of the gene are identified by standard methods including either DNA sequence homology to other acetyl CoA carboxylase genes or by RNAase protection analysis, as described by Sambrook et al., Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Press, Cold Spring Harbor, New York (1989). Once portions of the 5' end of the gene are 30 identified, complete copies of a plant acetyl CoA carboxylase gene can be obtained by standard methods, including by cloning or by polymerase chain reaction (PCR) synthesis using oligonucleotide primers complementary to the DNA sequence at the 5' end of the gene. The presence of an isolated full-length WO 99/67367 PCT/US99/14022 17 copy of a plant acetyl CoA carboxylase gene can be verified by hybridization, partial sequence analysis, or by expression of a plant acetyl CoA carboxylase enzyme. The maize acetyl CoA carboxylase gene cloned and expressed from a maize herbicide tolerant cell line can be assessed for tolerance to 5 cyclohexanedione or aryloxyphenoxypropanoic acid herbicides by standard methods, as described in Example I. An expression cassette of the invention can also contain an antisense DNA sequence. A antisense DNA sequence is a sequence that is substantially complementary to all or a portion of a coding sequence of a plant 10 acetyl CoA carboxylase gene. A substantially complementary sequence has about 90% to about 100% DNA sequence homology with that of the coding sequence of all or a portion of a plant acetyl CoA carboxylase. The antisense DNA sequence when expressed can act to inhibit the synthesis and expression of a native plant acetyl CoA carboxylase. Antisense sequences are preferably about 15 200 to 1000 nucleotides long in order to provide sufficient inhibition of synthesis and/or expression of a native acetyl CoA carboxylase. The inhibition of acetyl CoA carboxylase synthesis and gene expression by antisense DNA sequences can be confirmed in a transformed plant cell by standard methods for measuring the presence and/or activity of the enzyme such as described in Examples I and 20 V. An expression cassette of the invention can also include a functional mutant of a plant acetyl CoA carboxylase gene. Mutants of a plant acetyl CoA carboxylase gene are substantially homologous to a plant acetyl CoA carboxylase gene and are functional if the acetyl CoA carboxylase expressed 25 retains significant enzyme activity. A mutant substantially homologous to a plant acetyl CoA carboxylase can share about 90% to 99.99% DNA sequence with that gene. For example, a mutant acetyl CoA carboxylase gene can code for a herbicide tolerant acetyl CoA carboxylase, or for an acetyl CoA carboxylase with altered substrate specificity so that the total amount of oil content in the 30 plants or seeds is increased, or for an enzyme with an altered substrate specificity so that synthesis of secondary metabolites such as flavonoids or anthocyanins is decreased. A preferred mutant is a gene coding for an acetyl CoA carboxylase WO 99/67367 PCT/US99/14022 18 that is substantially tolerant to cyclohexanedione or aryloxyphenoxypropanoic acid herbicide Altered or functional mutants of a gene coding for a plant acetyl CoA carboxylase can be obtained by several methods. The alteration or 5 mutation of the ACCase gene can be accomplished by a variety of means including, but not limited to, the following methods. 1. spontaneous variation and direct mutant selection in cultures; 2. direct or indirect mutagenesis procedures on tissue culture of all cell types, seeds or plants; and 10 3. mutation of the cloned acetyl CoA carboxylase gene by methods such as site specific mutagenesis (Sambrook et al. cited supra), transposon mediated mutagenesis (Berg et al., Biotechnology, 1:417 (1983)) and deletion mutagenesis (Mitra et al., Molec. Gen. Genetic., 215:294 (1989)). 15 Mutants can be identified by a change in a functional activity of the enzyme encoded by the gene or by detecting a change in the DNA sequence using restriction enzyme mapping or partial sequence analysis. In a preferred version, a functional mutant gene encoding for a plant acetyl CoA carboxylase tolerant to cyclohexanedione and/or 20 aryloxyphenoxypropanoic acid herbicides is isolated from a maize herbicide tolerant cell line. The maize herbicide tolerant cell line was obtained as described in U.S. Patent No. 5,162,602, issued November 10, 1992, the disclosure of which is incorporated in Examples I-III. Briefly, partially differentiated cell cultures are grown and subcultured with continuous exposures 25 to low herbicide levels. Herbicide concentrations are then gradually increased over several subculture intervals. Maize cells or tissues growing in the presence of normally toxic herbicide levels are repeatedly subcultured in the presence of the herbicide and characterized. Stability of the herbicide tolerance trait of the cultured cells may be evaluated by growing the selected cell lines in the absence 30 of herbicides for various periods of time and then analyzing growth after exposing the tissue to herbicide. Maize cell lines which are tolerant by virtue of having an altered acetyl CoA carboxylase enzyme can be selected by identifying cell lines having WO 99/67367 PCT/US99/14022 19 enzyme activity in the presence of normally toxic levels of sethoxydim or haloxyfop. The tolerant maize cells can be further evaluated for whether acetyl CoA carboxylase is altered to a less sensitive form or increased in its level of expression. 5 Maize cell lines with a acetyl CoA carboxylase less sensitive to herbicide inhibition can be used to isolate a functional mutant gene of a plant acetyl CoA carboxylase. A DNA library from a maize cell line tolerant to herbicides can be generated and DNA fragments encoding all or a portion of an acetyl CoA carboxylase gene can be identified by hybridization to a cDNA probe 10 encoding a portion of the maize ACCase gene. A complete copy of the altered gene can be obtained either by cloning and ligation or by PCR synthesis using appropriate primers. The isolation of the altered gene coding for acetyl CoA carboxylase can be confirmed in transformed plant cells by determining whether the acetyl CoA carboxylase being expressed retains enzyme activity when 15 exposed to normally toxic levels of herbicides. 2. Promoters Once a plant acetyl CoA carboxylase gene or functional mutant thereof or an antisense DNA sequence is obtained and amplified, it is combined with a promoter functional in a plant cell to form an expression cassette. 20 Most genes have regions of DNA sequence that are known as promoters and which regulate gene expression. Promoter regions are typically found in the flanking DNA sequence upstream from the coding sequence in both procaryotic and eukaryotic cells. A promoter sequence provides for regulation of transcription of the downstream gene sequence and typically includes from 25 about 50 to about 2,000 nucleotide base pairs. Promoter sequences also contain regulatory sequences such as enhancer sequences that can influence the level of gene expression. Some isolated promoter sequences can provide for gene expression of heterologous genes, that is a gene different from the native or homologous gene. Promoter sequences are also known to be strong or weak or 30 inducible. A strong promoter provides for a high level of gene expression, whereas a weak promoter provides for a very low level of gene expression. An inducible promoter is a promoter that provides for turning on and off of gene expression in response to an exogenously added agent or to an environmental or WO 99/67367 PCT/US99/14022 20 developmental stimulus. Promoters can also provide for tissue specific or developmental regulation. An isolated promoter sequence that is a strong promoter for heterologous genes is advantageous because it provides for a sufficient level of gene expression to allow for easy detection and selection of 5 transformed cells and provides for a high level of gene expression when desired. The promoter in an expression cassette of the invention can provide for overexpression of acetyl CoA of a plant acetyl CoA carboxylase gene or functional mutant thereof. Overexpression of the gene is that amount of gene expression that results in an increase in tolerance of the plant cells to an 10 herbicide or that results in an increase in the total oil content of the cells. Overexpression of an acetyl CoA carboxylase gene is preferably about a 2- to 20-fold increase in expression of an acetyl CoA carboxylase over the expression level of the native acetyl CoA carboxylase. The promoter can also be inducible so that gene expression can be turned on or off by an exogenously added agent. 15 For example, a bacterial promoter such as the Pt promoter can be induced to varying levels of gene expression depending on the level of isothiopropylgalactoside added to the transformed bacterial cells. It may also be preferable to combine the gene with a promoter that provides tissue specific expression or developmentally regulated gene expression in plants. 20 Specific promoters functional in plant cells include the 35S cauliflower mosaic virus promoter, nopaline synthase (NOS) promoter and the like. Currently, a preferred promoter for expression in monocots is the 35S cauliflower mosaic virus promoter. An acetyl CoA carboxylase gene can be combined with the 25 promoter by standard methods as described in Sambrook cited supra. Briefly, a plasmid containing a promoter such as the 35S cauliflower mosaic virus promoter can be constructed as described in Jefferson, Plant Molecular Biology Reporter. 5,387 (1987) or obtained from Clontech Lab in Palo Alto, CA (e.g. pBIl21 or pBI22 1). Typically these plasmids are constructed to provide for 30 multiple cloning sites having specificity for different restriction enzymes downstream from the promoter. A gene for plant acetyl CoA carboxylase can be subcloned downstream from the promoter using restriction enzymes to ensure that the gene is inserted in proper orientation with respect to the promoter so that WO 99/67367 PCT/US99/14022 21 the gene can be expressed. In a preferred version, a maize acetyl CoA carboxylase is operably linked to a 35 S Came promoter in a plasmid such as pBIl21 or pBI221. Once a plant acetyl CoA carboxylase gene is operably linked to a promoter and the plasmid, the expression cassette so formed can be 5 subcloned into other plasmids or vectors. 3. Optional Sequences in the Expression Cassette The expression cassette can also optionally contain other DNA sequences. The expression cassette can further be comprised of a chloroplast transit peptide sequence operably linked between a promoter and a plant acetyl 10 CoA carboxylase gene. If the expression cassette is to be introduced into a plant cell, the expression cassette can also contain plant transcriptional termination and polyadenylation signals and translational signals linked to the 3' terminus of a plant acetyl CoA carboxylase gene. The expression cassette can also optionally be further comprised of a plasmid. 15 Because one site of action for biosynthetic pathways involving plant acetyl CoA carboxylase is the chloroplast, an expression cassette of the invention can be combined with an exogenous DNA sequence coding for a chloroplast transit peptide, if necessary. An exogenous chloroplast transit peptide is one which is not encoded within the plant acetyl CoA carboxylase 20 gene. A chloroplast transit peptide is typically 40 to 70 amino acids in length and functions post-translationally to direct the protein to the chloroplast. The transit peptide is cleaved either during or just after import into the chloroplast to yield the mature protein. The complete copy of a gene encoding a plant acetyl CoA carboxylase may contain a chloroplast transit peptide sequence. In that 25 case, it may not be necessary to combine an exogenously obtained chloroplast transit peptide sequence into the expression cassette. Exogenous chloroplast transit peptide encoding sequences can be obtained from a variety of plant nuclear genes, so long as the products of the genes are expressed as preproteins comprising an amino terminal transit peptide 30 and transported into chloroplast. Examples of plant gene products known to include such transit peptide sequences are the small subunit of ribulose biphosphate carboxylase, ferredoxin, chlorophyll a/b binding protein, chloroplast ribosomal proteins encoded by nuclear genes, certain heatshock proteins, amino WO 99/67367 PCT/US99/14022 22 acid biosynthetic enzymes such as acetohydroxy acid synthase, 3 enolpyruvylphosphoshikimate synthase, dihydrodipicolinate synthase, and the like. Alternatively, the DNA fragment coding for the transit peptide may be chemically synthesized either wholly or in part from the known sequences of 5 transit peptides such as those listed above. Regardless of the source of the DNA fragment coding for the transit peptide, it should include a translation initiation codon and an amino acid sequence that is recognized by and will function properly in chloroplasts of the host plant. Attention should also be given to the amino acid sequence at the 10 junction between the transit peptide and the plant acetyl CoA carboxylase enzyme where it is cleaved to yield the mature enzyme. Certain conserved amino acid sequences have been identified and may serve as a guideline. Precise fusion of the transit peptide coding sequence with the acetyl CoA carboxylase coding sequence may require manipulation of one or both DNA sequences to 15 introduce, for example, a convenient restriction site. This may be accomplished by methods including site directed mutagenesis, insertion of chemically synthesized oligonucleotide linkers and the like. Once obtained, the chloroplast transit peptide sequence can be appropriately linked to the promoter and a plant acetyl CoA carboxylase gene in 20 an expression cassette using standard methods. Briefly, a plasmid containing a promoter functional in plant cells and having multiple cloning sites downstream can be constructed as described in Jefferson cited supra. The chloroplast transit peptide sequence can be inserted downstream from the promoter using restriction enzymes. A plant acetyl CoA carboxylase gene can then be inserted 25 immediately downstream from and in frame with the 3' terminus of the chloroplast transit peptide sequence so that the chloroplast transit peptide is linked to the amino terminus of the plant acetyl CoA carboxylase. Once formed, the expression cassette can be subcloned into other plasmids or vectors. When the expression cassette is to be introduced into a plant cell, 30 the expression cassette can also optionally include 3' nontranslated plant regulatory DNA sequences. The 3' nontranslated regulatory DNA sequence preferably includes from about 300 to 1,000 nucleotide base pairs and contains plant transcriptional and translational termination sequence. Specific examples WO 99/67367 PCT/US99/14022 23 of 3' nontranslated regulatory DNA sequences functional in plant cells include about 500 base pairs of the 3' flanking DNA sequence of the pea ribulose biphosphate carboxylase small subunit E9 gene, the 3' flanking DNA sequence of the octopine synthase gene, and the 3' flanking DNA sequence of the nopaline 5 synthase gene. These 3' nontranslated regulatory sequences can be obtained as described in An, Methods in Enzymology, 151:292 (1987) or are already present in plasmids available from commercial sources such as Clontech, Palo Alto, CA. The 3' nontranslated regulatory sequences can be operably linked to the 3' terminus of a plant acetyl CoA carboxylase gene by standard methods. 10 An expression cassette of the invention can also be further comprised of a plasmid. Plasmid vectors included additional DNA sequences that provide for easy selection, amplification and transformation of the expression cassette in procaryotic and eukaryotic cells. The additional DNA sequences include origins of replication to provide for autonomous replication of 15 the vector, selectable marker genes, preferably encoding antibiotic resistance, unique multiple cloning sites providing for multiple sites to insert DNA sequences or genes encoded in the expression cassette, and sequences that enhance transformation of prokaryotic and eukaryotic cells. The preferred vectors of the invention are plasmid vectors. The especially preferred vector is 20 the pBl21 or pBI221 vector formed as described by Jefferson cited supra. Another vector that is useful for expression in both plant and procaryotic cells is the binary Ti vector PGA582. This binary Ti vector has been previously characterized by An, cited supra., and is available from Dr. An. This binary Ti vector can be replicated in procaryotic bacteria such as E. Cli and 25 Agrobacterium. The Agrobacterium plasmid vectors can be used to transfer the expression cassette to plant cells. The binary Ti vectors preferably include the nopaline T DNA right and left borders to provide for efficient plant cell transformation, a selectable marker gene, unique multiple cloning sites in the T border regions, the colEl replication of origin and a wide host range replicon. 30 The binary Ti vectors carrying an expression cassette of the invention can be used to transform both prokaryotic and eukaryotic cells, but is preferably used to transform plant cells.

WO 99/67367 PCT/US99/14022 24 B. Method for Screening for Expression and/or Overexpression of a Plant Acetyl CoA Carboxylase Gene A method for screening for expression or overexpression of a 5 plant acetyl CoA carboxylase gene is also provided by the invention. Once formed, an expression cassette comprising an acetyl CoA carboxylase gene can be subcloned into a known expression vector. The screening method in the invention includes the steps of introducing an expression vector into a host cell and detecting and/or quantitating expression of a plant acetyl CoA carboxylase 10 gene. This method of screening is useful to identify expression cassettes providing for an overexpression of a plant acetyl CoA carboxylase gene, antisense molecules that effectively inhibit acetyl CoA carboxylase synthesis, and expression of an acetyl CoA carboxylase in the chloroplast of a transformed plant cell. 15 Suitable known expression vectors include plasmids that autonomously replicate in prokaryotic and eukaryotic cells. Specific examples include plasmids such as the pBI121 or pBI221 plasmid constructed as described by Jefferson cited supra, a binary Ti vector such as PG582 as described by An cited supra, PUCI 19, or PBR322. The preferred expression system is a pBI121 20 or pBI221 plasmid. An expression cassette of the invention can be subcloned into an expression vector by standard methods. The expression vector can then be introduced into prokaryotic or eukaryotic cells by standard methods including protoplast transformation, silicon fiber-based transformation Agrobacterium 25 mediated transformation, electroporation, microprojectiles, e.g., preferably using gold particles of about 6 to about 10 microns in diameter although diameters that are about 100 to about 1000 times smaller may also be employed, and liposomes. The expression vector can be introduced into plant cells such as tobacco, Brassica, Black Mexican sweet corn, and Arabidopsis cells. The vector can also 30 be introduced into procaryotic cells such as E. coli or Agrobacterium. Transformed cells can be selected typically using a selection marker encoded on the expression vector. Transient expression of a plant acetyl CoA carboxylase gene can be detected and quantitated in the transformed cells. Gene expression can be WO 99/67367 PCT/US99/14022 25 quantitated by a quantitative Western blot using antibodies specific for the cloned acetyl CoA carboxylase or by detecting an increase in specific activity of the enzyme. The tissue and subcellular location of the cloned acetyl CoA carboxylase can be determined by immunochemical staining methods using 5 antibodies specific for the cloned acetyl CoA carboxylase. Sensitivity of the cloned acetyl CoA carboxylase to herbicides can also be assessed. Expression cassettes providing for overexpression of a plant acetyl CoA carboxylase or acetyl CoA carboxylase tolerant to herbicides can then be used to transform monocot and/or dicot plant tissue cells and to regenerate transformed plants and 10 seeds. C. Method of Imparting Cyclohexanedione or Aryloxyphenoxypropanoic Acid Herbicide Tolerance to a Plant The invention provides a method of conferring cyclohexanedione 15 or aryloxyphenoxypropanoic acid herbicide tolerance to a plant. The method includes the steps of introducing an expression cassette comprising a gene coding for a plant acetyl CoA carboxylase or a functional mutant thereof operably linked to a promoter into the cells of plant tissue and expressing the gene in an amount effective to render the cells of the plant tissue substantially 20 tolerant to herbicides. An effective amount of gene expression to render the cells of the plant tissue substantially tolerant to the herbicide depends on whether the gene codes for an unaltered acetyl CoA carboxylase gene or a mutant or altered form of the gene that is less sensitive to the herbicides. Expression of an unaltered plant acetyl CoA carboxylase gene in an effective amount is that 25 amount that provides for a 2- to 50-fold increase in herbicide tolerance and preferably increases the amount of acetyl CoA carboxylase from at least about 2 to 20-fold over that amount of the native enzyme. An altered form of the enzyme can be expressed at levels comparable to that of the native enzyme or less if the altered form of the enzyme has higher specific activity. Acetyl CoA 30 carboxylase substantially tolerant to herbicides is an enzyme that is tolerant of levels of herbicide which normally inhibit a native acetyl CoA carboxylase and preferably can function in concentrations of herbicide of about 2- to 20-fold greater than are toxic to the native enzyme.

WO 99/67367 PCT/US99/14022 26 Herbicide tolerance can be achieved by at least two methods including: 1) by increasing the level of gene expression of a native or unaltered acetyl CoA carboxylase gene; or 2) by introducing an altered gene coding for an acetyl CoA carboxylase that is less sensitive to herbicide inhibition. The level of 5 gene expression can be increased by either combining a plant acetyl CoA carboxylase gene with a promoter that provides for a high level of gene expression such as the 35S Came promoter or by introducing the gene into the cells so that multiple copies of the gene are integrated into the genome of the transformed plant cell. Formation of an expression cassette comprised of a plant 10 acetyl CoA carboxylase gene operably linked to a promoter that can be expressed in an effective amount to confer herbicide tolerance has been described previously. Most monocots, but not dicots, are sensitive to cyclohexanedione and/or aryloxyphenoxypropanoic acid herbicides. The preferred plant cells for 15 introducing an expression cassette of the invention to achieve herbicide tolerance for the plant cells then are monocot plants. Monocot plants include corn, wheat, barley, sorghum, rice, and others. An expression cassette of the invention can be introduced by methods of transformation, especially effective for monocots including biolistic transformation of Type II embryogenic suspension cells as 20 described by W.J. Gordon-Kamm et al., Plant Cell, 2, 603-618 (1990), M.E. Fromm et al., Bio/Technology, 8, 833-839 (1990) and D.A. Walters et al., Plant Molecular Biology, 18, 189-200 (1992) or by electroporation of type 1 embryogenic calluses described by D'Hafluin et al., The Plant Cel1, 4, 1495 (1992). Transformed cells can be selected for the presence of a selectable 25 marker gene. Transient expression of a plant acetyl CoA carboxylase gene can be detected in the transgenic embryogenic calli using antibodies specific for the cloned plant acetyl CoA carboxylase. Transformed embryogenic calli can be used to generate transgenic plants that exhibit stable inheritance of either the altered acetyl CoA carboxylase gene or overexpression of the acetyl CoA 30 carboxylase gene. Maize cell lines exhibiting satisfactory levels of tolerance to herbicide are put through a plant regeneration protocol to obtain mature maize plants and seeds expressing the tolerance traits such as described in D'Hafluin, cited supra., or An, cited supra. The plant regeneration protocol allows the WO 99/67367 PCT/US99/14022 27 development of somatic embryos and the subsequent growth of roots and shoots. To determine that the herbicide-tolerance trait is expressed in differentiated organs of the plant, and not solely in undifferentiated cell culture, regenerated plants are exposed to herbicide levels which will normally inhibit shoot and root 5 formation and growth. Mature maize plants are then obtained from maize cell lines that are known to express the trait. If possible, the regenerated plants are self pollinated. Otherwise, pollen obtained from the regenerated plants is crossed to seed-grown plants of agronomically important inbred lines. Conversely, pollen 10 from plants of these inbred lines is used to pollinate regenerated plants. The genetics of the trait are then characterized by evaluating the segregation of the trait in the first and later generation progeny. Stable inheritance of overexpression of a plant acetyl CoA carboxylase or a functional mutant of a plant acetyl CoA carboxylase conferring herbicide tolerance to the plant is 15 achieved if the plants maintain herbicide tolerance for at least about three to six generations. Seed from transformed monocot plants regenerated from transformed tissue cultures is grown in the field and self-pollinated to generate true breeding plants. Progenies from these plants become true breeding lines 20 which are evaluated for herbicide tolerance in the field under a range of environmental conditions. Herbicide tolerance must be sufficient to protect the monocot plants at the maximum labeled delivery rate under field conditions which cause herbicides to be most active. Appropriate herbicide concentrations and methods of application are those which are known and have been developed 25 for the cyclohexanedione and/or aryloxyphenoxypropanoic acid herbicides in question. In a preferred version, an expression cassette comprised of a maize acetyl CoA carboxylase gene isolated from a maize cell line tolerant to sethoxydim and haloxyfop and linked to the 35S Came promoter is introduced 30 into an herbicide sensitive monocot tissue using biolistic transformation. Transformed calli are selected and used to generate transgenic plants. Transformed calli and transgenic plants can be evaluated for tolerance to sethoxydim and haloxyfop and for stable inheritance of the tolerance trait.

WO 99/67367 PCTIUS99/14022 28 D. Method for Altering the Oil Content in a Plant The invention also provides a method of altering the oil content in a plant. The method include the steps of introducing an expression cassette comprising a gene coding for plant acetyl CoA carboxylase or functional mutant 5 thereof operably linked to a promoter functional in a plant cell into the cells of plant tissue and expressing the gene in an amount effective to alter the oil content of the plant cell. An alteration in the oil content of a plant cell can include a change in the total oil content over that normally present in that type of plant cell, or a change in the type of oil from that normally present in the plant 10 cell. Expression of the gene in an amount effective to alter the oil content of the gene depends on whether the gene codes for an unaltered acetyl CoA carboxylase or a mutant or altered form of the gene. Expression of an unaltered plant acetyl CoA carboxylase gene in an effective amount is that amount that may provide a change in the oil content of the cell from about 1.2- to 20-fold 15 over that normally present in that plant cell, and preferably increases the amount of acetyl CoA carboxylase about 2- to 20-fold over that amount of the enzyme normally present in that plant cell. An altered form of the enzyme can be expressed at levels comparable to that of the native enzyme or less if the altered form of the enzyme has higher specific activity. 20 An alteration in oil content of the plant cells according to the method of the invention can be achieved in at least two ways including: (1) an increase or decrease in expression of an unaltered plant acetyl CoA carboxylase gene; or (2) by introducing an altered or functional mutant plant acetyl 25 CoA carboxylase gene coding for an enzyme that exhibits a change in specific activity. The level of gene expression of an unaltered plant acetyl CoA carboxylase gene can be increased by either combining an unaltered plant acetyl CoA carboxylase gene with a promoter that provides for a high level of gene expression, such as 30 the 35S cauliflower mosaic virus or by introducing the expression cassette and/or selecting for plant cells having multiple copies of a plant acetyl CoA carboxylase gene integrated into the genome. A decrease in expression of an unaltered acetyl CoA carboxylase can be achieved by transformation with an ACCase antisense WO 99/67367 PCT/US99/14022 29 gene containing an expression cassette. When an altered or functional mutant plant acetyl CoA carboxylase gene codes for an enzyme that has an increase in specific activity, it may lead to an increase in total oil content of a plant cell even if the level of gene expression is comparable to that of the native enzyme. When 5 an altered or functional mutant acetyl CoA carboxylase gene codes for an enzyme having a decrease in specific activity, it may lead to a decrease in the total oil content of the plant cell compared to that normally present. An expression cassette as described above can be introduced into either monocots or dicots. Preferably, the expression cassette is introduced into 10 dicot plants such as soybean, canola, and sunflower. An expression cassette can be introduced by standard methods including protoplast transformation, Agrobacterium-mediated transformation, microprojectiles, electroporation, and the like. Transformed cells or tissues can be selected for the presence of a selectable marker gene. 15 Transient expression of a plant acetyl CoA carboxylase gene can be detected in transformed cells or tissues by immunoreactivity with antibodies specific for the cloned acetyl CoA carboxylase. Overexpression of a plant acetyl CoA carboxylase can be detected by quantitative Western blots. A change in specific activity of the enzyme can be detected by measuring enzyme activity in 20 the transformed cells. A change in total oil content can also be examined by standard methods, as described in Clark & Snyder, JACS, 66:1316 (1989). Transgenic plants and seeds can be generated from transformed cells and tissues showing a change in oil content or in the amount or specific activity of a plant acetyl CoA carboxylase using standard methods. It is 25 especially preferred that the oil content of the leaves, seeds, or fruits is increased. In a preferred version, a maize acetyl CoA carboxylase gene is combined with a 35S cauliflower mosaic virus promoter in a vector such as pBIl21 or pBI221 and introduced into soybean cells using the microprojectile method. Transformed soybean cells showing an increase in expression of acetyl 30 CoA carboxylase of at least about 2-fold or at least a 1.2-fold increase in oil content are selected. Transformed soybean cells exhibiting overexpression of acetyl CoA carboxylase or showing an increase in total oil content are used to generate transgenic plants and seeds.

WO 99/67367 PCT/US99/14022 30 E. Method of Producing Plant Acetyl CoA Carboxylase The invention also provides a method of producing plant acetyl CoA carboxylase in a host cell. The method includes the steps of introducing an expression cassette comprised of a gene encoding a plant acetyl CoA 5 carboxylase or functional mutant thereof into a host cell and expressing the gene in an amount sufficient to allow for crystallization of the plant acetyl CoA carboxylase. An amount sufficient to allow for crystallization of a plant acetyl CoA carboxylase is about 20- to 100-fold increase over the amount of plant acetyl CoA carboxylase that can normally be purified from plant cells, preferably 10 about 2 to 10 mg protein. Crystallized plant acetyl CoA carboxylase can be used to identify other herbicides that can bind to and inhibit acetyl CoA carboxylase function. In addition, the availability of large amounts of purified enzyme provides for screening of the efficacy of such herbicides. An expression cassette can include a promoter that is functional in 15 either a eukaryotic or prokaryotic cell. The expression cassette can be introduced into a prokaryotic cell such as E. cali, or a eukaryotic cell such as a plant or yeast. The preferred cell is a prokaryotic cell used routinely in producing recombinant proteins such as E. cali. The expression cassette can be introduced and transformed cells selected by standard methods. 20 The plant acetyl CoA carboxylase gene can be expressed in an prokaryotic cell until sufficient amount of the enzyme is produced so that it can be crystallized. Plant acetyl CoA carboxylase can be isolated from bacterial cells using standard methods, including those described in Example V. The purified acetyl CoA carboxylase can then be crystallized and characterized by standard 25 methods. EXAMPLE I Identification of Herbicide Mechanism and Site of Action The objective of this Example was to identify the mechanism 30 whereby sethoxydim and/or haloxyfop inhibit fatty acid synthesis in maize. The results, reported in J. D. Burton et al., Biochem. Biophys. Res. Comm., -148, 1039 (November 13, 1987), show that both sethoxydim and haloxyfop inhibit acetyl-coenzyme A carboxylase (ACCase) (EC 6.4.1.2) in maize chloroplasts.

WO 99/67367 PCTIUS99/14022 31 A. Chemicals Buffers and cofactors were purchased from Sigma Chemical Company (St. Louis, Missouri); [2- 1 4 C]acetate was purchased from Research Products International; [2- 1 4 C]pyruvate and [ 1 4 C]NaHCO 3 were purchased from 5 New England Nuclear; and [2-14C]malonyl coenzyme A was purchased from Amersham. Sethoxydim was a gift from BASF (Parsippany, New Jersey), and haloxyfop was provided by Dow Chemical USA (Midland, Michigan). B. Plant Growth Conditions Corn (Z.maysL., 'B37 x Oh43') seeds were germinated in 10 darkness for 96 hours in vermiculite in an incubation chamber maintained at 30'C, 80% RH. Seedlings were then transferred to a growth chamber with a 16 hour light (25 C) and an 8 hour dark (20'C) cycle, 90% relative humidity (RH). After greening 48 hours, seedlings were returned to the dark incubation chamber for 12 hours to deplete chloroplast starch reserves. Seedlings were harvested 6 15 days after planting. Pea (P. sativumI T., 'PI 9901-C') seedlings were grown in vermiculite in a growth chamber with a 16 hour light (21 *C) and 8 hour dark (16*C) cycle, 80% RH. Peas were harvested 10 to 13 days after planting. Black Mexican Sweet (BMS) corn suspension cultures were maintained in a supplemented Murashige-Skoog (MS) medium (C. E. Green, Hort. Sci., 12, 7-10 20 (1977)), and subcultured weekly by 20-fold dilution of the suspension culture into fresh medium. C. Chloroplast Isolation Chloroplasts from corn and pea seedlings were isolated at 4*C (K. Cline et al., J. Biol. Chem., 2fiQ, 3691-3696 (1985)). Seedlings (50 g of shoots) 25 were homogenized in 200 ml buffer A (50 mM HEPES-NaOH pH 7.5, 330 mM sorbitol, 0.1% w/v BSA, 1 mM MgCl 2 , 1 mM MnCl 2 , 2 mM EDTA, 5 mM isoascorbate, 1.3 mM glutathione) in an onmimixer (five, 3-second bursts at full speed). The homogenate was filtered through six layers of cheesecloth and two layers of miracloth, and then centrifuged at 3000 g for 3 minutes with hand 30 braking. The pellet was gently resuspended in buffer A and layered onto a preformed linear Percoll gradient (50 mM HEPES-NaOH pH 7.5, 330 mM sorbitol, 1.9 mM isoascorbate, 1.08 mM glutathione, 0.1% w/v BSA, 50% Percoll) which was centrifuged at 3000 g for 20 minutes in a Sorvall HB-4 rotor.

WO 99/67367 PCT/US99/14022 32 The lower band in the gradient, containing intact chloroplasts, was washed twice by gently resuspending it in 20 ml of buffer B (50 mM HEPES-NaOH, pH 7.5, and 330 mM sorbitol) followed by repelleting (3000 g, 5 minutes). The final pellet, consisting of intact chloroplasts, was resuspended in 2 to 3 ml of buffer B 5 and stored on ice in the dark until use. D. Fatty Acid Synthesis

[

1 4 C]acetate and [1 4 C]pyruvate were used as precursors to measure fatty acid biosynthesis in isolated chloroplasts (B. Liedvogel et al., Planta, 169, 481-489 (1986)). [ 1 4 C]acetate incorporation was assayed in a 0.5 ml-volume 10 containing: 50 mM HEPES-NaOH (pH 7.5), 330 mM sorbitol, 5 mM KH 2 P0 4 , 10 mM NaHCO 3 , 1 mM MgCl 2 , 1 mM ATP, 0.1 mM CoA, 0.15 mM [1 4 C]acetate (3.33 mCi/mmol), and chloroplasts (20 to 50 pg chlorophyll).

['

4 C]pyruvate incorporation into fatty acids was assayed in the same medium except that it included 2 mM TPP, 1 mM NAD*, 0.15 mM [1 4 C]-pyruvate (1.33 15 mCi/mmol), but no acetate. Assay suspensions were illuminated with 1400 gE/m 2 -second PAR at 25'C. Assays were initiated by the addition of the labeled substrate and stopped by the addition of 0.5 ml of 40% KOH. To determine the incorporation of radiolabel into a non-polar (fatty acid) fraction, each treatment was saponified at 90'C for 30 minutes in capped vials (P. B. Hoj et al., Carlsberg 20 Res. Common., 47, 119-141 (1982)). The vials were acidified with 0.5 ml 40%

H

2

SO

4 , and carrier fatty acids (20 pg each of C 14:0, C 16:0, and C 18:0) were added. The assay mixture was extracted twice with 4 ml hexane. The extracts were combined, dried under N 2 , and redissolved in 0.3 ml hexane. Aliquots (50 l) were counted for radioactivity by liquid scintillation spectrometry. 25 Incorporation of [ 1 4 C]malonyl-Coenzyme A into fatty acids (P. B. Hoj et al., supra; and J. B. Ohlrogge et al., Proc. Natl. Acad. Sci. USA, 26, 1194 1198 (1979)) was assayed using cell-free preparations from BMS tissue culture. Cells harvested during logarithmic growth phase were frozen in liquid nitrogen, ground with a mortar and pestle, and thawed in a medium containing: 0.1 M 30 HEPES-KOH, pH 7.5; 0.3 M glycerol, and 5 mM DTT (buffer:tissue, 2:1, v/w). The homogenate was centrifuged at 12,000 g for 20 minutes. The supernatant was filtered through miracloth and centrifuged (125,000 g) for 60 minutes and then filtered through miracloth and assayed. Assays were conducted at 25'C in a WO 99/67367 PCTIUS99/14022 33 0.4 ml volume containing: 1.0 mM ATP, 0.32 mM NADPH, 0.38 mM NADH, 25 pM CoA, 10 ptM acetyl-CoA, 25 pg acyl-carrier protein, and 12 gM malonyl CoA (11.54 pCi/ptmol). Reactions were initiated by addition of [ 1 4 C]malonyl CoA and stopped by addition of 0.4 ml 40% KOH. Label incorporation into 5 fatty acids was determined as above. Chlorophyll (D. I. Arnon, Plant Physiol., 24, 1-15 (1949)) and protein (P. K. Smith et al., Anal. Biochem., 15_0, 76-85 (1985)) were determined as described therein. E. Acetyl-Coenzyme A Carboxylase (ACCase) Activity Maize chloroplasts, isolated as described above, were suspended 10 in buffer C (0.1 M Tricine-KOH, pH 8.0; 0.3 M glycerol, and 1 mM DTT) and homogenized in a glass tissue homogenizer. The disrupted chloroplast fraction was centrifuged at 16,000 g for 15 minutes. The supernatant was desalted on a Sephadex G-25 column (1.5 x 5 cm equilibrated with 0.1 M Tricine-KOH, pH 8.0; and 0.3 M glycerol) and assayed directly. ACCase activity (B. J. Nikolau et 15 al., Arch. Biochem. Biophys., 211, 605-612 (1981)) was assayed at 30'C in a 0.2 ml volume which contained 1 mM ATP, 3 mM acetyl coenzyme A, 2.5 mM MgCl 2 , 50 mM KCl, 0.5 mM DTT, and 15 mM [ 1 4 C]NaHCO 3 (0.17 mCi/mmol). Reactions were initiated by addition of acetyl coenzyme A and stopped by addition of 25 pl of 12 N HCl. Product formation was determined by the 20 radioactivity found in an acid stable fraction by liquid scintillation spectrometry. Enzyme activity was linear for 15 minutes. F. Results To probe for the site of herbicidal activity of sethoxydim and haloxyfop, labeled acetate, pyruvate, and malonyl-CoA were used individually 25 as precursors for fatty acid synthesis. Isolated chloroplasts from corn seedlings incorporated [1 4 C]acetate and [14C]pyruvate into a non-polar fraction (fatty acids). Acetate incorporation was linear for 30 min after a 5 min lag period, and dependent upon the addition of free acetyl coenzyme A. Addition of either 10 pM sethoxydim or 1 gM haloxyfop inhibited [ 1 4 C]acetate incorporation into fatty 30 acids by 90% and 89%, respectively, as shown in Table I, below. Sethoxydim (10 pM) and haloxyfop (1 gM) also inhibited the incorporation of [ 1 4 C]pyruvate into fatty acids by 98% and 99%, respectively.

WO 99/67367 PCT/US99/14022 34 TABLE I Inhibition of [ 4 C]acetate and [ 14 C]pyruvate Incorporation into Fatty Acids in 5 Corn Seedling Chloroplasts by Sethoxydim (10 pM) and Haloxyfop (1 ptM), 10 minute assay time 10 Acetate Pyruvate ---Activity (nmol/mg chl-min)-- 15 Control 4.4 0.41 10.8 ±2.3 -------------- % Inhibition--------------- Sethoxydim 90 ±2.5 98± 1.1 Haloxyfop 89 ±3.1 99 ±0.3 20 'Results are expressed as mean of two experiments ± standard error. The effect of 10 tM sethoxydim and 1 pM haloxyfop on

[

4 C]malonyl-CoA incorporation into fatty acids was determined using cell-free 25 extracts from corn suspension cultures. Neither sethoxydim (10 pM) nor haloxyfop (1 iM) inhibited fatty acid synthetase activity. Thus, both herbicides inhibited fatty acid synthesis in intact chloroplasts from corn seedlings with either acetate or pyruvate as a precursor, but did not inhibit incorporation of malonyl-CoA into fatty acids. This suggests that ACCase which catalyzes the 30 formation of malonyl-CoA is the site of action of these herbicides. EXAMPLE II Selection and Characterization of Herbicide-tolerant Cell Lines A selection protocol to identify and isolate herbicide-tolerant 35 maize cells was developed to minimize the adverse effects of high herbicide concentrations on somatic embryo development and plant regeneration capacity. The procedure involved exposing tissue to gradually increasing concentrations of herbicide beginning with a sethoxydim concentration representing 1/20th of lethal dose and doubling the herbicide concentration at approximately two-week 40 intervals until the lethal dose (10 pM sethoxydim) was reached. In this way, the WO 99/67367 PCT/US99/14022 35 herbicide was allowed to take effect slowly with continuous selection pressure, thus permitting herbicide-tolerant cells to accumulate over time while not affecting the potential for plant regeneration. A. Selection of a Sethoxydim-Tolerant Cell Line 5 Many selections were carried out utilizing the selection protocol described in the preceding paragraph. The selection of one such sethoxydim tolerant cell line that was identified and characterized is described below in detail. Approximately 100 grams of vigorously growing, regenerable, 10 friable, embryogenic maize callus tissue established from an F, immature embryo resulting from the cross A188 x B73 were transferred to agar-solidified maintenance medium (Armstrong and Green, Planta, 164, 207 (1985)) in petri plates containing 0.5 ptM sethoxydim (BASF) (Parsippany, New Jersey). This callus line was designated 2167-9/2160-154. Forty plates were prepared and five 15 clumps of callus tissue weighing about 0.5 grams each were placed on each plate. The 0.5 [tM sethoxydim concentration was chosen from growth inhibition studies to provide less than 10-20% growth inhibition during the first two weeks of herbicide exposure. After 14 days, 0.25-0.5 gram pieces of tissue showing vigorous growth rate and retention of embryogenic morphology (i.e., presence of 20 somatic embryos) were subcultured on fresh medium containing 1.0 gM sethoxydim. Eighty plates containing five pieces of tissue per plate were prepared. For each subsequent transfer, all callus tissue showing growth and somatic embryo forming ability was placed on fresh media containing a two-fold increased sethoxydim concentration. Therefore, callus was transferred at two 25 week intervals to petri plates containing 0.5, 1.0, 2.0, 5.0 and 10.0 pM sethoxydim. During the course of the selection process, the total number of lines decreased as the herbicide-mediated growth inhibition became more intense. Cell lines exhibiting growth on 10 ptM sethoxydim were designated as herbicide tolerant and given an identification number. Two sethoxydim-tolerant lines were 30 recovered that exhibited uninhibited growth at 10 iM sethoxydim. These lines were designated 2167-9/2160-154 S-1 and 2167-9/2160-154 S-2.

WO 99/67367 PCTIUS99/14022 36 B. Characterization of Herbicide-Tolerant Maize Cell Line 2167-9/2160-154 S-2 Tolerant cell line 2167-9/2160-154 S-2 ("S-2") was characterized to evaluate: (1) the magnitude of sethoxydim tolerance; (2) cross-tolerance of 5 haloxyfop; and (3) the biochemical basis for the tolerance. Callus tissue from S 2 that had been maintained on 10 pM sethoxydim was transferred to media containing up to 100 gM sethoxydim. One-half gram of S-2 tissue was plated on a 7 cm filter paper as a lawn overlaying 50 ml agar-solidified culture medium containing 0, 0.5, 1.0, 2.0, 5.0, 10.0, 50.0 and 100 pM sethoxydim, and cultured 10 for two weeks. Control cell line 2167-9/2160-154 was plated similarly on medium containing the same levels of sethoxydim. The results of this study are summarized in Figure 2. The control cell line growth after two weeks was inhibited 50% at 1 gM sethoxydim. Growth of S-2 was not inhibited at 100 pM sethoxydim, indicating that S-2 was at least 100-fold more tolerant than the 15 control callus line. Growth of S-2 was inhibited with 0.65 pM haloxyfop, whereas the control cell line was inhibited 50% with 0.02 pM, indicating approximately a 30-fold increase in tolerance. C. Acetyl-Coenzyme A Carboxylase (ACCase) Activity of Maize Cell 20 Line S-2 Assays were conducted to determine if ACCase extracted from cell line S-2 was altered with respect to herbicide activity. ACCase activity of control tissue was 50% inhibited either by 1.5 pM sethoxydim, or by 0.25 pM haloxyfop. ACCase activity of S-2 tissue was inhibited 50% either by 70 pM 25 sethoxydim, or by 1.8 pM haloxyfop, indicating at least 40-fold and 7-fold decreases in herbicide sensitivity on concentration basis, respectively. EXAMPLE III Plant Regeneration and Production of Herbicide-Tolerant Seed A. Plant Regeneration Protocol 30 Sixteen ca. 150 mg clumps of S-2 callus were transferred per 25 x 100 mm petri plate containing agar-solidified N6 basal salts and 6% sucrose and incubated 7-14 days in low light (20 jEi m 2 s-). Several plates containing callus on plant regeneration medium were prepared. Callus was transferred to agar- WO 99/67367 PCTIUS99/14022 37 solidified Murashige-Skoog (MS) medium without hormones and incubated in high intensity light (200 tE m-2 s-) for shoot elongation. Developing plants (1 3 cm long) were isolated from the callus surface and transferred to magenta boxes containing agar-solidified MS salts, 2% sucrose with no hormones for two 5 weeks of further growth. When plants reached the 2-3 leaf stage, they were transplanted to peat pots containing potting soil, and were incubated in the growth room until growing stably. Surviving plants were transferred to soil in 4" diameter plastic pots and grown in the greenhouse. B. Expression of Herbicide Tolerance in Plants Regenerated from S-2 10 Callus Tissue Groups of eight control (2167-9/2160-154 unselected) and eight S-2 plants were sprayed with either 0.0, 0.01, 0.05, 0.11, 0.22 or 0.44 kg/ha sethoxydim to determine whole plant sethoxydim-tolerance of greenhouse grown plants. Control plants were killed by 0.05 kg/ha or more sethoxydim. 15 Plants regenerated from the S-2 cell line survived the 0.44 kg/ha sethoxydim treatment, indicating that S-2 plants exhibit at least 20-fold more tolerance of sethoxydim than control. Figure 3 shows the growth response of the regenerated plants seven days after treatment with 0.44 kg/ha sethoxydim. As shown in Figure 4, shoot height of regenerated S-2 plants was only slightly reduced 14 20 days after treatment with 0.44 kg/ha sethoxydim. C. Seed Production from S-2 Plants Plants surviving sethoxydim treatments of up to 0.44 kg/ha were transplanted to the genetics plot on the University of Minnesota campus, St. Paul, Minnesota. Additional S-2 plants were transplanted to the field that had 25 not been sprayed. Sixty-five 2167-9/2160-154 control plants and ninety-five S-2 plants were grown to maturity in the field. Plants were either self-pollinated or cross-pollinated to inbred maize lines A188, A619, A641, A661, A665, B37, B73, R806, and W153R. Control seed were produced by selfing 2167-9/2160 154 regenerated plants, or by crossing them with the inbreds listed above. 30 D. Expression of Herbicide Tolerance in Progeny of Regenerated Plants Seeds obtained by the crossing procedure described above were viable and germinated normally. Seeds from thirty S-2 selfed plants and fifteen 2167-9/2160-154 control plants were planted in 25 x 50 cm trays of soil (28 WO 99/67367 PCT/US99/14022 38 seeds from each plant in one tray) and grown in the greenhouse. Seedlings at the 3-4 leaf stage were treated with 0.1, 0.44, and 1.1 kg/ha sethoxydim and evaluated for visual herbicide damage and shoot height. Based on visual rating of herbicide damage two weeks after treatment, selfed progeny of S-2 plants 5 segregated approximately 1:2:1 for healthy, uninjured plants: to plants showing partial injury: to dead plants, respectively, at 0.44 and 1.1 kg/ha sethoxydim treatments. All control progeny of 2167-9/2160-154 control plants were killed by 0.1 kg/ha and greater levels of sethoxydim. These results demonstrate dominant expression of sethoxydim tolerance indicating that sethoxydim 10 tolerance in S-2 plants is a heritable trait. Similar tests were conducted on progeny of S-2 plants crossed to the other inbreds. In all cases, these test cross progeny treated with 0.44 kg/ha sethoxydim segregated 1:1 for growing shoots versus dead shoots whether S-2 plants were used as male or female parents. These results confirm that sethoxydim tolerance is controlled by a single 15 dominant nuclear gene. In all cases, control plants crossed to the other inbreds were killed and therefore sethoxydim-sensitive. E. Method for Obtaining Uniform Herbicide-Tolerant Seed Progeny of S-2 plants surviving sethoxydim treatments of 0.44 and 1.1 kg/ha and showing no herbicide injury were transferred to the 20 greenhouse and grown to maturity. These plants may be selfed and their progeny evaluated for sethoxydim and haloxyfop tolerance to identify pure breeding herbicide-tolerant maize lines. Progeny of S-2 plants crossed to inbred lines and exhibiting sethoxydim tolerance may be recurrently backcrossed to the same inbreds. 25 Progeny of each cross may be screened for sethoxydim-tolerance, and tolerant plants grown to maturity and again crossed to the recurrent parent. After six or seven cycles of backcrossing, sethoxydim-tolerant plants may be selfed and progeny screened for tolerance to produce homozygous sethoxydim tolerant maize inbreds.

WO 99/67367 PCT/US99/14022 39 EXAMPLE IV Selection of Additional Herbicide-Tolerant Maize Cell Lines One primarily sethoxydim-tolerant maize cell line, 2167-9/2160 5 154 S-1, and two haloxyfop-tolerant maize cell lines, 2167-9/2160-154 H-1 and 2167-9/2160-154 H-2, were selected and characterized as follows: A. Selection of Maize Cell Line 2167-9/2160-154 S-1 Maize cell line 2167-9/2160-154 S-1 was selected from maize cell culture using the protocol described in detail above for the selection of Line 10 2167-9/2160-154 S-2. Approximately 70 plants were regenerated from Line 2167-9/2160-154 S-1, and either self-pollinated or cross-pollinated to the inbred maize lines A188, A619, A641, A661, A665, B37, B73, R806, and W153R. B. Selection of Maize Cell Line 2167-9/2160-154 H-1 Line 2167-9/2160-154 H-1 was selected from maize cell culture 15 using a similar protocol described in detail above except maize callus tissue was selected using the herbicide haloxyfop. Maize callus tissue was initially plated on 0.01 ptM haloxyfop. At two-week intervals, surviving tissue was subcultured onto 0.05, 0.10 and 0.20 pM haloxyfop. Approximately 50 plants were regenerated from Line 2167-9/2160-154 H-1, and were self-pollinated. 20 C. Selection of Maize Cell Line 2167-9/2160-154 H-2 Line 2167-9/2160-154 H-2 was selected from maize cell culture using a similar protocol described in detail for line 2167-9/2160-154 H-1. No plants have been successfully regenerated from this line. D. Characterization of Lines 2167-9/2160-154 S-1, H-1 and H-2 25 The tolerant callus cultures were characterized to determine the magnitude of sethoxydim and haloxyfop tolerance. Callus tissue from these lines was evaluated in experiments as described above in the characterization of line 2167-9/2160-154 S-2. Table II summarizes the results of these studies. Line 2167-9/2160-154 S-1 and Line 2167-9/2160-154 H-2 showed a four-fold 30 increase in haloxyfop tolerance, while Line 2167-9/2160-154 H-1 exhibited approximately a 60-fold increase in haloxyfop tolerance. Neither haloxyfop selected line showed a significant degree of sethoxydim tolerance, while the WO 99/67367 PCTIUS99/14022 40 sethoxydim selected line S-I exhibited approximately a 100-fold increase in sethoxydim tolerance. TABLE 11 Herbicide Tolerance of Cell Lines S-1, H-1 and H-2 5 Herbicide Cell Line Haloxyfop Sethoxydim 2167-9/2160-154 S-1 41 100 2167-9/2160-154 H-1 61 0 10 2167-9/2160-154 H-2 4 0 'The numbers represent the fold increase in herbicide concentration that results in a 50% reduction in growth of the selected cell lines compared to the unselected control cell line 2167-9/2160-154. 15 E. Herbicide Inhibition of Acetyl Coenzyme A Carboxylase of Maize Cell Lines S-1, H-1 and H-2 Acetyl Coenzyme A Carboxylase (ACCase) was extracted from 20 cell lines S-1, H-1 and H-2 and assayed as described in detail for maize cell line S-2, above. Table III below summarizes the results of these studies. The ACCase from line S-1 was more tolerant of both sethoxydim and haloxyfop, while the ACCase from line H-I was more tolerant of haloxyfop, but not of sethoxydim. The ACCase from line H-2 showed no difference from the 25 unselected parent line 2167-9/2160-154 in sensitivity to either herbicide. However, cell line H-2 exhibited approximately a five-fold higher level of ACCase activity as compared to the unselected parent line 2167-9/2160 154. Thus, selection for sethoxydim or haloxyfop tolerance resulted in a less WO 99/67367 PCT/US99/14022 41 sensitive ACCase in cell line S-1 and H-1, as well as a higher level of ACCase activity in cell line H-2. TABLE III 5 Herbicide Inhibition of ACCase of Maize Cell Lines S-1, H-1 and H-2 Herbicide Cell Line Haloxyfop Sethoxydim 10 2167-9/2160-154 S-i 3 4 2167-9/2160-154 H-1 7 0 2167-9/2160-154 H-2 0 0 15 'The numbers represent the fold increase in herbicide concentration that inhibits ACCase activity of the selected cell lines by 50% compared to the unselected parent cell line 2167-9/2160-154. 20 Deposit of Seeds Seeds from representative S-2 plants (Ex. III (B)) and H-i plants (Ex. IV(B)) have been deposited with the American Type Culture Collection, 12301 Parklawn Drive, Rockville, MD 20852 USA on October 25, 1988 and assigned accession numbers ATCC 40507, and ATCC 40508, respectively. 25 EXAMPLE V Formation of cDNA Clones Encoding ACCase A. ACCase Purification The acetyl CoA carboxylase enzyme was isolated and purified from plant tissues and characterized. The purified enzyme was used to generate 30 antibody reagents useful in identifying cDNA clones encoding the gene or portions of the gene for ACCase. ACCase was extracted from frozen shoots of 7-d-old maize (Zea Mays L. inbred A619 or B73) seedlings grown in a growth chamber (24 C, 90% RH, 16-h daylength at 210 pE m- 2 s-). The outermost leaf and blade were 35 removed and the remainder of the shoot was frozen in liquid N 2 . Embryos and endosperm tissue from developing kernels were harvested from field-grown ears at 36 to 40 days after pollination (DAP). Black Mexican Sweet corn (BMS) WO 99/67367 PCT/US99/14022 42 maize suspension cells were obtained from cultures as previously described (W.B. Parker et al., Plant PhysioL, 92, 1220-1225 (1990)). Tissues were stored in liquid N 2 until used. Extraction and purification steps were performed at 0 to 4'C. 5 Crude extracts of leaf, bundle sheath strands, embryo, endosperm, and BMS cells were prepared from frozen tissue as described by W.B. Parker et al., Pro.c Nat'l Acad. Sci. USA, 2, 7175-7179 (1990), except that extraction buffer contained 0.1 M Tricine-KOH, pH 8.3, 0.3 M glycerol, 5 mM DTT, 2mM Na 2 EDTA, and 0.5 mM phenyl methonyl sulfonyl fluoride (PMSF). Triton X 10 100 (0.01% v/v) was added to bundle sheath strand extracts and to some whole leaf extracts. For some experiments, additional protease inhibitors (leupeptin, 2 ptg mL; pepstatin A, 100 ptg mL 1 ; benzamidine, 1 mM; c-amino-n-caproic acid, 5 mM; and soybean trypsin inhibitor, 10 jig mL-') were included. Filtered homogenates were centrifuged 20 minutes at 30,000g. A portion of the crude 15 supernatant fraction was immediately boiled 5 minutes in 1 volume of SDS sample buffer (W.B. Parker et al., Plant Physiol., 92, 1220-1225 (1990)) for SDS-PAGE analysis; the remainder was desalted on a 10-mL Sephadex G-25 column into extraction buffer minus PMSF. ACCase was purified from the crude extract supernatant in four 20 steps. This fraction was brought to 30% saturation with solid (NH 4

)

2

SO

4 , stirred 15 minutes, and centrifuged 20 minute at 20,000g. The supernatant was then brought to 40% saturation with (NH 4

)

2

SO

4 solution, stirred 30 minutes, and centrifuged. The pellet was dissolved in 5 mL extraction buffer, microfuged 5 minutes, and the resulting supernatant was applied to a Sephacryl S-300 gel 25 filtration column (Pharmacia; 2.5 x 46 cm) equilibrated with S-300 buffer (0.1 M Tricine-KOH, pH 8.3, 0.5 M glycerol, 0.5 mM DTT, 2 mM Na50 mM KCI). In later experiments a Sephacryl S-400 column was used. Fractions (2.5 mL) were eluted at 0.75 mL mif'. ACCase activity eluted shortly after the void A 280 peak

(V

0 = 75 mL). Active fractions were pooled, brought to 4.25 mM MgCl 2 (from a 30 0.5 M solution), and applied at 0.2 mL min' to a Blue Sepharose CL-6B (Pharmacia; 1.5 x 15 cm) equilibrated with Blue sepharose buffer (S-300 buffer containing 4.25 mM MgCl 2 and 10 mM NaHCO 3 ). The column was washed overnight with 150 mL buffer (0.45 mL min'). ACCase activity was then eluted WO 99/67367 PCTIUS99/14022 43 with 50 mL buffer plus 10 mM ATP (0.45 mL min-'). Active fractions were pooled and applied to an FPLC Mono-Q HR 5/5 anion-exchange column (Pharmacia) equilibrated with S-300 buffer minus KCl. The column was washed with 30 mL S-300 buffer minus KCl and then with a 48-mL, 0 to 500 mM KCI 5 gradient in S-300 buffer (0.25 mL min-'). Fractions (1 mL) from the two peaks of ACCase activity were pooled separately. All purification fractions were desalted into S-300 buffer and assayed for ACCase activity and protein. ACCase was also analyzed from mesophyll chloroplasts and bundle sheath strands. Mesophyll chloroplasts from homogenates of 7- to 8-day 10 old seedlings that were kept in the dark 24 hours prior to harvesting were isolated on a linear Percoll gradient according to J.D. Burton et al., Pesticide Biochemistry and Physiology, 3_4, 76-85 (1989), except that buffers contained 0.6 M sorbitol and centrifugation g-forces were reduced by 25%. Intact chloroplasts were taken from the discrete lower green band present after Percoll 15 gradient centrifugation (G. Morioux et al., Plant Physiol., 6, 470-473 (1981)). Pelleted chloroplasts were lysed by resuspending them in ACCase extraction buffer plus PMSF and 0.01% (v/v) Triton X-100. Bundle sheath strands were obtained from the original leaf homogenate material retained on a 70-jm filter after re-homogenizing the retentate five times in a total of 2 L buffer. Triton 20 X-treated, desalted leaf, mesophyll chloroplast, and bundle sheath strand extracts were assayed for activities of Rubisco (G. Zhu et al., Plant Physiol., 91, 1348 1353 (1991)), NADP-dependent malate dehydrogenase (M.D. Hatch et al., Biochem. Biophys. Res. Commun., 3_4, 589-593 (1969)), phosphoenolpyruvate carboxylase (R.C. Leegood et al., "Isolation of Membranes and Organelles from 25 Plant Cells," Academic Press, New York, 185-210( 1983)), catalase (Worthington Biochemicals, 1972), and fumarase (R.L. Hill et al., Methods Enzymol., 13, 91-99 (1969)), and for total chlorophyll (D.E. Arnon, Plant Physiol., 2_4, 1-5 (1949)). Mesophyll chloroplast preparations were judged to be relatively free of contamination by bundle sheath chloroplasts because they 30 contained 3-fold greater NADP-dependent malate dehydrogenase and one-tenth as much Rubisco activity (mg' chlorophyll) than bundle sheath strand extracts. Mesophyll chloroplast preparations also contained <2.6% as much catalase, fumarase, and phosphoenolpyruvate carboxylase activities (mg 1 chlorophyll) as WO 99/67367 PCT/US99/14022 44 did whole-leaf extracts, indicating they were relatively free of peroxisomal, mitochondrial, or cytoplasmic components. ACCase activity as measured by acetyl-CoA-dependent H1 4 CO; (ICN, 2.07 GBq mmol-') incorporation into acid-stable product previously shown 5 to be malonyl-CoA (J.D. Burton et al., Pesticide Biochemistry and Physiology, 3_4, 76-85 (1989)). Assays of desalted purification fractions or crude, desalted tissue extracts contained up to 50 and 25% (v/v) enzyme, respectively. In some experiments methylcrotonyl-CoA or propionyl-CoA were substituted for acetyl CoA (E.S. Wurtele et al., Archives of Biochemistry and Biophysics, 22M, 10 179-186 (1990)). Avidin (10 U mL-1) was included in some assays. Herbicide inhibition assays contained 1% (v/v) ethanol plus or minus 1 pLM haloxyfop (2 [4-[[3-chloro-5-(trifluoromethyl)-2-pyridinyl]oxy]phenoxy] propanoic acid, Dow Chemical Co. analytical grade racemic mixture) or 10 pM sethoxydim (2 [1[(ethoxylmino)butyl]-5-[2-(ethylthio)-propyl]-3-hydroxy-2-cyclohexene -1 15 one, Li salt, BASF Corp. technical grade). Data are means plus standard error of three assays. Protein concentrations were determined in duplicate with the Bio Rad Coomassie blue dye-binding assay as described by the manufacturer, using BSA as the standard. 20 Centrifuged crude extracts and proteins in purification fractions and immunoprecipitation supernatants were separated by SDS-PAGE in 6 or 7.5% gels as previously described (W.B. Parker et al., Plant Physiol, 92, 1220-1225 (1990)). Purification fractions were precipitated in 10% (v/v) TCA, washed with 80% (v/v) acetone, and air-dried 10 minutes prior to 25 electrophoresis. Proteins in gels were stained with silver (J. Heukeshoven et al., Electrophoresis, fi, 103-112 (1985)). High molecular weight protein standards for SDS-PAGE (Pharmacia) were used to estimate polypeptide masses. The four-step purification procedure shown in Table IV typically yielded 30 to 190 pg of highly purified ACCase from 50 grams (fresh weight) of 30 maize inbred A619 or B73 seedling leaves. ACCase activity in the crude supernatant fraction precipitated between 30 and 40% saturation with (NH 4

)

2 S0 4 , which appeared to increase total ACCase activity approximately 38%. Crude extract components might have depressed the reaction rate shown in Table IV WO 99/67367 PCT/US99/14022 45 because the assay mixture contained 50% enzyme (v/v). In tests of fractions from another purification, enzyme velocity was proportional to enzyme concentration in assay mixtures containing up to 25% (v/v) crude extract, but 50% (v/v) mixtures were not tested. ACCase activity eluted from the Sephacryl 5 S-300 gel filtration column slightly after the green void peak. Approximately 56% of the S-300 fraction ACCase activity was recovered from the Blue Sepharose column, primarily in the initial ATP-containing fractions (12.5 mL). Both 10 mM NaHCO 3 and 4.25 mM MgCl 2 (1- and 0.85-fold standard assay concentrations, respectively) were included in the Blue Sepharose buffer because 10 they improved the total and specific ACCase activity remaining after batch absorption to Blue Sepharose beads, elution with ATP, and desalting into extraction buffer minus PMSF. Neither NaHCO 3 nor MgCl 2 improved enzyme stability of crude extracts. Mono-Q anion-exchange chromatography resulted in separation of two ACCase activity peaks which eluted at approximately 210 mM 15 (designated ACCase II) and 250 mM KCl (designated ACCase I), as previously observed for a hybrid maize variety (J.L. Howard et al., FEBS Lett., 261, 261-264 (1990)). ACCase I comprised about 85% of the total activity recovered from the column (29% of the original crude extract activity) and had high specific activity (Table IV). The specific activity of ACCase II was less than 20 30% that of ACCase I. Both activities were inhibited >90% by avidin, as previously reported (J.L. Howard et al., FEBS Lett., 261, 261-264 (1990)). The mass of native ACCase I was estimated to be approximately 490 kD by gel filtration on Superose 6.

WO 99/67367 PCT/US99/14022 46 TABLE IV Purification of ACCase I From Maize Inbred A619 Seedling Leavesa 5 All fractions were desalted into S-300 buffer and assayed for protein and acetyl CoA dependent incorporation of [ 1 4 C]HCO3- into acid-stable products. 10 Step Protein Activity Specific Fold Activity Activity Purification Yield mg units units/mg % Crude extract 215 2.45 0.0114 1 100 30-40% (NH 4

)

2

SO

4 45.1 3.37 0.0748 6.56 138 S-300 10.7 3.35 0.313 27.5 137 15 Blue Sepharose 1.50 1.86 1.24 109 76 Mono-Q (ACCase I) 0.130 0.720 5.54 486 29 a Data are from one purification experiment starting with 50 g fresh weight of tissue and are representative of data obtained for eight purifications. 20 b Unit = 1 ptmol acid-stable product min-'. B. Formation and Specificity of Antibodies to ACCase 25 Antibodies are sensitive reagents that allow for the identification of gene products from cDNA and other cloned genes. Antibodies to purified ACCase were prepared and used to screen for cDNA clones encoding all or a portion of a gene for ACCase. Antiserum to maize ACCase was obtained by immunizing a 30 female New Zealand White rabbit (Egli et al., Plant Physiol., 101, 499 (1993)). An intramuscular injection of 100 gg of Mono-Q-purified, SDS-denatured ACCase I in Freund's complete adjuvant was followed by subcutaneous injections of 20 to 100 gg of gel-purified ACCase I polypeptide in acrylamide plus incomplete adjuvant every 4 to 6 weeks, for a total of six injections. Serum 35 was stored at -20'C in 0.02% (w/v) NaN 3 . For Western blots, proteins in SDS gels were electrophoretically transferred to Immobilon (W.B. Parker et al., Plant Physiol., 92, 1220-1225 WO 99/67367 PCT/US99/14022 47 (1990)) for 1 hour at 20 V in a Bio-Rad Transphor semi-dry blotter and then stained with Ponceau S (E. Harlow et al., "Antibodies - A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York (1988)). Destained blots were blocked with Tris-buffered saline plus 0.5% (v/v) Tween 5 20 (Bio-Rad), and 10% (w/v) bovine serum (for antiserum blots only). ACCase and biotinylated proteins were detected with immune serum (1/10,000) plus goat anti-rabbit IgG-alkaline phosphatase conjugate or with avidin-alkaline phosphatase (W.B. Parker et al., Plant Physiol, 92, 1220-1225 (1990)). Blots were repeated at least three times. 10 For immunoprecipitations, equal ACCase activities (0.58 nmol min-') in crude extracts were desalted into S-300 buffer containing 0.1 M KC and incubated 1 hour at 25 C with 16 pL buffer or with 16 pL serum consisting of 0 to 100% ACCase antiserum in preimmune serum. Immune complexes were incubated 1 hour at 25'C with a 2-fold (IgG binding) excess of Protein 15 A-agarose and then microfuged 5 minutes to obtain immunoprecipitation supernatant fractions. ACCase activity of supernatants was expressed as a percent of the 100% preimmune serum control. Data are means plus SE of three replicate assays for each of two sets of extracts. Western blots and silver-stained gels of purification fractions 20 separated by 7.5% SDS-PAGE showed that neither ACCase I nor ACCase II Mono-Q fractions contained biotinylated polypeptides smaller than 212 kD. A polypeptide > 212 kD was the primary protein component of the ACCase I Mono-Q fragment (Fig. 5). The ACCase II fraction contained a biotinylated polypeptide > 212 kD and a large amount of a 55 kD non-biotinylated 25 polypeptide. Fractions from earlier purification steps contained additional biotinylated proteins of approximately 74, 75, and 125 kD (Fig. 5). To better compare the biotinylated polypeptides > 212 kD in ACCase fractions I and II, we used 6% SDS-PAGE, which showed that the mass of ACCase II was approximately 8 kD less than that of ACCase I. Molecular 30 masses were estimated to be 219 kD (ACCase II) and 227 kD (ACCase I), based on comparisons with polypeptide standards and the observation (N.R. Palosaari, Plant Physiol., 99(S), 359 (1992)) that, on Phastgels (Pharmacia), ACCase I polypeptide was slightly smaller than dodecameric horse spleen ferritin (238 kD; WO 99/67367 PCT/US99/14022 48 M. Heusterspreute et al., FEBS Tett., 129, 322-327 (1981)). All purification fractions through the Blue Sepharose step contained both ACCase I and II polypeptides. Rapid extraction of leaves in buffer containing five additional protease inhibitors, or a 4 hour incubation of extracts at 25 C, had little or no 5 effect on the relative amounts of the two polypeptides, suggesting that ACCase II is not a breakdown product of ACCase I. Antiserum to ACCase I strongly recognized the ACCase I polypeptide in crude extracts and showed little or no recognition of ACCase II polypeptides. No bands were recognized by preimmune serum. Assuming that 10 avidin binds similarly to ACCase I and II polypeptides, it appears that the amount of ACCase II on the Western blot was slightly less than the amount of ACCase I. However, the relative staining with antibody compared to avidin indicated that the antibody had significantly less affinity for ACCase II than ACCase I. 15 To determine whether the same ACCase polypeptides were expressed in different maize cell types, proteins in mesophyll chloroplasts and crude extracts of leaves, endosperm tissue, embryos, and BMS cells were separated by SDS-PAGE. All preparations contained a predominant biotinylated polypeptide of approximately 227 kD (ACCase I) that was strongly recognized 20 by ACCase antiserum or avidin. Similar 227 kD band densities were observed when gel lanes were probed with either avidin or ACCase antiserum. The 219 kD ACCase II polypeptide was readily detected in leaves only by avidin binding, but was in low abundance or not detected in extracts from other tissues. Only the 227 kD ACCase I polypeptide was detected in purified mesophyll 25 chloroplasts, however, suggesting that the 219 kD ACCase II polypeptide is localized elsewhere in mesophyll cells or in other cell types of young leaves. ACCase activity and a > 212 kD biotinylated polypeptide(s) were also found in bundle sheath strand extracts, but low yields prevented us from determining the type of ACCase present. Two other major biotinylated polypeptides of 75 and 30 74 kD were found in all tissues. Other non-biotinylated proteins of 66 kD (faint) and 55 kD were also recognized by ACCase antiserum. The 55 kD polypeptide was only found in leaves; it was also present in both ACCase I and II Mono Q WO 99/67367 PCT/US99/14022 49 fractions (Fig. 5) and was identified as the Rubisco large subunit based on its comigration with protein immunoprecipitated by spinach Rubisco antiserum. ACCase antiserum immunoprecipitated at least 75% of ACCase activity from crude, desalted extracts of leaves, endosperm tissue, embryos, and 5 BMS cells (Fig. 6), indicating that most of the ACCase activity in these tissues is immunologically related to the ACCase I polypeptide of leaves. Less activity was precipitated from leaves (75%) than from other tissues, particularly embryos (98%). Compared to immunoprecipitation, inhibition of ACCase activity by antiserum in solution was less than 20% as effective in reducing ACCase 10 activity. The substrate specificity of ACCase from different purification fractions was examined to compare [ 1 4 C]HCOj incorporation in the presence of different acyl-CoA substrates. Both ACCase I and II utilized propionyl-Co-A 40 to 50% as rapidly as acetyl-CoA at 50 to 500 piM substrate even though they 15 contained no biotinylated polypeptides (Figure 5) the size of known propionyl CoA carboxylases (70 to 75 kD; see E.S. Wurtele et al., Archives of Biochemistry and Biophysics, 228, 179-186 (1990)). Activities in the presence of both acetyl-CoA and propionyl-CoA (250 or 500 ptM each) were approximately 90 (ACCase I) to 130% (ACCase II) that of 500 ptM acetyl-CoA 20 alone. Crude leaf extracts utilized propionyl-CoA and methylcrotonyl-CoA 60% as efficiently as acetyl-CoA. Methylcrotonyl CoA carboxylase activity was reduced 85% by gel filtration and was completely removed by Blue Sepharose affinity chromatography. ACCase I and II differed significantly in their inhibition by either 25 haloxyfop or sethoxydim (Fig. 7). Acetyl-CoA or propionyl-CoA-dependent H1 4

CO

3 incorporation by ACCase I was strongly inhibited (65 to 80%) by 1 pM haloxyfop or 10 gM sethoxydim, while ACCase II activity was inhibited less than 50% for all herbicide/substrate combinations examined. C. Cloning and Identification of Maize cDNA Clones Encoding ACCase 30 Maize cDNA clones encoding a portion of the ACCase game were identified by screening a DNA library generated from maize. The cDNA clones were used to identify the sequence of the ACCase gene and to identify the genomic DNA fragments encoding the gene or genes for ACCase.

WO 99/67367 PCT/US99/14022 50 A X gt 11 cDNA library from maize inbred A 188 seedlings was prepared by standard method for oligo-dT priming, as described for pea cDNA. (Gantt and Key, Eur. J. Biochem., 16fi:119-125 (1987). Plaque lifts of the maize cDNA library were screened with maize ACCase antiserum (Egli et al., Plant 5 Physiol., U1, 499 (1993)) to identify plaques expressing ACCase-like proteins, as described by Sambrook et al., cited supra. (1989). The initial screen of 800,000 plaques yielded 120 positives. Rescreening and plaque purification reduced the number of positives to 14. All 14 clones bound ACCase antibodies that, when eluted from plaque lifts (J. Hammarback et al., J. Biol. Chem., 10 265:12763 (1990)), recognized a 227-kD biotinylated polypeptide on SDS PAGE western blots of embryo and leaf crude extracts. The strongest western blot reaction was obtained with cDNA clone #15-14. The six best clones were digested with EcoRI to excise maize cDNA inserts. Total insert sizes ranged from 1.2 to 5.1 kb indicating the clones most likely did not contain the full 15 coding sequences for the mature 219-kD and 227-kD ACCase polypeptides (minimum estimates of 6.1 and 6.3 kb, respectively). Clone #15-14 contained three EcoRI fragments of 2.0, 1.2 and 0.23 kb shown in Figure 8. Southern blots showed that the 1.2 and 2.0-kb fragments of clone #15-14 each hybridized to different fragments in the other 20 five clones, with the exception of clone #4-4 which only contained a 1.2-kb fragment. The six maize cDNA clones contained EcoRI fragments that hybridized to a large transcript (ca. 7.8 kb) on Northern blots of total RNA from maize leaves, embryos and endosperm (Fig.9). BMS cell culture RNA also contained a 7.8 kb transcript, but the hybridization signal is not evident on this 25 exposure (Fig. 9). The relative abundance of the 7.8-kb transcript in embryos was higher than the other sources which is consistent with their ACCase activity. The three EcoRI fragments were subcloned from cDNA clone #15-14 into BlueScript vector and sequenced by the dideoxy chain termination method (Sequenase 2.0 USB) initially using T3 and T7 primers and then 30 oligonucleotide primers based on insert sequence. A clone #16-6 was also sequenced in a similar manner. Clone #16-6 included three EcoRI fragments of 3.1 kb, 1.2 kb, and 0.23 kb and had additional sequence located upstream from that of clone #15-14. After comparing the sequence and determining that the WO 99/67367 PCT/US99/14022 51 sequence was the same, the additional 1.2 kb sequence at the 5' end was sequenced. Clone #18-5 was sequenced in a similar manner. Clone #18-5 included 3.9 kb, 1.2 kb, and 0.23 kb EcoRI fragments and contains an additional 5 1.9 kb 5' sequence upstream from clone #15-14. Subelone #18-51 (3.9 kb EcoRI fragment) has been deposited with the American Type Culture Collection and given Accession No. 69236. GenBank, PIR-29, and Swiss-Prot 19 data banks have been searched for amino acid homology with the corresponding amino acid sequences 10 of the three subclones of clone #18-5. Peptide sequences corresponding to the maize cDNA subclones had higher similarity to chicken, rat, and yeast ACCases than to any other peptide sequence in the data banks. Figure 8 illustrates the relative organization of the 3.9, 1.2 and 0.23-kb EcoRI fragments of clone #18-5 and their co-linearity and extent of amino acid identity with chicken ACCase 15 cDNA sequence. This comparison shows that the maize clone #18-5 has a large region near the 3' end with high amino acid identity (40 to 61%) to chicken ACCase, a longer region with 23% identity in the middle of the 3.9-kb sequence, and a short region with 52% identify near the 5' of the 3.9 kb sequence. Portions of the sequence of the #18-5I subclone have been 20 identified as encoding domains of the ACCase enzyme of functional significance. Those functional regions include a fragment that spans the presumed transcarboxylase active site in the enzyme having the following presumed sequence SEQ ID NO: 2: 25 1112-856 GTT CCT GCA AAC ATT GGT GGA CCT CTT CCT ATT ACC AAA CCT CTG GAC CCT CCA GAC AGA CCT GTT GCT TAC ATC CCT GAG AAC ACA TGC GAT CCA CGT GCA GCT ATC TGT GGT GTA GAT GAC AGC CAA GGG AAA TGG TTG GGT GGT ATG TTT GAC AAA GAC AGC TTT GTG GAG ACA TTT GAA GGA TGG GCA 30 AAA ACA GTG GTT ACT GGC AGA GCA AAG CTT GGA GGA ATT CCT GTG GGC GTC ATA GCT GTG GAG ACA WO 99/67367 PCTIUS99/14022 52 This functional domain is contained in the sequence 1112 to 856 base pair from the 3' stop codon or carboxy terminus region of the ACCase coding sequence of maize. This transcarboxylase active sequence is also present in clone #15-14. Another functional region that has been identified spans the 5 12 base pair sequence encoding the biotin binding site having the following peptide sequence SEQ ID NO: 3: 5' GTT ATG AAG ATG 3' Val Met Lys Met 10 The biotin binding site is encoded approximately 30% in from the 5' (N terminus) end of rat, chicken and yeast ACCase genes. These functional domains are useful in mapping and further identifying other cDNA and/or genomic fragments encoding ACCase genes. 15 The cDNA clones encoding portions of the acetyl CoA carboxylase genes are useful to identify the sequence of the gene or genes and are useful as probes to locate the genomic copies of the gene or genes. Because the ACCase antibodies used to screen the X gtl 1 library recognize both the 219 and 227 kD ACCase polypeptides, it has not been determined which polypeptide 20 is encoded by these less than full length clones. It is likely that the majority of the clones encode the 227 kD polypeptide since that polypeptide is more abundant in the leaf tissue source of the DNA library and the antibodies have a higher affinity for the 227 kD ACCase polypeptide. EXAMPLE VI 25 Isolation and Sequencing of Genomic Encoded ACCase Genes and a Complete cDNA Sequence of a Maize ACCase Gene The maize genome has been analyzed to identify copy number and location of the genomic copies of ACCase gene or genes. Four distinct types of maize ACCase genomic clones have been identified, termed A1, A2, B1 30 and B2 (see below). To obtain genomic copies of ACCase genes, a maize B73 genomic library (Clontech, Palo Alto, CA) was screened with the 2 kb subclone from #15-14 and several clones of about 15 kb were identified as having homology to the ACCase cDNA. Restriction mapping and partial sequence 35 analysis revealed two types of genomic clones (Type A and Type B) that differed WO 99/67367 PCT/US99/14022 53 in restriction sites and in their position relative to the ACCase partial cDNA sequence as shown in Figure 8. The 2.5 kb EcoRI-SaII fragment (#16) from the Type A genomic clone and the 3.0 kb EcoRI-EcoRI fragment (#34) from the Type B genomic 5 clone were shown to hybridize to the 3.9 kb probe from #18-5 and were subcloned into the Bluescript vector and sequenced. Approximately 1.5 kb of DNA sequence from the genomic Type A 2.5 kb fragment were 100% identical to coding sequence from the 3.9 kb cDNA subclone #18-51 described in Example V; the remaining sequence exhibited no identity with the cDNA clone and 10 presumably represents a noncoding intron sequence. A 350 nucleotide sequence derived from the genomic Type B 3.0 kb fragment was about 95% identical to the cDNA clone indicating that its coding sequence differs from that of genomic Type A. These results also indicate that the maize genome encodes at least two different genes encoding a polypeptide having acetyl CoA carboxylase activity. 15 To identify and clone the remainder of the gene representing the amino-terminus of maize ACCase, additional regions from the Type A genomic clone have been subcloned and partly sequenced. The 3.5 kb HindIII-HindlIl fragment (#28) has been sequenced for about 400 nucleotides from each end. The 3' end of #28 shows significant homology to the amino acid sequence from 20 the chicken sequence located about 0.5 kb from the start of the chicken gene. The complete sequence for fragment #28 can be obtained and analyzed to determine whether it contains the 5' end of the ACCase coding region. The start of the transcribed region, and thus the likely start of the coding region for ACCase, can be identified by using the genomic clones in RNAse 25 protection analysis (J. Sambrook et al., "Molecular Cloning - A Laboratory Manual," Cold Spring Harbor Press, Cold Spring Harbor, New York (1989)). Based on sequence data from the genomic clone, alignment, as shown in Figure 8, with sequences of other ACCases and identification of potential open reading frames, oligonucleotide primers can be constructed to synthesize cDNA 30 molecules representing the amino terminus of the ACCase gene. These molecules can be hybridized to genomic Type A DNA fragments such as #28 and the nonhybridizing portions digested with Sl nuclease. The end of the protected fragment are determined by analysis on a DNA sequencing gel.

WO 99/67367 PCTIUS99/14022 54 To synthesize the remaining coding region between the end of the cDNA clone #18-5 and the start of transcription, two oligonucleotide primers were synthesized. Primer 1 is complementary to the DNA sequence: (SEQ ID NO:4) 5 5' GCCAGATTCC ACCAAAGCAT ATATCC 3' near the 5' end of cDNA subclone #18-5I and can be used as a primer for synthesis of cDNA molecules from maize seedling, leaf or embryo RNA. A primer corresponding to a DNA sequence near the transcription 10 start site can be used in combination with primer 1 for the amplification of DNA by the polymerase chain reaction (PCR). Several independent clones are then sequenced and their sequences compared to the known sequence of the Type A genomic clone to determine the exact coding sequence corresponding to that maize gene for ACCase. A similar strategy can be used to obtain the complete 15 coding sequence for genomic Type B ACCase. The remaining cDNA sequence was obtained by three successive rounds of RT-PCR using oligonucleotide primers based on genomic apparent exon (5') and known cDNA (3') sequences. The primers used to amplify nucleotides 1-240 of the cDNA were 28sst-a5+ (SEQ ID NO. 7) and 20 28sst-6at3+ (SEQ ID NO:8), nucleotides 217-610 of the cDNA were 28sst-5+ (SEQ ID NO:9) and 28-2t3+ (SEQ ID NO:10), and nucleotides 537-2094 of the cDNA were ACCPCR5' (SEQ ID NO:1 1) and 155- (SEQ ID NO:30) (Table V). PCR products corresponding to nucleotides 1-240, 217-610, and 537-2094 of the final sequence were cloned into PCR-script (Stratagene).

WO 99/67367 PCTIUS99/14022 55 00 0 0 u z H H H H Hn H 00 00 C> 0C C)1r Hq 0 f WO 99/67367 PCT/US99/14022 56 The original 5.4-kb cDNA clone #18-5 and PCR products from at least three individual PCR per oligonucleotide pair were sequenced in both directions by the dideoxy chain-termination method, using either Sequenase II (U.S. Biochemicals) or ABI 373 (Applied Biosystems, Inc.) protocols. No 5 sequence differences were found in regions of clone overlaps. The complete sequence of the cDNA of maize ACCase (nucleotides 1-7470 SEQ ID NO:5) and its corresponding amino acid sequence (amino acids 1-2325 SEQ ID NO:6) are shown in Figures 13 and 14. The 7470 bp cDNA includes a 459 nucleotide 3' untranslated region and 36 nucleotides of 5' untranslated sequences. 10 The first Met codon in the cDNA (nucleotides 37-39) was identified as the start codon based on its similarity to consensus initiation sequences (Kozak, J. Cell. Biol., 10, 229 (1989); Lutcke et al., Embo. J., fi, 43 (1987)). An in-frame stop was found in the genomic sequence 6 nucleotides upstream of the sequenced cDNA, and RT-PCR analysis of this region suggested 15 that the in frame stop codon was also present in the cDNA. The 3' end of the coding sequence was defined by two stop codons present in the large open reading frame after nucleotide 7011. The translated coding sequence predicted a polypeptide of 2325 amino acids (257 kD; SEQ ID NO:6) which was 79 to 81% identical to the multifunctional (MF) ACCases from alfalfa (Shorrosh et al., 20 Proc. Nat'l. Acad. Sci., 91, 4323 (1994)) and wheat (Gornicki et al., Proc. Nat'l Acad. Sci., 91, 6860 (1994)), and to a 118-amino acid predicted polypeptide of a rice expressed sequence tag (Genbank accession # D39099, T. Sasaki), but only 53 to 55% identical to ACCase from other eukaryotes. In a pileup alignment of plant ACCases (Genetics Computer 25 Group, Madison, Wisconsin), Met 1 of both maize and Brassica napus ACCases was located about 130 amino acids upstream of the conserved sequence VDEFCKALGG, compared to only 25 amino acids upstream for other plant ACCases. The predicted 2325 amino acids of maize ACCase contains a biotinylation site at position 806, within the conserved MKM motif (Ton et al., 30 Eur. J. Biochem., 215, 687 (1993)). The arrangement and amino acid sequence of binding sites (Shorrosh et al., Proc. Nat'l. Acad. Sci., 91, 4323 (1994)) for ATP (amino acids 318-333), biotin (amino acids 799-811; biotin at 806), WO 99/67367 PCTIUS99/14022 57 acetyl-CoA (amino acids 1952-1961), and carboxybiotin (amino acids 1662-1711) were highly conserved among all MF ACCases. EXAMPLE VII Characterization of other Genomic Clones 5 The initial restriction fragment length polymorphism (RFLP) analysis of EcoRI-digested total DNA from three maize inbred lines showed one band when probed with the 2 kb subclone from #15-14 (internal to gene) and two bands when probed with the 1.2 kb subclone (near the 3' end of the gene). Fragments homologous to the 2 kb probe were monomorphic and the more 10 intense of the two bands hybridizing with the 1.2 kb probe was dimorphic. As discussed in Example V, these results support the view that maize contains at least two distinguishable ACCase genes and that they may be quite similar for much of the coding region. Additional genomic Southern blots of a set of recombinant inbred lines were used to map polymorphisms for the ACCase 15 probes to maize chromosomes. One polymorphism was mapped to the short arm of chromosome 2; other polymorphisms were not evident in these initial tests to identify a chromosomal location for other maize ACCase genes. The isolation and restriction mapping of additional genomic clones from a B73 genomic library (Clontech) resulted in the identification of 20 four different types of clones termed Al, A2, BI and B2 (Figures 16-19) which had 96% nucleotide sequence identity. Types A and B correspond to previously published pA3 and pA4 cDNAs (Ashton et al., Plant Mol. Biol., 24, 35 (1994)) and differ from pA3 and pA4 by ~ 4% in their coding sequences. Type A and B genomic clones have linear sequence homology 25 except for an insertion in an intron of the Type B genes about 1400 bp 3' of the Al(SEQ ID NO:5) translation start site. Analysis of the insert boundaries revealed a 3-bp target site duplication and a 6-bp direct repeat, and further sequence analysis showed the presence of two new and unique LINE elements (Long Interspersed Nuclear Elements) in B1 and B2. Mammalian LINE 30 elements are highly abundant (104 to 105 copies), 6 to 7 kb long, and have frequent 5'-end deletions and an A-rich 3' terminus. They are flanked by short direct repeats, and contain two ORFs, one encoding a reverse transcriptase. Three LINE elements (Cin4, 50-100 copies in maize; del2, 250,000 copies in WO 99/67367 PCT/US99/14022 58 lily; BNR1, 2-5% of genome in sugarbeet) have been described in plants (Leeton et al., Mol. Gen. Geneti., 232, 97 (1993); Schmidt et al., Chromo. Res., 3., 335 (1995); Schwarz-Sommer et al., EMBO J., i, 3873 (1987)). Maize ACCase B1 has one unique LINE element and B2 has two. The two B2 LINE elements were 5 characterized by differences in their reverse transcriptase sequence. The B genomic clone inserts have characteristic LINE features including cysteine motifs and a possible polyA tail, and high abundance. The LINE insert also has been found in an intron of the maize Shrunken-2 gene (Hannah et al., Plant Physiol., 9, 1214 (1992)). 10 The partial nucleotide sequence (3489 nucleotide) of a Type Al ACCase genomic clone is shown in Figure 16 (SEQ ID NO:12). The clone is a HindII fragment which includes nucleotides 1-931 of the cDNA in Figure 13 (SEQ ID NO:5), and the first four introns within the coding region, at positions 240 (460 nucleotides), 296 (480 nucleotides), and 872 (76 nucleotides) of SEQ 15 ID NO:5. The clone also has 1395 nucleotides 5' to the cDNA of SEQ ID NO:5 (i.e., 1431 nucleotides 5' of the translational start at nucleotide position 1432). The partial nucleotide sequence (1328 nucleotide) of another Type A clone is shown in Figure 17 (SEQ ID NO: 13). The partial sequence is all 5' untranslated sequence and contains a 7 base insert between nucleotides 20 279-290, but is otherwise identical to SEQ ID NO:12. The partial nucleotide sequence of six Type A2 clones is shown in Figure 18 (1565, 1168, 638, 558, 976 and 852 nucleotides; SEQ ID NOs 14, 15, 16, 17, 18 and 19, respectively). Within the Al -A2 clone pair, identified differences are in introns 25 and 5' UTR sequences. The A2 genomic clone is weakly amplified with Type Al PCR primers specific for the 5' UTR if the 3' primer employed is for a conserved amino acid sequence found in all ACCases (e.g., 28sst-1 OF, ACTGTGCGTTTGAGAAGGTC, SEQ ID NO:28, and 28sst-2T3+, CCTCTACGTAATTGGTCAGC, SEQ ID NO:29). The A2 amplified product is 30 the same size as that from the Type Al genomic clone, and restriction analysis indicated a difference in sequence from Al. Sequence differences in the 5' region should provide a means to distinguish between expression of Al and A2 ACCase genes and to determine whether A2 also encodes a CTP.

WO 99/67367 PCT/US99/14022 59 The partial nucleotide sequence (231, 207 and 180 nucleotides) of three Type B clones is shown in Figure 19 (SEQ ID NOs 20, 21 and 22, respectively). The cDNAs corresponding to genomic clones A2, B1 and B2 are 5 cloned and sequenced in a manner similar to that described above. The derived amino acid sequences are aligned with known ACCase sequences. If putative CTP sequences are identified, functionality is tested as described below. Also if the tissue specificity and developmental timing of expression differ for different ACCase genes, the sequences of the promoter regions of the corresponding 10 genomic clones are compared. Gene-specific probes for specific ACCase genes can provide more information on their roles in lipid synthesis (plastid and cytoplasmic isoforms), secondary metabolism (cytoplasmic isoforms), and herbicide resistance (likely plastid isoforms). A 3' Type Al ACCase cDNA probe mapped to chromosome 2S 15 (Egli et al., Maize Genetics Newsletter, 6, 92 (1994)) and to 1OL (Caffrey et al., Maize Gen. Coop., f2, 3 (1995)). Two 5' Type Al cDNA probes which span the transit peptide mapped to chromosome 2S in the same location as the 3' probe (see maize genetic map, 1996 version, Maize Genomic Database). PCR primers 28sst-97F (CCTTTTTATGGCACTGTGCG, SEQ ID NO:25) and 28sst-6t3+; 20 (CATCGTAGCCTATATGAGGACG, SEQ ID NO:26) located in non-coding regions of Al that span the chloroplast transit peptide were used to amplify a B73 chromosome-specific product which segregated with the resistance trait. A nearby 5' primer (28sst-a5+, SEQ ID NO:7) amplified all genotypes and functioned as a positive control. Herbicide resistance due to the Accl-S3 25 mutation segregates (29/29 individuals to date) with production of a Type A 5' end-specific PCR product derived from the mutant parent while herbicide sensitive plants lack the transit sequence (15/17 progeny). Two individual plants which contained B73-specific DNA at this location died of unknown causes while grown in the presence of herbicide. 30 Mutations in maize that confer resistance to cyclohexanedione and aryloxyphenoxypropionate herbicides by means of an altered ACCase target are found at two non-allelic loci, Acc1 and Acc2. Al and A2 appear to encode plastidic ACCases and correspond to the Acc1 and Acc2 herbicide resistance WO 99/67367 PCT/US99/14022 60 loci. Acc2 has been mapped to 10 L (VanDee, M.S. Thesis, University of Minnesota (1994)). Acc1 is the site of five allelic mutations including Accl-S2 and -S3 (Marshall et al., Theor. Apple. Genet., 83, 435 (1992)), and has been mapped to chromosome 2. 5 Only one plastidic ACCase polypeptide was identified by SDS PAGE of maize leaf extracts, although 2-D gel analyses might provide evidence for a second, highly similar isoform. Of the two ACCase isoforms, only ACCase I shows altered herbicide inhibition in Accl-S2 mutants, and most of the ACCase activity in leaves and developing embryos is herbicide-resistant and 10 thus attributed to the Accl-S2 gene product. Although a 3' ACCase probe has been mapped both to 1 OL near Acc2-S5 and to 2S, the conserved sequence of ACCase genes and lack of polymorphism in multiple bands complicates identification of genes encoded at these loci. The Type Al ACCase gene is probably located on chromosome 2, 15 since (i) 5' untranslated and chloroplast transit peptide probes from Type Al hybridize to two bands (dark and light) in maize inbreds, and (ii) analysis of maize-oat addition lines carrying maize chromosomes 2 through 9 indicates the dark band is on chromosome 2 and the light band is on chromosome 1 or 10. Type B ACCase genes are likely to encode cytosolic isoforms. 20 Given that cytosolic malonyl-CoA is a precursor in the synthesis of many secondary metabolites including flavonoids (e.g. maysin, a corn silk component associated with corn earworm resistance), these cytosolic ACCases can have agronomic utility. Northern blot analysis of total maize RNA with an ACCase probe 25 (nucleotides 3400-5932) showed a single 8.3 kilobase band. To determine whether the expression of ACCase RNAs was developmentally regulated, blots of total RNA from 16 to 42 DAP (days after pollination) embryos were probed with an ACCase cDNA fragment. Transcript abundance peaked about 23 DAP and the steady state pattern was similar to in vitro ACCase enzyme activities and 30 protein measured from developing embryos. Type A- and B- specific 32

P-CTP

labeled antisense transcripts were 780 nt long (662 nt of ACCase sequence + 118 nt of vector/promoter sequence) and were identical except for 15 base mismatches scattered along their length. Each antisense transcript was WO 99/67367 PCTIUS99/14022 61 hybridized to total RNA from embryos at 16, 20, 23, and 42 DAP and digested with RNAse A/TI mixture to yield a 662-base fragment specific to the probe used. The results showed that the Type A transcript was more abundant than Type B at all tested stages, and that only Type A remained high in older 5 embryos. Types A and B had similar expression patterns and peaked around 20-23 DAP. The ratio of Type A:B mRNA in leaves was about 2:1, similar to its relative abundance in cDNA expression libraries. EXAMPLE VIII Expression of the Maize ACCase Chloroplast Transit Peptide 10 The N-terminus of the predicted maize ACCase polypeptide is longer than that of predicted cytosolic ACCase isoforms and has several characteristics typical of chloroplast transit peptides within the first approximately 73 amino acids of the predicted N-terminal sequence. The CTP cleavage site motif is not found in the putative maize ACCase CTP, although 15 only about 30% of known CTPs contain this consensus sequence (Gavel and von Hejne, FEBS Lett., 2f1, 455 (1990)). However, the maize ACCase N-terminus appears to have several other properties typical of known CTPs: (1) a lack of acidic residues in amino acids 1-49, (2) high Ser + Thr content (69% within amino acid residues 23-35), (3) an R-rich region between S- and D-rich regions 20 in amino acid residues 36-49, and (4) a predicted turn - P sheet within amino acid residues 58-73 (von Hejne and Nishikawa, FEBS Lett., 22, 1 (1991)). The ability of the amino acid sequence contained within the N-terminal 100 amino acids of the translated maize acetyl-CoA carboxylase (ACCase) cDNA to direct the N-terminal portion of the maize ACCase biotin 25 carboxylase domain into chloroplasts was tested in vitro by methods used extensively in the literature (see Cline et al., J. Biol. Chem., 260, 3691 (1985); Lubben and Keegstra, Proc. Nat'l. Acad. Sci., 83., 5502 (1986)). The criteria for import was that (1) in vitro-synthesized, 1 5 S-labeled protein was imported into chloroplasts, and (2) the transported protein was smaller than the original 30 translation product, by an amount which corresponds to the removal of the expected CTP. Import studies utilized either maize or pea chloroplasts. Pea chloroplasts have been reported to correctly import proteins from many different species, including maize (Nieto-Sotelo et al., Plant Physiol., 93, 1321 (1990)).

WO 99/67367 PCT/US99/14022 62 Alternatively, the function of the putative maize ACCase CTP is tested by inserting the first 258 coding nucleotides of maize ACCase in frame with and 5' of a GUS reporter gene in pBI221 (Clontech). This construct and the pBAR plasmid are used to co-transform maize "Black Mexican Sweet" suspension cells 5 by particle bombardment. Basta-resistant transformants are selected, and GUS activity and/or protein is assayed in surviving cultures or in plasmids isolated from transformants. A partial ACCase construct consisting of nucleotides 1-833 of SEQ ID NO: 5 including the putative CTP (nucleotides 37 to 256) and the first 10 domain within the biotin carboxylase region (identified by amino acid sequence comparison with E. ccli biotin carboxylase; see Waldrop et al., Biochem., 33, 6249 (1994)) was amplified by PCR and cloned into the EcoRV site of PCR script (Stratagene) to create the plasmid pBCN1. A corresponding plasmid lacking CTP sequences (nucleotides 278-833) was also made (-pBCN1). The 15 protein encoded by -pBCN1 begins at amino acid residue 83 (Val - Met). Constructs were transformed into E. ali SURE cells (Stratagene). Restriction analysis of pBCN1 with BamH1 and HindIII indicated that the 5' end of the ACCase was located adjacent to the T7 RNA polymerase binding site in PCR-Script. A partial sequence of pBCN1 obtained by using the T7 sequencing 20 primer and the AB1373 automated sequencing protocol confirmed this orientation and showed that the pBCN1 insert sequence was identical to maize ACCase cDNA for at least the first 300 nucleotides and that it included the maize ACCase Met 1 ATG. An acyl carrier protein clone containing a CTP (spinach ACPII, a gift of Dr. John Ohlrogge, Michigan State University) can be 25 used as a positive control. These constructs are used for in vitro transcription, translation, chloroplast import, and SDS-PAGE analysis of products in the same manner as pBCN1. Purified pBCN1 was digested with EcoRI to linearize the plasmid at the 3' end of the BCN1 insert, electrophoresed in 1.5% agarose, and the 30 plasmid band at approximately 3.8 kb was excised and Gene-Cleaned (BioLab 101). The purified band was digested with 20 ptg proteinase K to remove any residual RNAse, extracted with phenol and then chloroform under RNAse-free conditions. DNA content was estimated by ethidium bromide fluorescence in WO 99/67367 PCT/US99/14022 63 droplets, relative to X DNA standards (Sambrook et al., Molecular Cloning: A Laboratory Manual, 2nd ed. (1989). One tg of pBCN1 DNA was transcribed into capped RNA with the T7/mMessage mMachine kit (Ambion). Uncapped transcripts (Sp6 RNA polymerase; Promega) of pea RUBISCO small subunit 5 (SSU; Anderson et al., Biochem. J., 240, 709 (1986)) were also transcribed. RNA yield was estimated by determining the % incorporation of a 32 P-ATP (Amersham) into a precipitable product, according to the Ambion kit instructions. Electrophoresis and autoradiography of 32 P-labeled product showed that it contained a single RNA band of approximately 895 nucleotides, as 10 expected. The RNA transcripts were translated into 1 5 S-labeled polypeptides with Ambion's wheat germ IVT kit and approximately 45 gCi 3 1S-methionine (Amersham; 37 TBq/mmol) in a 50-pl reaction. Labeled proteins were held on ice 6 hours prior to their use in chloroplast import experiments. 15 Pea (cv. "Little Marvel") and maize (inbred A188) plants were grown in a growth chamber at 25'C, 16 hour day length. Chloroplasts were isolated from pea and maize leaves 7 days after planting, respectively, as previously described (Burton et al., Pestic Biochem. and Physiol., 3_4, 76 (1989); Egli et al., Plant Physiol., 101, 499 (1993)). Intact mesophyll chloroplasts were 20 washed in resuspension buffer [50 mM HEPES-KOH, pH 7.8 plus 0.33 M (pea) or 0.66M (maize) sorbitol] in preparation for import assays. Suspensions were diluted to obtain 75 pg chlorophyll/0.3 ml (Arnon, Plant PhysioL, 24, 1 (1949)). Import experiments were carried out essentially as described by Cline et al.(J. Biol. Chem., 260, 3691 (1985)). Import reactions containing 0.3 25 ml pea or maize chloroplast suspension, 40 pil 3 1S-translation mixture, 3 mM Mg-ATP and 10 mM Met were incubated under light for 1-30 minutes at 25 C. Unimported proteins were digested for 30 minutes with 40 pg of thermolysin, and proteolysis was stopped with 10 mM EDTA. Chloroplasts were re-isolated by centrifuging them through 1-ml 30 40% v/v Percoll gradients in the presence of resuspension buffer plus 3 mM Mg ATP, 10 mM Met, and 20 mM EDTA, washed twice in the same buffer, and resuspended in 65 p1 of 1 mM MgCl 2 /10 mM Tris buffer, pH 8.0. Chloroplasts were lysed by three cycles of freeze-thawing in liquid N 2 , microfuged, and WO 99/67367 PCT/US99/14022 64 aliquots of the supernatants and of the original in vitro-translated proteins were analyzed by SDS-PAGE in 8-25% gradient Phast gels (Pharmacia), followed by direct detection of radiolabeled proteins in the wet gels (AMBIS) (Figure 20). As estimated by SDS-PAGE, a 30 minute import converted the 5 original 32-kD BCN1 polypeptide to a doublet of 27.2 and 27.5 kD in maize and produced an additional 30-kD band in pea (Figure 20A). Neither maize or pea chloroplasts imported -pBCN1-derived polypeptides. Formation of the 27.2-kD polypeptide likely resulted from cleavage after amino acid #47-49, a likely cleavage site because it lies between S- and D-rich regions, and R residues are 10 located at -2, -7, and -8 (S. Gavel and G. Von Heijne, FEBS Lett., 2i, 455 (1990)). Time-dependence of BCN1 import was further examined (1 30 minutes) (Figure 20B) to determine if any imported polypeptides were a result of incomplete processing or proteolysis. Import was maximal after 15 15 minutes, but import time had no effect on the relative amounts of different-sized import products. Higher amounts of ATP (5 mM) stimulated import relative to lower amounts of ATP (< 0.2 mM). The data suggest that, in maize, efficient cleavage of BCN1 occurs at two closely adjacent sites and that partially processed products are also formed during BCN1 import by pea chloroplasts. 20 Therefore, nucleotides 1-833 of the maize ACCase gene encode a CTP. EXAMPLE IX Expression of a cDNA Clone or Genomic Clones Encoding the ACCase Gene The cDNA and genomic clones encoding all or a portion of the 25 ACCase gene can be subcloned into a known expression system and the gene products reactive with the antibodies specific for maize ACCase can be identified using a Western blot. For example, the ACCase cDNA clones are inserted into two transformation plasmids: (i) Glblexp which contains the embryo-specific maize Globulin] (Glb1) promoter and 3' regions (Belanger et 30 al., Genetics, 129, 863 (1991)); and (ii) pAHC17 which contains the maize ubiquitin (Ubi-1) constitutive promoter and first exon and intron, and the NOS 3' terminator (Christensen et al., Plant Mol. Biol., 1h, 675 (1992); Toki et al., Plant Physiol, 1f_, 1503 (1992)). The 3' end of the Al cDNA has a unique SalI site WO 99/67367 PCT/US99/14022 65 just 3' of the stop codon which is used to ligate into a SalI site in both plasmids ahead of the construct terminator. Other cloning sites will be added as needed to the plasmids or cDNA to complete the ligation of the 5' end. The gene products can also be further characterized structurally and/or enzymatically. This will 5 ensure that the genomic and cDNA clones that encode acetyl CoA carboxylase can be screened for promoters that provide for overproduction of the native or herbicide tolerant ACCase enzyme in plants. For example, the 2 kb EcoRI fragment from clone #15-14 can be subcloned into a plant transformation plasmid pBl21 or pBI221 downstream 10 from the 35S CaMV promoter and upstream from the nopaline 3' polyadenylation signal sequence, as described in Jefferson, Plant Molec. Biol. RepIr., 5, 387-405 (1987). This plasmid can then be used to transform plant cells such as tobacco, Brassica and Arabidopsis cells using protoplast or biolistic transformation, as described by W.J. Gordon-Kamm et al., Plant Cell, 2, 603-618 15 (1990); M.E. Fromm et al., Bio/TechnolQgy, &, 833-839 (1990); An, Methods in Enzymology, 153, 292 (1987); and D'Hafluin, The Plant Cell, 4, 1495 (1992). An increase in transient expression can be detected using quantitative Western blotting with antibodies specific for the ACCase enzymes. Polyclonal antibodies to maize ACCase most likely do not substantially crossreact with ACCase from 20 dicots like tobacco or Arabidopsis. Alternatively, the ACCase gene can be subcloned along with the 35S Came promoter into a binary Ti vector pGA482, as described in An, cited supra., which is a binary Ti vector system and can be used to transform plant cells by Agrobacterium-mediated transformation. Stably transformed plants can 25 be generated by standard methods as described in Example III, and levels of expression of ACCase genes can be determined by quantitative Western blots, as described in Harlow and Lane, Antibodies, Cold Spring Harbor Laboratories (1988). The ability to monitor expression of cloned ACCase genes will permit the identification of promoters that provide for enhanced expression of the 30 ACCase gene. The expression system can be used to screen for those promoters that enhance gene expression of the ACCase gene at least about 5 to 10-fold over the endogenous levels of ACCase produced normally in the plant cells. Because the 35S Came promoter is known as a strong promoter, it is likely this promoter WO 99/67367 PCT/US99/14022 66 will provide for at least a 5-fold increase in the expression of ACCase over that normally produced in the plant cell. In addition, this expression system can be used to screen antisense DNA sequences. For example, an antisense sequence can be obtained that is 5 complementary to an about 0.5 kb region of the maize ACCase cDNA that has high homology with a portion of the chicken ACCase gene and contains the sequence for the presumed transcarboxylase active site domain, as shown in FIG. 8. The antisense sequence could be subcloned into a pBI121 or pBI221 expression under the control of an inducible plant promoter, such as nitrite 10 reductase promoter (Back et al., Plant Molec. Biol., 12:9-18 (1991)). The ability of the antisense sequence to inhibit expression of the native ACCase gene can be evaluated in transformed cells, for example as described in Hamilton et al., Nature, 346:284-287 (1990). EXAMPLE X 15 Identification and Cloning of the Gene From Herbicide Resistant Maize Cell Lines Herbicide resistant maize cell lines were generated as described in Examples I, II, and IV. These herbicide resistant cell lines have been shown to produce an ACCase enzyme that is less sensitive to inhibition by sethoxydim or 20 haloxyfop. The genes encoding the herbicide resistant forms of the ACCase will be identified and cloned using standard methods as described in Sambrook et al., Guide to Molecular Cloning: A Laborator Manual (1989). The genes encoding the herbicide resistant forms of ACCase can then be introduced into herbicide sensitive plant species by standard methods to confer herbicide resistance . For 25 example, the ACCase enzyme in the maize cell line 2167-9/2160-154 S-1 is at least 100-fold less sensitive to sethoxydim than the wild-type. DNA from the cell line or plants will be obtained and digested with EcoRI and/or other appropriate restriction enzymes, according to standard methods. The restriction enzyme digest will be separated out by agarose gel 30 electrophoresis and probed with either the 2 kb or the 3.9 kb cDNA ACCase probe described in Example V. Fragments hybridizing to the 2 kb or 3.9 kb probe will be subcloned into a Bluescript vector and portions of the gene will be WO 99/67367 PCT/US99/14022 67 sequenced, as described in Example V, to verify that the entire ACCase gene has been isolated. To confirm that the clone encodes the ACCase gene, it will be subcloned into the pBI121 or pBI221 expression vector, as described in Example 5 VIII. The ACCase gene product expressed by the clone in either Black Mexican sweet corn cells or tobacco cells will be evaluated for reactivity with ACCase specific antibodies, by enzyme activity, and/or resistance of the enzyme activity to inhibition with sethoxydim and/or haloxyfop. It is likely that the cloned gene will encode an ACCase which is resistant to inhibition by sethoxydim and 10 haloxyfop. This gene can then be introduced into an herbicide-sensitive embryogenic plant cell or an embryo, including maize cells or immature embryos, to confer herbicide resistance to that plant species upon regeneration. The complete coding sequence encoding the herbicide resistant form of the ACCase enzyme will be cloned into a plant transformation vector 15 such as pBIl21 or pBI221 as described in Jefferson, Plant Molec. Biol. Reporter, 5:387-405 (1987). This vector contains the 35S Came constitutive promoter, the P-glucuronidase structural gene, and the nopaline synthase 3' polyadenylation signals. The P-glucuronidase gene is replaced with a cloned ACCase gene. Optionally, the cloned ACCase gene can be combined with natural or 20 synthetically produced chloroplast transit peptide sequence from pea, as described in Keegstra & Olsen, Ann. Rev. Plant. Physiol./Mol. Biol., 4_0:471-501 (1989) and/or unique restriction sites introduced so the cloned gene can be distinguished from the endogenous maize ACCase gene. Standard methods of subcloning will be utilized as described in Sambrook et al., cited supra. 25 For transformation of maize cells, type II calli can be transformed using biolistic transformation, as described by W.J. Gordon-Kamm et al., Plant Cel, 2, 603-618 (1980); M.E. Fromm et al., Bio/Technology, -, 833-839 (1990); and D.A. Walters et al., Plant Molecular Biology. 19, 189-200 (1992). Alternatively, type I embryogenic calli can be transformed using electroporation 30 after mechanically or enzymatically wounding calli, as described by D'Hafluin et al., The Plant Cell, 4:1495 (1992). Once the cloned gene is introduced into these cells and transformants are selected, typically by antibiotic resistance, fertile transgenic maize plants can be regenerated, as described by D'Hafluin et al. cited WO 99/67367 PCT/US99/14022 68 supra. Fertile transgenic plants can be evaluated for herbicide tolerance, as described in Example III. It is likely that the fertile transgenic plants having and expressing a cloned ACCase gene as an ACCase resistant to sethoxydim and/or haloxyfop will exhibit herbicide tolerance as compared to the corresponding 5 untransformed plant. EXAMPLE XI Generation of Transgenic Plants Having an Increase in Oil Content Once identified and cloned, the gene or genes from maize acetyl CoA carboxylase can be introduced into monocot or dicot plant species, 10 including maize, under the control of a promoter that provides for overexpression of the ACCase enzyme. The overexpression of the ACCase enzyme is likely to lead to an increase in the oil content of the plants and seeds. Naturally occurring soybeans that have a high oil content and soybeans that have a low oil content have been identified. The acetyl CoA 15 carboxylase from both types of soybeans was isolated, as described in Example V. The activity of the enzyme was measured as a function of the time of seed development and the results are shown in Figure 11. The results in the Figure 11 indicate that higher oil content soybean is associated with a 2-fold increase in the ACCase activity during early 20 to mid stages of development when compared with a low oil content soybean. Thus, increased expression of the ACCase gene correlates with an increase in the oil content of the seed. Total oil content of the seed was also measured at maturity (60 days). The high oil producing cell lines, Anoka and P128C. 134, have a total oil content of 21.8% and 19.9%, respectively. In contrast, the low 25 oil soybean line of M76-395, has an oil content of 13.6% oil. Thus, the increase of ACCase expression early in seed development correlates with a higher total oil content in the seed at maturity. A gene encoding a genomic maize acetyl CoA carboxylase can be isolated, as described in Example V, and used to transform plant species by 30 protoplast or biolistic transformation. If the gene is combined with a strong promoter, such as the 35S cauliflower mosaic virus promoter, overexpression of the ACCase gene is likely. Alternatively, selecting transformed cells with multiple copies of the gene can also result in transformed cells overexpressing WO 99/67367 PCT/US99/14022 69 the ACCase gene. The gene can be cloned into a vector such as pBI121 or pBI221, as described by Jefferson, cited supra. This vector contains the 35S cauliflower mosaic virus promoter, the -glucuronidase structural gene, and the nopaline synthase 3' polyadenylation signals. The cloned ACCase gene can 5 replace the p-glucuronidase gene and then be used to transform plant cells, including maize, as described in Example VIII. Transformed cells can be screened for overproduction of ACCase. The presence of the cloned gene can be verified by identifying the unique restriction enzyme sites incorporated into the cloned gene. ACCase levels can 10 be assessed by standard enzyme assay methods and quantitative Western blots using antibodies specific for maize ACCase. Fatty acid and lipid content in cells lines overproducing ACCase are likely to be elevated and can be assessed using standard methodologies, as described in Clark & Snyder, JACS, fi:1316 (1989). Transformed cell lines overproducing ACCase and having increased total oil 15 content will be used to regenerate fertile transgenic plants and seeds, as described in D'Hafluin, cited supra. EXAMPLE XII Generation of Transgenic Plants Having an Increase in Plastidic ACCase Activity 20 Maize embryos are transformed with sense and antisense cDNA constructs encoding the plastidic Al ACCase. Selected transgenic cultures and regenerated transgenic plants and progeny are subjected to detailed analyses of: ACCase transcript levels; activity of ACCase I (plastidic) and ACCase II 25 (presumably cytosolic) in various plant tissues; fatty acid synthesis; lipid accumulation in kernels (primarily embryos); and other plant traits. The culmination of these tests indicates whether plastidic ACCase activity can be modified via transformation and whether fatty acid synthesis is affected. 1. Transformation of maize embryos and plant regeneration 30 Immature maize embryos of the Hi-II genotype are transformed by particle bombardment according to slight modifications from previously described procedures (Fromm et al., Biotech., -, 833 (1990); Koziel et al., Biotech., 11, 194 (1993)). This procedure has been employed in transformation studies for bombardment of approximately 15,000 embryos which were then WO 99/67367 PCT/US99/14022 70 selected for Basta-resistant callus (bar selectable marker gene expression), and regenerated into plants. Transformed (Basta-resistant) plants are obtained from 1-2% of the initial embryos and, when separate plasmids are used for co transformation, the nonselected transgene is recovered in about 50% of the 5 Basta-resistant plants. Basta-resistant, hemizygous transformed (To = F 1) plants will be tested by PCR or Southern blots for the presence of Al ACCase sequences unique to the transformation vector, grown to maturity in growth chambers, greenhouse or field, and self-pollinated when possible, or backcrossed to a 10 nontransformed parent. F2 or backcross progeny are grown in the greenhouse and field and tested for Basta resistance and presence of the Al ACCase transgene to identify homozygous transgenic plants. Homozygous inbred transgenic lines are then developed. 2. Analysis of ACCase Al transformants 15 Plants recovered from at least 100 independent transformation events (i.e., from different bombarded embryos) for both the UBI1 and GLB1 vectors are recovered. Regenerated plants are tested for the presence of the intact ACCase Al transgene and its cosegregation with the Basta-resistance marker. Homozygous and heterozygous transgenic lines are assayed for total 20 ACCase activity in leaves for UBIl transformants and in developing embryos (22-26 DAP) for both UBIl and GLB1 transformants. Sethoxydim and haloxyfop inhibition are used to quickly determine the levels of herbicide sensitive ACCase I (plastidic) and the herbicide-insensitive ACCase II activity in both leaves and embryos. Increased expression of the ACCase Al transgene 25 contributes to plastidic ACCase activity and not to ACCase II activity. Kernel fatty acid and oil content are analyzed at maturity and their relationship with ACCase I determined by methods well known to the art. Transformants that differ in kernel ACCase activity and/or fatty acid and oil content are then selected for more detailed analysis of embryos 30 throughout development (4-day intervals from 16 DAP to maturity). These analyses include RNAse protection assays to determine total Al + A2 transcript levels using a non-specific probe and to determine relative levels of endogenous Al versus transgene Al transcripts by use of antisense riboprobes spanning the WO 99/67367 PCTIUS99/14022 71 5' UTR region of the Al transgene constructs. Western blots of total proteins separated by SDS-PAGE gels are probed with the ACCase I-specific antibody described hereinabove or with avidin and analyzed by densitometry to distinguish changes in the 227-kD ACCase I and 219-kD ACCase II isoforms. 5 ACCase I activity, fatty acid and lipid content are determined in embryos at each stage of development. These analyses determine whether expression of an additional gene(s) for plastidic ACCase increases ACCase activity and consequently fatty acid and oil content in maize tissues, especially in embryos. 3. Transformation with maize ACCase Al antisense transformation vectors 10 Antisense transformation vectors were constructed by blunt-end ligation of nucleotides 1-833 of SEQ ID NO:5 in reverse orientation into multicloning sites of both the GLB1 and UBIl plasmids. A sense construct with the same 833-bp cDNA sequence also was made with the GLB 1 plasmid to serve as a transformation control. Insert orientations were verified by restriction 15 mapping. UBI1 antisense, GLB1 antisense, and GLB1 sense constructs were introduced into > 2100, > 2900 and > 2000 embryos, respectively, and Basta resistant callus were selected. If antisense expression results in significant reduction in ACCase activity, it may not be possible to obtain viable callus or plants from the 20 constitutive UBIl antisense transformants. Similarly, plants transformed with the embryo-specific GLB 1 antisense construct may exhibit deleterious effects on embryo development. Thus, failure to obtain transgenic progeny containing the antisense ACCase gene from these transformations may indicate that ACCase activity cannot be downregulated without loss of viability. 25 4. Analysis of ACCase Al antisense transformants All Basta-resistant cultures will be regenerated. The presence of the UBIl and GLB 1 antisense constructs will be determined by PCR analysis for unique transgene sequences such as the Ubi-] intron/ACCase Al junction or Gib1 5' UTR/ACCase Al junction, or by Southern blotting to detect unique 30 fragments. Plants and lines homozygous or heterozygous for the antisense transgene are analyzed for steady state level of the ACCase Al antisense transcripts in appropriate tissues/organs (such as leaves, tassels, ears, embryos and endosperm for UBIl; leaves and embryos for GLB1) by using ACCase Al WO 99/67367 PCT/US99/14022 72 sense riboprobes for hybridization on RNA blots or for RNAse protection assays. Total ACCase activity (both ACCase I and ACCase II isoforms) and fatty acid and lipid content are determined for the antisense transgenic lines and for corresponding tissues from nontransformed control plants. These analyses show 5 whether ACCase Al antisense transgenes are expressed in plants and, if so, whether expression is associated with reduced ACCase activity and altered fatty acid and lipid content in maize. EXAMPLE XIII Expression of Plastidic and Cytosolic ACCases during Plant Development 10 Intact embryos are isolated from developing kernels of field grown inbred B73 at 2 to 4 day intervals between 16-42 DAP and frozen immediately in liquid nitrogen. Samples also are saved for fresh and dry weight determinations. Subsamples from each stage are analyzed for total lipid and fatty acid content. Seedling leaves are sampled along the leaf blade ranging 15 from the etiolated, meristematic basal region to the fully expanded, green tip. Leaves and other tissues (e.g., epidermis) of maize genotypes that accumulate anthocyanin pigments are also analyzed to assess whether a specific ACCase (such as a cytosolic ACCase) is more highly expressed in tissues in which malonyl-CoA also is required as a substrate for chalcone synthase in the 20 flavonoid pathway leading to anthocyanin synthesis. Gene-specific antisense riboprobes in RNAse protection assays are employed to determine Al, A2, BI and B2 transcript levels. The corresponding sense transcripts are produced in vitro and used as standards to verify specificity and quantitate the sample transcript levels. Quantitation is 25 done on an AMBIS radioanalytic image system. Herbicide inhibition of total ACCase activity provides an assessment of levels of herbicide-sensitive ACCase I (plastidic) and the herbicide-insensitive ACCase II activity in these tissues. ACCase I and II isoforms are separated by ion-exchange chromatography. Total proteins are separated by SDS-PAGE and Western blots probed with avidin to 30 detect the biotinylated 227-kD ACCase I and 219-kD ACCase II isoforms or probed with ACCase I-specific antibodies. EXAMPLE XIV WO 99/67367 PCT/US99/14022 73 Identification and Isolation of an Oat Gene Associated with Groat Oil Content Cultivated oat, Avena sativa L. (2n=6x=42), is grown mostly in temperate regions. In many parts of the world, oat grain is produced for many 5 purposes such as for animal feed, human food, and cosmetic and pharmaceutical additives (Webster, 1986). Seventy-eight percent of the world oat production is used for livestock feed (Schrickel, 1986). Among the cereals, oat groats have the highest lipid concentration (Morrison, 1978), but due to their high fiber content oat is lower in energy value than many other cereals when used for livestock 10 feed. Oat is also unique among the cereals because more than 50% of total seed lipid is deposited in the starchy endosperm rather than the embryo of developing caryopses (Peterson and Wood, 1997; Youngs, 1986). An increase in oil content is desirable for making oat a higher-energy feed grain. On the other hand, a decrease in oil content would make oat a more attractive commodity for use in 15 many current human diets where oat is desired for the type and concentration of its fiber (Ripsin et al., 1992). A polygenic pattern of inheritance has been indicated for groat oil content with heritability estimates ranging from 63 to 93% (Schipper and Frey, 1991 a) with primarily additive gene action (Thro and Frey, 1985). The average 20 groat oil concentration in contemporary cultures is 6 to 7% but variation from 2 to 16% oil occurs in diverse germplasm (e.g., Kanota has about 11%, Ogle about 9% and Marion about 8% of oil (tryglycerides plus polar lipids) content, per dry weight). Genotype by environment interaction for this trait is small (Gullard, 1986; Branson and Frey, 1989). After five cycles of a recurrent selection 25 program, Schipper and Frey (1991b) were able to produce oat lines with groat oil content greater than 16%. This response to selection is similar to the classical Illinois long-term selection experiments for oil content in maize (Zea mays L.) that resulted in dramatic progress toward altering the oil content in the original variety (Dudley et al., 1974). These studies demonstrate that, except when 30 physiological limits are encountered, long-term selections for a quantitative, highly heritable and environmentally stable trait can prove successful. Analyses of 5987 oat samples representing a wide range of Avena germplasm showed a range of 2 to 12% for free non-polar lipids, with polar lipids contributing an WO 99/67367 PCT/US99/14022 74 additional 2 to 3% (Youngs et al., 1982). Combinations of alleles for changing the groat oil content beyond this range may exist because this is mostly an unselected trait that exhibits large phenotypic diversity. Therefore, genetic dissection of this trait could expedite a program aimed at modifying the oil level 5 of oat. The recent explosion in development of restriction fragment length polymorphism (RFLP) linkage maps and their use in identifying genomic regions which affect quantitatively inherited traits is well documented (Phillips and Vasil, 1994). Molecular genetic studies relating to seed oil content have 10 been reported for major oil-producing crop plants such as rape seed (Brassica rapa spp. oleifera; Tanhuanpaa et al., 1995a and 1995b), soybean (Glycine max L., Diers et al., 1992; Mansur et al., 1993; Brummer et al., 1997) and maize (Goldman et al., 1994; Alrefai et al., 1995). These studies indicate that genomic regions can be identified which significantly influence seed oil content. In the 15 maize long-term Illinois selection material, two populations, the high oil by low oil population (Alrefai et al., 1995) and the high protein by low protein population (Goldman et al., 1994) were analyzed to identify genomic regions influencing kernel oil content. In both populations a region on chromosome 6 was identified near the linoleic acid locus. In the one case, this region explains 20 63% of the phenotypic variation in the ratio of oleic to linoleic (18:1 to 18:2) fatty acids and also influenced the total kernel oil concentration (Alrefai et al., 1995). To date, oat plants with an altered fatty acid composition as a result of a qualitative genetic mutation have not been identified, possibly due to the hexaploid nature of the genome. 25 In monocots, graminicides cause plant cell death by inhibiting ACCase and cessation of fatty acid biosynthesis (Burton et al., 1989). Maize mutants have been isolated that grow in the presence of normally lethal concentrations of ACCase-inhibiting herbicides, exhibit up to 20-fold increase in wild-type ACCase activity, and synthesize fatty acids in the presence of 30 herbicide (Parker et al., 1990). Other results point to the role of ACCase in regulating seed oil deposition. A herbicide-insensitive ACCase mutation, Accl, exhibits increased ACCase activity in corn embryos paralleling the reporting time course of oil deposition in this tissue (Somers et al., 1993). In developing WO 99/67367 PCT/US99/14022 75 seeds of caster bean and rape, measured activity of ACCase correlates well with lipid accumulation providing additional evidence as to the role of this enzyme in fatty acid biosynthesis (Simcox et al., 1979; Turnham and Northcote, 1983). Moreover, expression of Arabidopsis cytosolic ACCase in B. napus plastids 5 alters fatty acid composition, with the largest effect being an increase in oleic acid, and increasing total seed oil content by about 5% (Roesler et al., 1997). This increase in oleic acid is consistent with studies in maize indicating a positive correlation between percentage of oleic acid and total seed oil content (Alrefai et al., 1995). 10 Materials and Methods Genetic Material Two oat F 2

:

6 -derived recombinant inbred populations developed by single seed descent (SSD) were used in this study: a population of 137 recombinant inbred lines (RILs) developed from the cross of a facultative winter 15 type cultivar, Kanota, by a spring cultivar, Ogle, (KO) and a second population of 139 RILs from the cross of Kanota by another spring cultivar, Marion, (KM). A portion (71 RILs) of the first population had been employed in previously reported oat mapping and QTL studies (O'Donoughue et al., 1995; Siripoonwiwat et al., 1996; Holland et al., 1997). The second population was 20 chosen on the basis of having a common parent, Kanota, to provide a degree of "biological replication". These materials were grown in 1992-1995 in a randomized complete block design at the following locations; Kanota X Ogle (KO): Aberdeen, Idaho, 1992 and 1993; Ithaca, New York, 1993; and St. Paul and Rosemount, 25 Minnesota, 1994. Kanota X Marion (KM): St. Paul and Rosemount, Minnesota, 1994; and St. Paul, Minnesota, 1995. At Aberdeen, Idaho, the materials were grown in 3 replicates of four-row plots of 2.5 m in length, 30 cm between rows with the center 2 rows harvested. At 30 Ithaca, New York, the materials were grown in 3 replicates of six-row plots of 4 m length and 18 cm between rows. At St. Paul and Rosemount, Minnesota, the materials were grown in 4 replicates of hill plots, with 30 seeds planted per hill WO 99/67367 PCT/US99/14022 76 on 30 cm grid spacing. The exact parents used to generate the populations were included as single or multiple entries in each replicate. Oil Content Analysis To reduce the number of samples needed for analysis, equal 5 amounts of sample from each replicate at a given location/year were bulked to represent a single balanced sampling per line. This resulted in five measurements for each RIL in the Kanota X Ogle population and three for the Kanota X Marion population. Groat oil content was measured by a Bran+Luebbe InfraAlyzer 10 500 on dehulled ground samples. The Near Infrared (NIR) calibration was based on total fat as determined by acid hydrolysis. The calibration set contained 348 oat samples (representing commercial and elite breeding lines from North American oat breeding programs) with fat contents that ranged from 4.95 to 14.01%. The calibration equation had a correlation coefficient of 0.985, a 15 Standard Error of Estimate of 0.299, and a F-value of 11500. Acid hydrolysis extracts the total lipid content of the groat. Thus, the reported values represent the sum of the neutral lipids (triglycerides) plus the polar lipid fractions (Youngs et al., 1982). Thus, these values are higher than those obtained from ether extracts which only contain the neutral lipid fraction. 20 Oat ACCase Clones or PCR Products An oat groat cDNA clone was isolated based on its hybridization to a maize cDNA fragment which included portions of the less conserved mid peptide sequence and the transcarboxylase domain (Shorrosh et al., 1994). Sequence identity indicated that the oat cDNA (3031 bp) corresponded to maize 25 ACCase nt 3124-6169 (Egli et al., 1995). The 1010-aa predicted polypeptide was nearly identical to portions of wheat (89%) and maize (84%) plastidic multifunctional ACCases; it had lowef identity to cytosolic ACCases, including wheat (67.5%) and alfalfa (64.6%) (Gornicki et al., 1994; Shorrosh et al., 1994). An amplified 'Prairie' oat developing groat cDNA library (R. 30 Skadsen, USDA, Madison, WI) cloned into the EcoRI site of AZAPII (Stratagene) was screened with a [ 32 P-dCTP]-labelled maize ACCase cDNA probe (Ausabel et al., 1997) derived from an EcoRI digest of clone 18-5 (nt 2020-5927; Egli et al., 1995). Plaques from eleven partially purified positives WO 99/67367 PCTIUS99/14022 77 were eluted in SM medium (Sambrook et al., 1989) and excised in vivo to obtain pBlueScript plasmids carrying oat cDNA inserts, according to the manufacturer's instructions (Stratagene Corp.). Five colonies per transformation reaction were screened (as colony lifts) with an AflII-BbrPI fragment of maize 5 ACCase cDNA (4281-5797; Egli et al., 1995). One clone, oat-91, hybridized to the probe. Oat-91 was sequenced using the ABI Prizm system (Perkin-Elmer Corp.) at the Advanced Genetic Analysis Center (Univ. of Minnesota, St. Paul, MN). Sequence identity indicated the oat-91 EcoRI insert corresponded to maize ACCase nt 3124-6169 (Egli et al., 1995). Oat ACCase cDNA sequence 10 was compared to other plant ACCases with the BESTFIT program (BESTFIT, Wisconsin Package V. 9.1, Genetics Computer Group (GCG), Madison, WI). For PCR, Pileup alignments of maize, wheat, barley, and rice ACCase coding sequence (GCG, Madison, WI; 12, 18, F, R) were used to identify conserved regions (i) for design of new primers or (ii) regions of identity 15 to available maize ACCase oligos, to be used for PCR amplification of oat ACCase. Two maize primers, PCR2 (5'-ggcagagcaa agcttggagg; SEQ ID NO:25) and PCR1 (5'-tccaccagag aagcctctcc; SEQ ID NO:26) appeared like to amplify oat cDNA based on their identity to other monocots and previous use to amplify soybean ACCase. These primers are located within the transcarboxylase 20 domain, surrounding the acetyl-CoA binding site of ACCase. An amplified 'Prairie' oat seedling leaf cDNA library cloned into the EcoRI site of IZAPII (Skadsen, University of Wisconsin) was used as a source of cDNA. One I of the oat cDNA library was amplified with maize primers (PCR2/PCR1), using the Expand High Fidelity PCR system (BMB) in a GeneAmp PCR System 9600 25 thermal cycler (Perkin-Elmer) after denaturing phage heads at 94'C for 10 minutes. The single, 222-bp product was purified by Wizard PCR Plus (Promega Corp.) and sequenced in an AB1377 (Perkin-Elmer) using PCR1 as primer. The resulting sequence appeared to be from oat since it had high identity to maize plastidic ACCase (89%). Additional 5' cDNA was obtained by (i) PCR 30 of the oat cDNA library with T3/PCR1 primers to generate a complex band mixture and (ii) reamplification of 0. 11 products with maize ACCase primers 1324+ (5'-caatcacgtcagactg; SEQ ID NO:27)/PCR1 to obtain a 1.9-kb product with the same 3' sequence as the PCR2/PCR1 product.

WO 99/67367 PCTIUS99/14022 78 RFLP and Statistical Analysis Briefly, two statistical procedures were used to determine the location, effect and genotype by environment interaction (GxE) for significant QTLs. The first procedure was single factor ANOVA (ANOVA and GLM 5 procedures; SAS Institute, 1990) with the added criteria for consistency in association of genomic regions with a trait over environments and linkage distances (Kianian et al., 1998). If the association were significant when considering the average of all the environments, the association was also significant over the majority of the environments (3 of 5 in KO and 2 of 3 in 10 KM), and closely linked markers within that region, if available, also showed significant associations. A "backward elimination" process as described by Kennard and Havey (1995) was used to construct a multi-locus model containing the most important markers identified by the ANOVA procedure. Marker loci were excluded one at a time from this model, based on the criterion of least 15 significant (P < 0.05) Type-III sum of squares (Kennard et al., 1994). In addition, MQTL software (Tinker and Mather, 1995a & b) was used to strengthen and validate the ANOVA results. This software performs Simple Interval Mapping (SIM) and Simplified Composite Interval Mapping (sCIM). It also has the advantages of searching for one QTL while simultaneously 20 accounting for the effect of others, providing separate analyses for each environment, and calculation of a threshold to control the type-I error rate. One thousand permutations were performed to estimate the threshold for a type-I error rate below 5%. Significant QTL regions were determined as described by Tinker et al., 1996. These regions were grouped in a multi-locus linear model to 25 estimate the overall phenotypic variance explained by the model. However, additional criteria were used to include markers in the final MQTL model, which gave 10 cM spacing between QTLs, and minimized the number of regions excluding those markers explaining less than 5% of the phenotypic variance, while maintaining most of the overall R 2

.

WO 99/67367 PCT/US99/14022 79 Results Trait Data Groat oil concentration for both populations showed a relatively continuous distribution with many of the individuals having values within one 5 standard deviation of the parents (Table VI). Among the three parents, Kanota showed the highest concentration of groat oil content followed by Marion and then Ogle. The groat oil content value for the Kanota line derived from the parent used in generating the KM population was higher than that used for the KO population (Table VI). This is not unexpected: Kanota is a heterogeneous 10 population that shows variation for other traits, but individual parent lines were used in population development and were maintained as controls for each population. Large numbers of individuals having values outside the range of parents were observed for both populations (Table VI). Transgressive segregants tended more toward the lower groat oil concentration than the high. The KO 15 population contained individuals with the highest and lowest groat oil concentrations, possibly due to a larger difference in parental values (3% between Kanota and Ogle, 2% between Kanota and Marion). Overall, the KM population contained more transgressive segregants than the KO population (Table VI). 20 WO 99/67367 PCTIUS99/14022 80 00 Ln00 C V) r rq Q) m C 00 00 C - c - C)"D0 Cu C \C 00 -- r C) V) C' C)) t oo c~ - En A C Cd o> C)d

CIS

WO 99/67367 PCTIUS99/1 4022 81 646 C4N o ONON 00 WO 99/67367 PCT/US99/14022 82 The correlation among the different environments for oil content was high (Table VII). Measurements for the KO population had an average R of 0.93 with the lowest value of 0.90 between material grown at Aberdeen, ID 1992 and St. Paul, MN 1994 (P < 0.0001). Values for the KM population were 5 less well correlated (average R of 0.79) with the lowest correlation between the samples grown in Rosemount, MN 1994 and St. Paul, MN 1995 (P < 0.0001). The difference in the size of the population correlations could be due to a smaller range of values for the KM population (5%) as opposed to the KO population (7%), or fewer measurements (3 vs. 5 environments), or both. 10 Kanota X Ogle Population RFLP Data A hexaploid oat RFLP map that contains 561 loci has been generated from 71 RILs in the KO population (O'Donoughue et al., 1995). Initial quantitative analysis was performed using this subset of the KO 15 population, the available trait data, and 360 of the 561 loci that are co dominantly segregating and evenly distributed across the oat genome. To take advantage of the entire population and to add power to the analysis, the remaining 66 RILs in this population were also mapped. To date, 153 loci have been mapped with these individuals. Detection of QTL with either the subset 20 (360 markers and 71 individuals) or the entire data set (153 markers and 137 individuals) of the KO population did not make a major difference, except in two situations. In both cases, QTLs were detected in the subset, but were not detected in the entire population due to the absence of RFLP data for the remaining 66 RILs. Results of the entire population are presented herein except 25 for the aforementioned differences. An RFLP map containing all 137 individuals and the 153 mapped loci was generated and was in accord with the published map (O'Donoughue et al., 1995) except for linkage group 3. An inter-varietal translocation is believed to involve this linkage group. In the final analysis, the published map was used 30 for this region. Additional information from aneuploid analysis (Kianian et al., 1997) was also taken into account when two linkage groups were assigned to the same chromosome. Thus, linkage groups 5 and 7 assigned to chromosome 5C, linkage groups 6 and 20 assigned to chromosome 16, and linkage groups 4 and WO 99/67367 PCTIUS99/14022 83 12 assigned to chromosome 21, were analyzed as three distinct units. In order to have a complete coverage of the genome for SIM and sCIM, 344 mapped loci (191 loci scored on only 71 RILs) were used in the MQTL analysis. Quantitative Analysis 5 Analysis of variance using the criteria mentioned revealed eight linkage groups and one unlinked marker significantly associated with groat oil content (P < 0.05 for individual environment and P < 0.0001 using means across environments). These regions are on linkage groups 6 and 7 (Table VIII). Two additional genomic segments on linkage group 37 and an unlinked locus 10 Coleoptal were significant in tests of the first 71 individuals of the mapping population but could not be verified on the whole population due to a lack of RFLP data for the entire set. The positive alleles (those that cause an increase in groat oil content) for all of the genomic regions influencing this trait were from the Kanota parent (Table 3). Overall, the alleles from Kanota contributed 2.8% 15 (4.1% if linkage group 37 and Coleoptal loci were included) to the groat oil content. Regions on linkage groups 6 and 11 are considered highly significant and remained in the full model after the backward elimination process (Table 4). However, the two additional loci from linkage group 37 and Coleoptal were also included in the model if only the first 71 individuals in this population were 20 being analyzed. However, the two additional loci from linkage groups 37 and Coleoptal were also included in the model if only the first 71 individuals in this population were being analyzed. Locus Xedo665B on linkage group 11 accounted for 36% to 48% (depending on the environment) of the phenotypic variance for this trait (P < 0.0001). For these linked loci, a substitution of 25 Kanota alleles for their Ogle counterparts caused an average increase of 2% in groat oil content. Substitution of Kanota alleles for Ogle alleles at the locus Xcdo1414B (or Xumn5lB) on linkage group 37 caused an average increase of 1% in groat oil content and was the second most significant locus considering only the first 71 individuals in this population.

WO 99/67367 PCTIUS99/14022 84 0~ cl,~ co~ $ I- 0 0 ~ 0 ~0 o~~ C 00 CI 0 to 0 0 C 0j)2 0 0v C0 0 m .0 -eq 4- t E4 0 cn~ WO 99/67367 PCT/US99/14022 85 00 000 (ZS 05 0 -0 cz 00 WO 99/67367 PCT/US99/14022 86 Quantitative analysis using the MQTL software gave a slightly different result. Like the ANOVA results, a highly significant locus (P < 0.0001), Xcdo665B on linkage group 11, was identified at a test statistic (TS) value of 290.2 that is roughly equal to a LOD score of 64 (0.22 x TS). This 5 locus overshadowed any other significant region and accounted for more than 37% of the phenotypic variance [Variance QTL Main effect (VM)/Variance Phenotypic (VP) = 38%, Variance QTL x environment Interaction and VM (VI)/VP = 38% and Variance Genetic from background markers (VG)/VP = 7%)] and a change of 2% in groat oil content (Figure 23). However, secondary 10 loci that showed significance below the threshold value (threshold TS = 58.2 after 1000 permutations) were inferred when either SIM or sCIM gave evidence for QTLs. Xumn51B on linkage group 37 and the Coleoptal locus, which is unlinked, were significant at test statistic values of 49.9 and 44.8 respectively. The three loci together accounted for 47% of the phenotypic variance and a 15 change of 3.3% in groat oil content (VM/VP = 47%, VI/VP = 48% and VG/VP = 7%). For all three loci the Kanota parent contributed the positive allele. These loci showed a remarkable consistency over the five environments. Thus, there were no significant peaks inferred from the QTL x E interaction scan (Figure 23). 20 Kanota X Marion Population RFLP Data The results from an initial quantitative analysis of the KO population (first 71 RILs) for groat physical and chemical traits were used to determine markers for mapping the KM population. The markers mapped were 25 not chosen for their association with groat oil content in the KO subset population, but for their association with other groat chemical and physical characteristics (Kianian et al., 1998). Since the level of polymorphism in hexaploid oat is relatively high, most probes detected sequences mapping to more than one location. A total of 60 loci were mapped with the 139 RILs in the 30 KM population. These markers mapped to 20 of the 38 linkage groups that had been identified in the KO population (linkage groups 1, 2, 4, 5, 8, 9, 18, 19, 21, 24, 25, 26, 30, 33, 34, 35, 36, 37, 38 were not covered). The linkage groups not covered were, in general, the smaller ones consisting of only two to a few WO 99/67367 PCTIUS99/14022 87 markers in the KO map. The data from the KM population generated an additional linkage group of two loci, and 13 markers remained unlinked. Except for linkage group 37 and Coleoptal locus, this KM map covered all regions with significant associations with groat oil concentration in the KO population. 5 Quantitative Analysis Single factor ANOVA revealed four linkage groups significantly associated with groat oil content in the KM population. These were groups 3, 11, 22, and 5X (P < 0.05 for individual environment and P < 0.0001 using means across environments). As in KO, all of the genomic regions influencing this trait 10 were from the Kanota parent. Overall, the Kanota genomic regions contributed 2.1% to the groat oil content. Regions on linkage groups 11, 22, and 5X (group 5X has been tentatively assigned to linkage group 5 but more markers are needed to accurately place this linkage group on the hexaploid oat RFLP map) are considered highly significant and remained in the full model after the backward 15 elimination process. As in the KO population, locus Xcdo665B on linkage group 11 was highly significant for its effect on groat oil content (P < 0.0001), accounting for 16% to 20% (depending on the environment) of the phenotypic variance for this trait. Overall, the three significant loci explained 15% of the phenotypic variance and together accounted for a total difference of 1.7% in 20 groat oil content. Analysis by SIM or sCIM gave the same result as ANOVA. Linkage groups 11, 22, and 5X were significant in the KM population. Locus Xcdo665B on linkage group 11 was significant at a test statistic value of 70.5 (LOD = 15.5). This locus was not as dominant in the KM population as it was in 25 the KO population. This locus alone explained 18% of the phenotypic variance (VM/VP = 18%, VI/VP = 19% and VG/VP = 3%) and accounted for 0.9% change in groat oil content. The other significant locus above the threshold value was Xcdo1199C on linkage group 5X (TS = 33.1) accounting for a change of 0.2% in groat oil concentration. A secondary locus inferred was Xcdo506B on 30 linkage group 22, which accounted for 0.4% change in oil content. These three loci together accounted for 20% of the phenotypic variance and 1.5% in groat oil content (VM/VP = 20%, VI/VP = 22% and VG/VP = 3%). As expected from the trait data, these loci showed a varied effect over the three environments and one WO 99/67367 PCT/US99/14022 88 interaction on linkage group 11 is notable (Figure 23). The QTL x E interaction peak on linkage group 11 was at TS = 6.6, slightly below the significant threshold, and was due to a range of 0.5% in groat oil content explained [the values for the groat oil content were 0.6% (St. Paul, Minnesota, 1995) and 1.1% 5 (Rosemount, Minnesota, 1994)]. ACCase and its Relationship to Groat Oil Content QTLs An oat groat cDNA was isolated based on its hybridization to a maize cDNA fragment which included portions of the less conserved mid peptide sequence and the transcarboxylase domain (Shorrosh et al., 1994). The 10 3031 -bp oat cDNA was nearly identical to wheat (90%) and maize (84%) genes encoding plastidic ACCase (Egli et al., 1995; Gornick et al., 1997). A 424 nucleotide sequence form the 3' end of the insert was translated and compared to other plant ACCase peptides. The oat peptide sequence was 93% identical to plastidic isozymes from maize and wheat (Egli et al., 1995; Gornicki et al., 15 1997), but only 78 to 83% identical to cytosolic ACCase from maize, wheat (Gronicki et al., 1994), or alfalfa (Shorrosh et al., 1994). The 3031-bp oat ACCase cDNA clone identified a single polymorphic locus in the KO population (AccaseA) and two polymorphic loci in the KM population (AccaseA and AccaseB). The AccaseA locus maps 2.3 cM 20 from Xcdo665B and 3.7 cM from Xisu2287 on linkage group 11. The AccaseB locus maps 11 cM from Xcdo665D and 7.3 cM from Xbcd1 729B. In the KO population AccaseA explains on average 37% of the phenotypic variance and in KM 18% of the phenotypic variance for groat oil content (for both cases P < 0.0001 using means across all environments). The individuals in both 25 populations can be divided into two groups based on their genotype for this locus and groat oil content using the mean across all environments (Figure 3). These two classes differ in their oil content by an average of 1.67% in KO and 0.78% in KM. Results of QTL analysis for the AccaseA locus are the same as that reported for Xcdo665B (Table 4). 30 Table X. Peptide sequence identity of oat and other multifunctional ACCases. Partial, internal sequences are 1010 amino acids unless noted otherwise. Presence of an N-terminal plastid transit peptide is noted, if known.

WO 99/67367 PCT/US99/14022 89 ACCase Gene source GenBank # Transit Peptide? % identity Avena sativa AF072737 ? 100 Triticum aestivum AF029895 yes 89.2 Zea mays U19183 yes 84.1 5 Triticum aestivum U10187 no 67.5 Zea mays Egli, unpubl. ? 78.6 # Hordeum vulgare X99102 ? 71.9 § Arabidopsis thaliana L27074 no 63.7 Brassica napus X77576 yes 63.9 10 # based on 145 aa at the C-terminus of oat-91 § based on 498 aa at the C-terminus of oat-91 partial cDNA, unknown N-terminus 15 Discussion The ACCase cDNA clone from developing oat seeds hybridizes to a genomic region that is linked to a QTL locus on linkage group 11 having a major influence on the total lipid content of oat groats. The influence of this 20 locus was detected in various environments and two genetic backgrounds, thus confirming the value of biological as well as environmental replication. These results provide strong support for the hypothesis that ACCase has a major role in determining groat oil content in crop plants. Either allelic variation in ACCase activity controls oil content in oat groats, or a gene(s) closely linked to ACCase 25 is responsible. In other studies of oat directed at transferring a major crown rust resistance locus on linkage group 11 (Bush et al., 1994), lines with the introduced rust resistance gene often had altered groat oil content. These results further indicate the likely presence on linkage group 11 of a genetic locus with a critical role in the fatty acid biosynthetic pathway in oat. Mutations, which 30 cause a gradual but demonstrable major effect on a biochemical pathway, are believed to be the most common type in polyploids (Sears, 1972). Identification of a major quantitative trait locus influencing the biochemical composition of grain is not unusual. Studies of fatty acid, starch, and protein concentrations and composition provide good examples (Goldman et 35 al., 1993 and 1994; Alrefai et al., 1995; Hu et al., 1995). In many of these cases, WO 99/67367 PCTIUS99/14022 90 the significant region is in close proximity to a known genetic locus with a demonstrated effect on the trait of interest. For instance, in maize, a major genomic region for fatty acid concentration and composition has been identified near the linoleic acid-i locus, and one influencing protein and starch 5 concentration was identified near the Shrunken-2 locus (Goldman et al., 1993 and 1994; Alrefai et al., 1995). These and numerous other reports bolster the hypothesis that allelic variants at known genetic loci may be responsible for quantitative effects (Robertson, 1985). Reliable identification of genomic regions with an influence on a 10 particular quantitative trait has always been difficult. In a number of studies conducted on a given quantitatively inherited trait in various populations and environments, the most significant regions tended to be in common (Goldman et al., 1994; Alrefai et al., 1995). However, the accuracy of identifying a quantitative trait locus will depend, in part on the degree to which it influences 15 the trait in a given population. This is demonstrated in our study by the locus on linkage group 11 which has similar but varied effects in different populations. Marion has a higher concentration of groat oil than Ogle and more loci with positive influence on this trait. The locus on linkage group 11 increased the groat oil content to a lesser degree in the KM population than in the KO 20 population. Correspondingly, the relative proportion of variance due to the environmental influence on this locus was much higher in the KM population than in the KO population. This example illustrates the power of biological replication (i.e., having populations with a common parent) in identifying quantitative trait loci. 25 Identification of the most significant locus does not preclude the identification of other less important loci due to the influence of genetic background and environment on the overall effect. The consistency in identification of significant regions by various quantitative analysis methods in the KM population is a good example. The regions on linkage groups 5, 11 and 30 22 and Xcdo395C locus are considered to make important contributions to the groat oil content in this population. The Xcdo1199C locus located near the Xacor242 locus on linkage group 5 of the published hexaploid oat RFLP map (O'Donoughue et al., 1995) has as great an effect as Xcdo665B-XaccaseA region WO 99/67367 PCTIUS99/14022 91 on linkage group 11 in this population. This locus was not detected in the KO population presumably because the Ogle allele had a similar influence as Kanota and in the KM population the Marion parent was carrying the allele with positive influence on the groat oil content. The other two significant loci in the KM 5 population have a relatively minor effect on this trait and would be more difficult to detect in a population segregating for regions with a larger influence on oil such as in the KO population. Results from other QTL studies can help identify possible pleiotropic effects of other traits such as heading date and vernalization response 10 on groat oil concentration. A region on linkage group 3 with alleles being contributed by Kanota is implicated as having a major influence on heading date and possibly yield, thus reducing its importance as having a direct effect on groat oil concentration (Siripoonwiwat et al., 1996; Holland et al., 1997). The cultivar Ogle was found to have loci on linkage groups 22 and 37 that influence plant 15 height in the KO population (Siripoonwiwat et al., 1996). Loci on linkage group 11 is believed to have an influence on test weight and other groat physical characteristics, thus providing additional evidence of their role in oat groat morphophysiology (Siripoonwiwat et al., 1996; Cakir et al., 1996). These and other studies of quantitative traits in oat are helping identify genomic regions 20 that influence the growth, development and physiology of grain, the primary product. Thus, genetic and molecular manipulation of the oat genome to improve grain quality and quantity can be directed at critical segments. Cultivated oat is a hexaploid composed of three ancestral genomes designated A, C, and D (Rajhathy and Thomas, 1974). Recent studies 25 have indicated that the organization of hexaploid oat genomes into discrete homoeologous groupings such as that of hexaploid wheat is difficult (Rooney et al., 1994; O'Donoughue et al., 1995; Kianian et al., 1997). However, C-genome chromosomes, due to their distinctiveness, have been used as a basis to identify homoeologous segments among the other chromosomes (Kianian et al., 1997). 30 Thus, it is believed that a type of segmental homology instead of whole chromosome homology best describes the genome organization in hexaploid oat (Kianian et al., 1997). Association of homoeologous genomic segments with a quantitative trait further substantiates that relationship.

WO 99/67367 PCTIUS99/14022 92 The region marked by the probe CDO1414 can identify both the significant region on linkage group 11 and the one on linkage group 37 indicating possible homoeology of these regions. The same is true for probes CDO1 199 for linkage groups 6 and 5X and CD0708 for linkage groups 6 and 5 22. Thus, the association of groat oil content with RFLP markers in oat is explained by at least three sets of homoeologous genomic segments. In hexaploid wheat plastidic ACCase maps to the short arm of homoeologous group 2 near the telomere (Gornicki et al., 1997). Plastidic ACCase genes in wheat have been shown to be transcriptionally active in 10 seedling leaves but their effect on seed lipid content are unknown (Gornicki et al., 1997). ACCase in maize has been mapped to short arm of chromosome 2 near the centromere and long arm of chromosome 10 (Egli et al., 1995; Van Dee, 1994). A QTL with moderate effect on oil concentration in progeny of Illinois High and Low oil maize lines was located near the ACCase locus on maize 15 chromosome 2 (Berke and Rocheford, 1995). Additionally, there are numerous defective kernel (dek), reduced endosperm (ren), and Endosperm factor (E) loci influencing grain morphophysiology located on chromosomes 2 and 10 of maize near these synthetic segments (MaizeDB). In maize, oil is stored in embryos and embryo size is associated with oil content of whole seeds (Dudley, 1974). 20 Unlike maize, oat grains contain significant amounts of lipid in the endosperm (Peterson and Wood, 1997), and thus QTLs affecting pathway flux could be more important than those altering seed anatomy. The regions in wheat and maize agree well in terms of conserved syntony and correspond to chromosome 4 of rice, linkage group C of diploid oat and linkage group 11 of hexaploid oat 25 map (Ahn and Tanksley, 1993; Van Deynze et al., 1995). This comparative evidence from other grass species and homoeologous evidence from oat further help reinforce the validity of the association of these genomic regions with groat oil content and composition in oat. 30 Summarv Oat groats are unique among cereals for the high level and the embryo-plus-endosperm localization of lipids. Genetic manipulation of groat quality traits such as oil is desired for optimizing the value of oat in human and WO 99/67367 PCT/US99/14022 93 livestock diets. A locus having a major effect on oil content in oat (Avena sativa L.) groats was located on linkage group 11 by single factor analysis of variance, simple interval mapping and simplified composite interval mapping. A partial oat acetyl-CoA carboxylase (ACCase) cDNA clone with >84% identity to genes 5 for plastidic ACCase of maize and wheat, which catalyzes the first committed step in de novo fatty acid synthesis, identified a polymorphism linked to this major locus. Similar QTL and ACCase locus placements were obtained with two recombinant inbred populations having one parent in common (Kanota X Ogle and Kanota X Marion) and containing 137 and 139 individual lines, 10 respectively. By having a common parent, these populations provide biological replication of the results in that significant genomic regions should be evident in analysis of multiple cross combinations. Similar QTL and ACCase placements were obtained with two recombinant inbred populations having one parent in common (Kanota X Ogle 15 and Kanota X Marion) and containing 137 and 139 individual lines, respectively. By having a common parent, these populations provide biological replication of the results in that significant genomic regions should be evident in analysis of multiple cross combinations. These populations were mapped with RFLP loci (150 and 60) distributed over the genome and were grown over diverse locations 20 (4 and 2) and years (3 and 2) to provide replication for measurements of groat oil content. The locus on linkage group 11 linked to AccaseA accounted for up to 48% of the phenotypic variance for groat oil content. Additional QTLs in both populations were identified which accounted for an additional 10% of the phenotypic variance. Thus, the major locus plus additional QTLs represent three 25 homologous genomic segments in oat that can be followed by as few as three RFLP markers to produce a large shift in the groat oil content. In contrast to oat, a maize ACCase gene was linked only to a moderate oil QTL. Unlike maize, where embryo size affects seed lipid concentration, oat stores significant amounts of lipid in the endosperm, thus 30 QTLs affecting pathway flux may be more important than those altering seed anatomy. If an ACCase gene is indeed the major groat oil QTL, differential ACCase gene expression and/or enzyme activity may control groat oil content.

WO 99/67367 PCT/US99/14022 94 Moreover, if an ACCase DNA is a QTL in oat, it may be useful as a marker for marker-assisted breeding in oat and other cereals, e.g., maize and wheat. Table XI. Genomic regions significantly influencing groat oil content in oat. 5 Groat oil values were averages of five (KXO) or three (KxM) environments. Kanota x Ogle Locus Linkage High-Oil Coefficient of Groat oil, Group Parent variation % dry wt Xcdo665B 11 K 0.38 2.0 # 10 XaccaseA 11 K 0.38 2.0 Coleoptal unlinked K 0.13 0.3 Xcdol414B 37 K 0.10 1.0 Xcdol357C 6 K 0.10 0.8 4.1 Total 15 Kanota x Marion Locus Linkage High-Oil Coefficient of Groat oil, % Group Parent variation dry wt Xcdo665B 11 K 0.18 0.70 # 20 XaccaseA 11 K 0.18 0.70 Xcdol199C 5X K 0.11 0.70 Xbcd1562A 3 K 0.05 0.40 1.80 Total 25 # Xcdo665B and XaccaseA were associated with the same QTL. Groat oil values were averages of five (KxO) or three (KxM) environments. QTLs were identified by single factor analysis of variance. 30 EXAMPLE XV Method to Decrease Oil Content in Oat Seed Youngs (1986) reviewed 11 studies reporting lipid content of oat groats and found an overall range of 2.0 to 11.8% oil concentration in oat groats. Contemporary oat cultivars exhibit an average of 6 to 7% oil (1994-96 Midwest 35 Regional Uniform Nursery data). Recently, Peterson and Wood (1997) reported on the composition and structure of high-oil oat selections ranging from 6.9 to 18.1% oil. While this level of lipid is beneficial for animal feeding because high WO 99/67367 PCTIUS99/14022 95 oil is associated with higher energy value in the feed, high oil content in oats causes rancidity problems during processing. Rapid liberation of fatty acids from triglycerides by endogenous lipases results in fatty acid oxidation. Furthermore, use of current oat cultivars results in relatively high-fat food 5 products, e.g., oatmeal. Oat processors indicate that a 50% reduction of groat lipid content, while maintaining the superior agronomic characteristics of the currently existing cultivars, would be highly desirable. Oat is unique among the cereals in regard to seed lipid deposition (Welch, 1995; Youngs, 1986). Most other cereals accumulate storage lipid 10 primarily in the embryo of developing caryopses. Mature oat groats contain more than 50% of total seed lipid content in the starchy endosperm (Peterson and Wood, 1997). Additionally, significant amounts of oil are found in the bran fraction of processed oats. A substantial portion of the bran is in starchy endosperm and aleurone cell layers which do not separate cleanly from the bran 15 as in wheat. Together the oat endosperm and bran fractions contain >90% of the total lipid in the groat. Thus, the endosperm is the major source of the oil in the groat. Two distinct types of ACCase are known. The first type is typical of eukaryotic organisms where a single, nuclear gene encodes the biotin 20 carboxylase, biotin carboxyl carrier protein, and carboxyl transferase domains within a single, multifunctional polypeptide of 210-280 kDa that, in plants, forms a dimeric active enzyme. Plant lipid synthesis begins in plastids, where ACCase synthesizes the malonyl-CoA needed to initiate and elongate fatty acids of up to 18 carbons (Guenther, 1996). To date, multifunctional ACCase genes 25 from maize, wheat and Brassica have been shown to encode functional chloroplast transit peptides. Both monocots and dicots also appear to contain a cytoplasmic form of the multifunctional ACCase. The cytoplasmic ACCase may be involved in the synthesis of malonyl-CoA needed for such reactions as the synthesis of very long chain fatty acids and flavonoids in plants. 30 The second, multisubunit type of ACCase consists of three functional domains formed from four separable polypeptides and is found in dicot plastids and in prokaryotic organisms, but appears not to be present in the Gramineae. The biotinylated polypeptide component of this multisubunit form WO 99/67367 PCT/US99/14022 96 of ACCase is absent from the Gramineae and the plastid-localized multifunctional ACCase is believed to be responsible for fatty acid biosynthesis. Measurements of the levels of acetyl-ACP and malonyl-ACP in spinach leaves (Post-Beittenmiller et al., 1991) and in isolated spinach and pea 5 chloroplasts (Post-Beittenmiller et al, 1991) in the light and dark have provided evidence that ACCase plays a regulatory role in chloroplast fatty acid biosynthesis in vegetative tissues. In monocots, there also is evidence that ACCase activity regulates fatty acid biosynthesis. Several related cyclohexanedione and aryloxyphenoxypropionate graminicides specifically 10 inhibit monocot multifunctional, plastid-localized ACCase. Complete herbicide inhibition of ACCase results in plant cell death due to cessation of fatty acid biosynthesis (Burton et al., 1989). Based on the apparent flux control coefficient for ACCase in barley and maize leaves exposed to low levels of graminicides, ACCase was shown to be the major flux-controlling enzyme for light-stimulated 15 lipid synthesis in these leaves (Page et al., 1994). Parker et al. (1990b) isolated maize tissue cultures that were selected for growth in the presence of normally lethal concentrations of ACCase inhibiting herbicides. Incubation of wild-type cell cultures in the presence of herbicide inhibited incorporation of acetate into the lipid fraction, while variant cell lines exhibiting up to 20-fold increases in 20 wild-type ACCase activity synthesized fatty acid in the presence of herbicide (Parker et al., 1990b). These results provide further evidence that ACCase is rate-limiting in monocot vegetative tissues. Monocot plastid-localized, multifunctional ACCase is likely involved in regulating seed oil deposition. Herbicide-resistant maize mutant 25 plants were isolated using tissue culture selection (Parker et al., 1990a). The herbicide-resistant phenotype of these plants is conferred by mutations converting plastid-localized, multifunctional, herbicide-sensitive ACCase to herbicide-resistant ACCase (Marshall et al, 1992; Parker et al., 1990a). Two mutant genes for herbicide-insensitive ACCase, Acc1 and Acc2, have been 30 characterized and both are partially dominant (VanDee et al, 1992). Mutant alleles of Acc1 were used to study the role of ACCase in regulating kernel development and fatty acid synthesis. ACCase activity in corn embryos increased during development paralleling the reported time course of oil WO 99/67367 PCT/US99/14022 97 deposition in this tissue (Somers et al., 1993). Furthermore, the maximum value of corn embryo ACCase was two-fold greater than the average ACCase activity determined in seedling leaf extracts, suggesting that the enzyme was involved in elevated synthesis of fatty acids required for oil deposition in the embryo. Using 5 the herbicide-insensitivity of the mutant ACCase as a marker for the Acc1 gene, we determined the tissue-specific expression pattern of that gene. ACCase activity assayed in extracts of developing kernels harvested from herbicide resistant plants was tested for herbicide inhibition. The embryo and endosperm ACCase activities exhibited similar reductions of herbicide inhibition as ACCase 10 extracted from the leaves of these herbicide-resistant mutants, indicating that the Acc1 gene was the major ACCase expressed in the embryo and endosperm of the developing corn kernel. To further investigate this gene-enzyme relationship, homozygous herbicide-resistant and wild-type susceptible kernels were isolated four days after pollination and cultured in vitro to maturity in the presence and 15 absence of a range of herbicide concentrations (Somers et al., 1993). Wild-type kernels cultured on all levels of herbicide tested were shrunken and ceased development presumably because of herbicide inhibition of kernel ACCase activity and thus fatty acid synthesis. Mutant kernels developed normally in the presence of herbicide, indicating that the Acc1-encoded ACCase is the major 20 isoform expressed in the developing kernel. The results from the corn studies indicate that multifunctional, plastic-localized ACCase is a rate-limiting enzyme in fatty acid biosynthesis in vegetative and seed tissues. Complete coding sequences of several multifunctional ACCases from higher plants have been described. They show high amino acid homology 25 within the ATP, carboxybiotin, and acetyl-CoA binding domains and an overall amino acid similarity of 70% or greater. Maize multifunctional ACCase isoforms are encoded by at least four genes, Al, A2, B1, and B2 (Egli et al., 1997). Type A cDNAs are > 98% identical, and are 95% identical to a type B cDNA. Types A and B encode highly similar N-termini and are both likely to 30 encode plastid-targeted isoforms (Egli and Gengenbach, 1996; Egli et al., 1997; Egli et al., 1995). Herbicide resistance in corn (Marshall et al., 1992; VanDee et al., 1992) has been mapped to the same chromosomal locations as an ACCase cDNA probe with high identity to both A and B types (Caffrey et al., 1995; Egli WO 99/67367 PCT/US99/14022 98 et al., 1994; VanDee et al., 1992). Herbicide-tolerant ACCase is expressed in developing corn kernels (Somers et al., 1993) and both A- and B-type cDNAs are found in immature embryos at the time of peak ACCase activity (Egli and Gengenbach, 1996; Somers et al., 1993). In hexaploid wheat, six cDNAs 5 encoding cytosolic ACCase were attributed either to three homoeologous sets of duplicate genes or to allelic differences in each genome (Podkowlinski et al., 1996). Hexaploid wheat also contains three transcriptionally active genes encoding plastidic ACCase with 98% exon sequence identity. These genes are located on homoeologous chromosomes (Gornicki et al., 1997). 10 Oat is used for both human food and animal feed. Crop improvement goals for oat involve two divergent end uses in that high oil oat is desired as an animal feed whereas a lower oil oat is desired to make low-fat oat products for human consumption. Although progress in decreasing oil content in oat can be achieved by plant breeding, a genetic engineering approach may also 15 be used to create a single, dominant low-oil gene that can be introduced into any oat variety either directly by genetic engineering or via plant breeding for oil reduction in contemporary and future cultivars. Low oil in oat may be a value added trait for the crop that may allow processors to produce low-fat oat products that contain previously identified cholesterol-lowering constituents 20 including, but not limited to, p-glucans, thus making oats a more healthful human food. Targeted antisense down-regulation of ACCase in the endosperm using an endosperm-specific promoter may avoid problems that would be caused by inhibition of ACCase in the embryo. Determination of the antisense 25 phenotype by detection of reduced ACCase activity and by quantitation of groat oil content in transgenic lines provides direct evidence for the involvement of ACCase in oat seed oil deposition. Current oat varieties contain about 7% oil which causes rancidity problems during processing and results in high-fat food products. To reduce oil 30 (triglyceride) content in groats, ACCase cDNA, e.g., oat ACCase DNA, is employed in an antisense orientation to specifically down-regulate oat endosperm ACCase activity and thus reduce groat oil content. Preferably, a plant having a dominantly inherited low-oil gene useful for production of low-oil WO 99/67367 PCTIUS99/14022 99 varieties is prepared. As oat oil is primarily localized in the starchy endosperm, down-regulation of endosperm ACCase activity is less likely to cause deleterious effects on the embryo. Moreover, converting oil reduction to a single, dominantly inherited antisense trait is useful in rapid production of low oil 5 varieties in this allohexaploid. Seed from transgenic plants are analyzed for expression of the transgene during development and for oil content at maturity. In addition, low oil oat varieties are produced. Reduction of groat oil content may decrease processing problems associated with oil rancidity and can decrease caloric value resulting in a lower 10 fat food. Reduced-fat oatmeal and other oat products may increase the use of oat as a human food and create a value-enhanced crop for producers of the low-fat oat. In addition, the preparation of high-energy (high oil) oat feed through use of a high-oil gene is also contemplated. Isolation and Characterization of Oat ACCase cDNAs. 15 An oat groat cDNA clone was isolated based on its hybridization to a maize cDNA fragment which included portions of the less conserved mid peptide sequence and the transcarboxylase domain (Shorrosh et al., 1994). Sequence identity indicated that the oat cDNA (3031 bp) corresponded to maize ACCase nt 3124-6169 (Egli et al., 1995). The 1010-aa predicted polypeptide 20 was nearly identical to portions of wheat (89%) and maize (84%) plastidic multifunctional ACCases; it had lower identity to cytosolic ACCases, including wheat (67.5%) and alfalfa (64.6%) (Gornicki et al., 1994; Shorrosh et al., 1994). Therefore, the oat-91 cDNA is likely to encode a portion of a plastid-targeted, multifunctional ACCase polypeptide. 25 An amplified 'Prairie' oat developing groat cDNA library (R. Skadsen, USDA, Madison, Wisconsin) cloned into the EcoRI site of IZAPII (Stratagene) was screened with a [ 32 P-dCTP]-labeled maize ACCase cDNA probe (Ausabel et al., 1997) derived from an EcoRI digest of clone 18-5 (nt 2020-5927; Egli et al., 1995). One to three plaques/plate of eleven partially 30 purified positives were eluted in SM medium plus CHC 13. In vivo excision was carried out to transform SOLR cells with pBlueScript plasmids carrying oat cDNA inserts, according to the manufacturer's instructions (Stratagene Corp.). Five colonies per transformation reaction were rescreened (as colony lifts) with WO 99/67367 PCTIUS99/14022 100 an AflII-BbrPI fragment of maize ACCase cDNA (4281-5797; Egli et al., 1995). One clone, oat-91, hybridized to the probe. Oat-91 was sequenced using the ABI Prizm system (Perkin-Elmer Corp.) at the Advanced Genetic Analysis Center (University of Minnesota, St. Paul, Minnesota). Sequence identity indicated the 5 3.05-kb EcoRI insert of oat-91 corresponded to maize ACCase nt 3124-6169 (Egli et al., 1995). Oat ACCase cDNA sequence was compared to other plant ACCases with the BESTFIT program (BESTFIT, Wisconsin Package Version 9.1, Genetics Computer Group (GCG), Madison, Wisconsin). For PCR, Pileup alignments of maize, wheat, barley, and rice 10 ACCase coding sequences (GCG, Madison, Wisconsin; 12, 18, F, R) were used to identify conserved regions (i) for design of new primers or (ii) regions of identity to available maize ACCase oligos, to be used for PCR amplification of oat ACCase. Two maize primers, PCR2 (5'-ggcagagcaa agettggagg) and PCR1 (5'-tccaccagag aagcctctcc) appeared likely to amplify oat cDNA based on their 15 identity to other monocots and previous use to amplify soybean ACCase. These primers are located within the transcarboxylase domain, surrounding the acetyl CoA binding site of ACCase (50). An amplified 'Prairie' oat seeding leaf cDNA library cloned into the EcoRI site of IZAPII (Skadsen, University of Wisconsin) was used as a source of cDNA. One I of the oat cDNA library was amplified 20 with maize primers (PCR2/PCR1), using the Expand High Fidelity PCR system (BMB) in a GeneAmp PCR System 9600 thermal cycler (Perkin-Elmer) after denaturing phage heads at 94'C for 10 minutes. The single, 222-bp product was purified by Wizard PCR Plus (Promega Corp.) and sequenced in an AB1377 (Perkin-Elmer) using PCR1 as primer. The resulting sequence appeared to be 25 from oat since it had high identity to maize plastidic ACCase (89%; 4, 12, Egli, unpublished). Additional 5' cDNA was obtained by (i) PCR of the oat cDNA library with T3/PCR1 primers to generate a complex band mixture and (ii) reamplification of 0.1 I products with maize ACCase primers 1324+ (5' caatcacgtcagactg)/PCR1 to obtain a 1.9-kb product with the same 3' sequence as 30 the PCR2/PCR1 product. To test whether the oat ACCase cDNA clone oat-91 encodes a plastidic isoform, 422 nt of 3' insert sequence was translated and compared to predicted amino acid sequences of maize plastidic ACCases, a putative maize WO 99/67367 PCT/US99/14022 101 cytosolic ACCase, and to multifunctional ACCases from wheat, barley, alfalfa, soybean, Arabidopsis, Brassica, chicken, human, Ustilago, yeast and Cyclotella. Aligned, degapped peptides were used to construct a phylogram by the UPGMA method based on Kimura protein distances (GCG, Madison, Wisconsin). The 5 oat ACCase sequence was 93% identical to plastidic maize and wheat ACCase sequences, but only 78-83% identical to cytosolic ACCases from maize, wheat, or dicots. The corresponding nucleotide identities were 88-92% and 73-78%, respectively. In the phylogram, monocot ACCases were separated from dicots and grouped according to their known or expected cellular locations (Figure 25). 10 Both sequence identity and phylogenetic grouping indicate that oat cDNA encodes a plastidic ACCase. Additional 5' sequence is obtained by screening the cDNA library with the most 5' oat probes or by PCR of the library with near 5' (upstream) oat plus vector-specific oligonucleotide primers. Alternatively, the 5' end of 15 ACCase message may be obtained by 5' RACE PCR (Frohman, 1993). Total RNA is isolated from developing groats, and digested with DNAse. Complementary DNA is synthesized with Superscript II reverse transcriptase (Gibco BRL) by priming with oat ACCase primers. PCR amplification of cDNA utilizes oat primers plus 5' maize primers. Because chloroplast transit peptide 20 sequences are not well conserved between ACCase genes of maize, wheat, and Brassica, or between different gene families of a single species, and because 5' non-coding sequences also are unlikely to be conserved, completion of the 5' sequence of oat ACCase cDNA is accomplished by 5' RACE-PCR (GIBCO BRL; Gornicki et al., 1994). 5' RACE products and downstream sequences are 25 sequenced to verify that a chloroplast transit peptide coding sequence is present and to identify primers that may be used to generate a complete or partial oat ACCase cDNA product by PCR. Oat cDNA is reamplified using either the Expand High Fidelity or Long Template PCR kits (Boehringer Mannheim Biochemicals), depending on the length of desired product. Products are blunt 30 cloned into pBluescript (Stratagene) or a similar vector, or extended with dATP and cloned into pGEM-T (Promega). Adjacent sites in vector polycloning regions or unique sites introduced by PCR are utilized to move oat ACCase cDNA into expression vectors.

WO 99/67367 PCTIUS99/14022 102 Co-segregation of Oat ACCase cDNA and a Major Groat Oil QTL A major quantitative trait locus (QTL) controlling groat oil content in oat was recently identified by the oat genome mapping program at the University of Minnesota (See Example XIV). Single factor ANOVA showed 5 that RFLP marker Xcdo665B on linkage group 11 was associated with a QTL accounting for approximately 37% of the variance in groat oil in a Kanota by Ogle mapping population (O'Donoughue et al., 1995). The 1.9 kb oat ACCase cDNA was placed on this map to within 2.6 cM of Xcdo665B in the same population (n = 136). This ACCase polymorphism also accounted for about 10 37% of the variation in groat oil content. These results provide strong support for the hypothesis that ACCase has a major role in determining groat oil content. Either allelic variation in ACCase activity controls oil content in oat groats or a gene(s) closely linked to ACCase is responsible. Suppression of ACCase expression by antisense transformation can provide a direct test of whether 15 ACCase activity indeed has a major effect on groat oil. Antisense ACCase Transformation Construct Production Antisense ACCase cDNAs are constructed using standard procedures in the plasmids that are routinely used for oat transformation (Torbert et al., 1995). The cDNA is ligated in antisense orientation relative to an 20 endosperm-specific promoter and is terminated using the nopaline synthase (NOS) 3' untranslated sequence of the Agrobacterium tumefaciens NOS gene, although other termination regions may be employed. This antisense cassette is ligated into a plasmid backbone carrying a neomycin phosphotransferase II (nptII) gene under control of the cauliflower mosaic virus 35S promoter as a 25 selectable marker (Torbert et al., 1995). An investigation of down-regulated phenotypes conferred by expression of various granule-bound starch synthase antisense constructs in potato indicated that the size of the antisense transcript did not affect efficacy of inhibition (Kuipers et al., 1995). Choice of promoter sequences controlling 30 expression of the antisense construct appeared to have important effects in the degree of antisense inhibition of potato granule-bound starch synthase. Effective antisense RNAs have been reported which (relative to the mRNA) are either full length or fragments designed to block 5' end capping, intron splicing, or the poly WO 99/67367 PCT/US99/14022 103 A tail (Guerineau et al., 1994; Inoye, 1988). Preferably, the longest oat ACCase clones are employed in an anti-sense construct. Complete inhibition of vegetative- or seed-expressed ACCase is likely to be lethal because the enzyme is required for general fatty acid 5 biosynthesis, as shown by the lethality of herbicide inhibition of ACCase in susceptible monocots. In oat, it is not known whether tissue-specific isozymes or homologs of ACCase exist, or if they do, whether they vary significantly in nucleotide sequence. Therefore, it is important to target antisense down regulation of ACCase to the developing oat endosperm and that the antisense 10 gene exert a partial down-regulation of endosperm ACCase homologs to avoid deleterious effects of possible endosperm ablation due to complete inhibition of ACCase. Because antisense suppression of heterologous genes or members of a gene family can occur (deLang et al., 1993), and because the partial inhibition of plastidic ACCases but not cytosolic ACCases is desired, partial antisense cDNAs 15 designed from functional domain sequences may be used, or conversely, less conserved regions between functional domains. Comparison of wheat genes for cytosolic and plastidic ACCases (Gornicki et al., 1997; Gornicki et al., 1994) may aid in the design of antisense cDNAs to specifically target plastidic ACCase. 20 The wheat Dy10 high molecular weight glutenin promoter (Belchi and Anderson, 1994) confers strong expression of GUS in oat and expression is specifically limited to the endosperm. Because of its tissue specificity and the likelihood that its expression pattern does not directly coincide with the expression of ACCase during oat endosperm development, the DylO promoter 25 may be useful in restricting the ACCase antisense transcripts to the oat endosperm resulting in a partial inhibition of ACCase expression and the normal development of endosperm cells. Oat Transformation Elite cultivars of oat can be transformed and on average one 30 transgenic tissue culture is produced that regenerates fertile, transgenic plants per two microprojectile bombardments. Generally, from 15 to 30 transgenic plants are produced from the microprojectile bombardments conducted during one experiment. The oat ACCase antisense plasmid is introduced into regenerable WO 99/67367 PCT/US99/14022 104 oat tissue cultures of oat genotypes that exhibit high oil content using microprojectile bombardment as described by Torbert et al. (1995). Tissue cultures are selected for resistance to paromomycin conferred by the nptll selectable marker gene. Paromomycin-resistant tissue cultures are regenerated 5 and plants analyzed for the nptII protein using an NPTII ELISA kit (3' > 5' Inc., Boulder, CO). Plants that express the nptII protein are grown to maturity. The progeny from at least 100 independent transgenic events are then evaluated. Characterization of the ACCase Antisense Phenotype 10 Homozygous T 2 seed are divided for maintenance, advancement to produce greater quantities of seed for more detailed analyses, and for preliminary analysis. Single oat plants may produce up to 100 seed. Therefore, sufficient seed should be available to determine if a transgene phenotype is detectable. Dried mature seed of antisense homozygotes and nontransgenic 15 sister lines are analyzed for oil content initially using nondestructive Near Infrared Reflectance (NIR) spectroscopy. This method requires only 1 g of dehulled groats and is inexpensive. NIR is employed to rapidly identify transgenic lines that exhibit lowered oil content. Selected lines identified by NIR are reanalyzed using either NMR or the gravimetric procedure (Peterson 20 and Wood, 1997). If differences in oil content are observed, oil composition is analyzed by gas chromatography. If the results of these analyses warrant further investigation of the materials, a replicated greenhouse trial is conducted to confirm the reduced oil phenotype in the transgenic plants. This test is conducted on T 3 seed produced from multiple T 2 plants of each transgenic line. 25 If a low-oil phenotype is observed, these materials are entered into a breeding program to initiate variety development. Selected low-oil lines are advanced an additional generation. Seed is isolated at different stages of development and stored under conditions to preserve ACCase activity and mRNA for Northern blot analysis. ACCase 30 activity levels and sense and antisense transcript levels are determined to further characterize the antisense phenotype of these plants. While oat is considered a beneficial alternative crop with multiple uses, production acreage continues to shrink due to poor profitability and WO 99/67367 PCT/US99/14022 105 international competition. Yet oat is used as a high quality animal feed and for production of nutritious human food. Numerous positive human health qualities are attributed to consumption of oatmeal. Foremost among these is the cholesterol-lowering property of oats. A major problem confronting oat 5 processors is that current oat varieties have elevated oil content that precludes the production of a low-fat oat product. Thus, in the production of oat products such as oatmeal, the high oil content offsets the benefit of the cholesterol lowering constituents of the product. Oat oil content is relatively easy to manipulate using conventional plant breeding. However, the trend has been to 10 increase oil content of oats for use as animal feed and only recently has low-oil oat been recognized as important for food products. Thus, oat varieties with reduced oil content may be prepared using genetic engineering. Moreover, a single dominant gene system can vastly simplify conversion of existing agronomically adapted oat varieties to low oil as 15 well as being useful in the production of new low-oil varieties. Based on the demand by processors for low-oil oat varieties, it is warranted to consider low oil as a value-added trait for oat. Increasing the value of oat may increase the acreage devoted to growing oats in the U.S. because of possible increased profitability of the crop. 20 References Abu-Elheiga, L., Jayakumar, A., Baldini, A., Chirala S. S., Wakil, S. J. Human acetyl-CoA carboxylase: characterization, molecular cloning and evidence for two isoforms. Proc. Nat. Acad. Sci. U.S.A., 92:4011-4015 (1995). 25 Ahn, S., Tanksley, S. D. Comparative linkage maps of the rice and maize genomes. Proc. Nat. Acad. Sci. U.S.A., 90:7980-7984 (1993). Alban, C., Baldet, P., Douce, R. Localization and characterization of two structurally different forms of acetyl-CoA carboxylase in young pea leaves, of which one is sensitive to aryloxyphenoxypropionate herbicides. Biochem. J., 30 310:557-565 (1994).

WO 99/67367 PCT/US99/14022 106 Al-Feel, W., Chirala, S. S., Wakil, S. J. Cloning of the yeast FAS3 gene and primary structure of yeast acetyl-CoA carboxylase. Proc. Natl. Acad. Sci. U.S.A., .a9:4534-4538 (1992). Alrefai, R., Berke, T. G., Rocheford, T. R. Quantitative trait locus 5 analysis of fatty acid concentrations in maize. Genome, 38:894-901 (1995). Anderson, J. V., Lutz, S. M., Gengenbach, B. G., Gronwald, J. W. Genomic sequence for a nuclear gene encoding acetyl-coenzyme A carboxylase (GenBank L42814) in soybean. Plant Physiol., 109:338 (1995). Ashton, A., Jenkins, C. L. D., Whitfeld, P. R. Molecular cloning of two 10 different cDNAs for maize acetyl-CoA carboxylase. Plant Mol. Biol., 24:35-49 (1994). Ausubel F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A., Struhl, K., eds. Current Protocols in Molecular Biology, John Wiley & Sons, Inc. (1997). 15 Bailey, A., Keon, J., Owen, J., Hargreaves, J. The ACCI gene, encoding acetyl-CoA carboxylase, is essential for growth in Ustilago maydis. Mol. Gen. Genet., 249:191-201 (1995). Belchl, A. E., Anderson, 0. D. Expression of a novel high-molecular weight glutenin subunit gene in transgenic wheat. Nature Biotechnology, 14 20 (1996). Berke, T.G., Rocheford, T.R. Quantitative trait loci for flowering, plant and ear height, and kernel traits in maize, Crop Sci., 35, 1542-1549 (1995). Betty, M., Ireland, R. J., Smith, A. M. Purification and characterization of acetyl-CoA carboxylase from developing pea embryos. Plant Physiol. J., 25 140:513-520 (1992). Branson, V., Frey, K. J. Recurrent selection for groat oil content in oat. Crop Sci., 28:1382-1387 (1989). Brummer, E. C., Graef, G. L., Orf, J., Wilcox, J. R., Shoemaker, R. C. Mapping QTL for seed protein and oil content in eight soybean populations. 30 Crop Sci., 3-2:370-378 (1997). Burton, J. D., Gronwald, J. W, Somers, D. A., Gengenbach, B. G., Wyse, D. L. Inhibition of corn acetyl-CoA carboxylase by cyclohexanedione and aryloxyphenoxypropionate herbicides. Pest Biochem. Physiol., 34:76-85 (1989).

WO 99/67367 PCT/US99/14022 107 Bush, A. L., Wise, R. P., Rayapati, P. J., Lee, M. Restriction fragment length polymorphisms linked to genes for resistance to crown rust (Puccinia coronata) in near-isogenic lines of hexaploid oat (Avena sativa). Genome, 32:823-831 (1994). 5 Caffrey, J. J., Wurtele, E. S., Nikolau, B. J. Genetic mapping of two acetyl-CoA carboxylase genes. Maize Genetics Cooperation Newsletter, 69:3-4 (1995). Cakir, M., Reeves, D. L., Kahler, A. L. Linkage analysis of RFLP markers and quantitative traits in hexaploid oats. Agronomy Abstracts. 10 American Society of Agronomy, Madison, Wisconsin, pp. 163 (1996). Charles, D. J., Cherry, J. H. Purification and characterization of acetyl CoA carboxylase from developing soybean seeds. Phytochem., 25:1067 (1986). de Lange, P., de Boer, G-J, Mol, J. M. N., Kooter, J. M. Conditional inhibition of P-glucuronidase expression by antisense gene fragments in petunia 15 protoplasts. Plant Molec. Biol., 23:45-55 (1993). Diers, B. W., Keim, P., Fehr, W. R., Shoemaker, R. C. RFLP analysis of soybean seed protein and oil content. Theor. Apple. Genet., 83:608-612 (1992). Dudley, J. W., Lambert, R. J., Alexander, D. E. Seventy generations of selection for oil and protein concentration in the maize kernel. In: Dudley, J. W. 20 (ed.) Seventy generations of selection for oil and protein concentration in the maize kernel. Crop Sci. Soc. Am., Madison, Wisconsin, pp. 181-212 (1974). Egli, M., Gengenbach, B. Plastid localization of a multifunctional acetyl CoA carboxylase in maize. Maize Genetics Cooperation Newsletter, 2U:58-59 (1996). 25 Egli, M. A., Gengenbach, B. G., Gronwald, J. W., Somers, D. A., Wyse, D. L. Characterization of maize acetyl-coenzyme A carboxylase. Plant Physiol, 101:499-506 (1993). Egli, M., Lutz, S., Gengenbach, B. Genetic analysis of maize plastidic acetyl-CoA carboxylase. 1997 Biochemistry and Molecular Biology of Plant 30 Fatty Acid and Glycerolipids Symposium, South Lake Taho, CA, P-4, (1997). Egli, M., Lutz, S., Somers, D., Gengenbach, B. Identification and mapping of maize acetyl-CoA carboxylase genes. Maize Genetics Cooperation Newsleter, 6:92-93 (1994).

WO 99/67367 PCT/US99/14022 108 Egli, M. A., Lutz, S. M., Somers, D. A., Gengenbach, B. G. A maize acetyl-coenzyme A carboxylase cDNA sequence. Plant PhysioL, 108: 1299-1300 (1995). Elborough, K. M., Simon, J. W., Swinhoe, R., Ashton, A. R., Slabas, A. 5 R. Studies on wheat acetyl CoA carboxylase and the cloning of a partial cDNA. Plant Mol. Biol.,24:21-34 (1994). Evenson, K. J., Gronwald, J. W., Wyse, D. L. Purification and characterization of acetyl-coenzyme A carboxylase from diclofop-resistant and susceptible Lolium multiflorum. Plant Physiol., 105:671-680 (1994). 10 Frohman, M. A. Rapid amplification of complementary DNA ends for generation of full-length complementary DNAs: Thermal RACE. Methods in Enzymology, 218:340-356 (1993). Gale, M. D., Devos, K. M., Proc. Nat]. Acad. Sci. USA, 95, 1971 (1998). Goldman, I. L., Rocheford, T. R., Dudley, J. W. Quantitative trait loci 15 influencing protein and starch concentration in the Illinois Long Term Selection maize strains. Theor. Apple. Genet., 82:217-224 (1993). Goldman, I. L., Rocheford, T. R., Dudley, J. W. Molecular markers associated with maize kernel oil concentration in an Illinois high protein X Illinois low protein cross. Crop. Sci., 3_4:908-915 (1994). 20 Gornicki, P., Faris, J., King, I., Podkowinski, J., Gill, B., Haselkorn, R. Plastid-localized acetyl-CoA carboxylase of bread wheat is encoded by a single gene on each of the three ancestral chromosome sets. Proc. Natl. Acad. Sci. U.S.A., 9_4:14179-14184 (1997). Gornicki, P., Haselkorn, R. Wheat acetyl-CoA carboxylase. Plant Mol. 25 BioL,22:547-552 (1993). Gornicki, P., Podkowinski, J., Scappino, L. A., DiMaio, J., Ward, E., Haselkorn, R. Wheat acetyl-coenzyme A carboxylase: cDNA and protein structure. Proc. Natl. Acad.Sci. U.S.A., 91:6860-6864 (1994). Gornicki, P., Faris, J., King, I., Podkowinski, J., Gill, B., Haselkorn, R. 30 Plastid-localized acetyl-CoA carboxylase of bread wheat is encoded by a single gene on each of the three ancestral chromosome sets, Proc. Nat. Acad. Sci. USA, 94, 14179-14184 (1997).

WO 99/67367 PCT/US99/14022 109 Gornicki, P., Podkowinski, J., Scappino, L.A., DiMaio, J., Ward, E., Haselkorn, R. Wheat acetyl-coenzyme A carboxylase: cDNA and protein structure, Proc. Natl., Acad. Sci. IJSA, 91, 6860-6864 (1994). Guenther, S. Genbank accession #X99102 (barley) (1996). 5 Guerineau, F., Waugh, R. Brown, J. W. S. Parameters affecting the activity of antisense RNA sequences in tobacco protoplasts. Plant Cell Rep., 13:703-708 (1994). Guchhait, R. B., Polakis, E., Dimroth, P., Stoll, E., Moss, J., Lane, M. D. Acetyl coenzyme A carboxylase system of Escherichia coli. J. Biol. Chem., 10 249:6633-6645 (1974). Gullord, M. Oil and protein content in oats (Avena sativa L.). In: Lawes, D. A., and Thomas, H. (eds.) Proc. 2nd Int. Oats Conf. Aberystwyth, Wales. 15 18 July 1985. Martinus Nijhoff Publishers, Dordrecht, The Netherlands, pp. 210-213 (1986). 15 Harwood, J. L. Fatty acid metabolism. Ann. Rev. Plant Physiol. Plant Mol. Biol., 39:101-138 (1988). Holland, J. B., Moser, H.S., O'Donoughue, L. S., Lee, M. QTLs and epistasis associated with vernalization responses in oat. Crop Sci., 32:1306-1316 (1997). 20 Hosaka, K., Kianian, S. F., McGrath, J. M., Quiros, C. F. Development and chromosomal localization of genome-specific DNA markers of Brassica and the evolution of amphidiploids and n = 9 diploid species. Genome, 33.:131-142 (1990). Hu, J., Quiros, C. F., Arus, P., Struss, D., Robbelen, G. Mapping of a 25 gene determining linolenic acid concentration in rapeseed with DNA-based markers. Theor. Apple. Genet., 90:258-262 (1995). Inouye, M. Antisense RNA: its functions and applications in gene regulation - a review. Gen, 72:25-34 (1988). Keegstra, K., Olsen, L. J., Theg, S. M. Chloroplastic precursors and their 30 transport across the envelope membranes. Annu. Rev. Plant Physiol. Plant Mol. BiL,40:471-501 (1989).

WO 99/67367 PCTIUS99/14022 110 Kennard, W. C., Havey, M. J. Quantitative trait analysis of fruit quality in cucumber: QTL detection, confirmation, and comparison with mating-design variation. Theor. Apple. Genet., 91:53-61 (1995). Kennard, W.C., Slocum, M.K., Figdore, S.S., Osborn, T.C. Genetic 5 analysis of morphological variation in Brassica oleracea using molecular markers, Theor. Apple. Genet., 82, 721-732 (1994). Kianian, S. F., Wu, B-C, Fox, S. L., Rines, H. W., Phillips, R. L. Aneuploid marker assignment in hexaploid oat with the C genome as a reference for determining remnant homoeology. Genome, 4_0:386-396 (1997). 10 Kolattukudy, P. E. Chemistry and Biochemistry of Natural Waxes. P. E. Kolattukudy, ed., Elsevier, New York, pp. 289-347 (1976). Kondo, H., Shiratsuchi, K., Yoshimoto, T., Mauda, T., Kitazono, A., Tsuru, D., Anai, M., Sekiguchi, M., Tanabe, T. Acetyl-CoA carboxylase from E. coli: gene organization and nucleotide sequence of the biotin carboxylase 15 subunit. Proc. Natl. Acad. Sci. J.S.A., M8:9730-9733 (1991). Konishi, T., Shinohara, K., Yamada, K., Sasaki, Y. Acetyl-CoA carboxylase in higher plants: most plants other than Gramineae have both the prokaryotic and the eukaryotic forms of this enzyme. Plant Cell Physiol., 32:117-122 (1996). 20 Kuipers, A. G. S., Soppe, W. J. J., Jacobsen, E., Visser, R. G. F. Factors affecting inhibition by antisense RNA of granule-bound starch synthase gene expression in potato. Mol. Gen. Genet., 246:745-755 (1995). Lee, E-S J., Gronwald, J. W., Orf, J. H. Acetyl-CoA carboxylase activity in high-oil and low-oil soybeans (Glycine max L.). 11th International Meeting 25 on Plant Lipids. Paris, France. Abstract A14 (1994). Li, S-J, Cronan, J. E. The gene encoding the biotin carboxylase subunit of E. coli acetyl-CoA carboxylase. J. Biol. Chem., 262:855-863 (1992). Li, S-J, Cronan, J. E. The genes encoding the two carboxyltransferase subunits of E. coli acetyl-CoA carboxylase. J. Biol. Chem., 262:16841-16847 30 (1992). Lopez-Casillas, F., Bai, D-H, Luo, X., Kong, I-S, Hermodson, M. A., Kim, K-H. Structure of the coding sequence and primary amino acid sequence WO 99/67367 PCT/US99/14022 111 of acetyl-coenzyme A carboxylase. Proc. Natl. Acad. Sci. U.S.A., 85:5784-5788 (1988). Mansur, L. M., Lark, K. G., Kross, H., Oliveira, A. Interval mapping of quantitative trait loci for reproductive, morphological, and seed traits of soybean 5 (Glycine max L.). Theor. Appl. Genet., 8_:907-913 (1993). Marshall, L. C., Somers, D. A., Dotray, P. A., Gengenbach, B. G., Wyse, D. L., Gronwald, J. W. Allelic mutations in acetyl-coenzyme A carboxylase confer herbicide tolerance in maize. Theor. Apple. Genet., 83:435-442 (1992). Marten, G. C., Shenk, J. S., Barton II, F. E. (ed.) Near infrared 10 reflectance spectroscopy (NIRS): Analysis of forage quality. Agric. Handb. 643. USDA-ARS, Washinton, D. C. (1989). Morrison, W. R. Cereal lipids. In: Pomeranz, Y. (ed.) Advances in cereal science and technology. Vol. 2. Am. Assoc. Cereal Chem., St. Paul, Minnesota, pp. 221-348 (1978). 15 Nikolau, B.J., Hawke, J.C. Purification and characterization of maize leaf acetyl-CoA carboxylase, Arch. Biochem. Biophys., 22, 86-96 (1984). O'Donoughue, L. S., Kianian, S. F., Rayapati, P. J., Penner, G. A., Sorrells, M. E., Tanksley, S. D., Phillips, R. L., Rines, H. W., Lee, M., Fedak, G., Molnar, S. J., Hoffiman, D., Salas, C. A., Wu, B., Autrique, E., Van Deynze, 20 A. A molecular linkage map of cultivated oat. Genome, 38:368-380 (1995). Ohlrogge, H. B., Jaworski, J. B., Regulation of fatty acid synthesis, Annu. Rev. Plant Physiol. Plant Mol.Biol., 4&, 109-136 (1997). Page, R. A., Okada, S., Harwood, J. L. Acetyl-CoA carboxylase exerts strong flux control over lipid synthesis in plants. Biochem. Biophys. Acta, 25 1210:369-372 (1994). Page, R.A., Okada, S., Harwood, J.L. Acetyl-CoA carboxylase exerts strong flux control over lipid synthesis in plants, Biochem. Biophys. Acta, 1210, 369-372 (1994). Parker, W. B., Marshall, L. C., Burton, J. B., Somers, D. A., Wyse, D. L., 30 Gronwald, J. W., Gengenbach, B. G. Dominant mutations causing alterations in acetyl-coenzyme A carboxylase confer tolerance to cyclohexanedione and aryloxyphenoxypropionate herbicides in maize. Proc. Natl. Acad. Sci. IU.S.A., .2:7175-7179 (1990a).

WO 99/67367 PCTIUS99/14022 112 Parker, W. B., Somers, D. A. Wyse, D. L., Keith, R. A., Burton, J. D., Gronwald, J. W., Gengenbach, B. G. Selection and characterization of sethoxydim-tolerant maize tissue cultures. Plant Physiol., 92:1220-1225 (1990b). 5 Peterson, D. M., Wood, D. F. Composition and structure of high oil oat. J. Cereal Sci., 26:121-128 (1997). Phillips, R. L., Vasil, I. K. (eds.) DNA-based markers in plants. Kluwer Academic Publishers. Dordrecht, The Netherlands (1994). Podkowinski, J., Sroga, G. E., Haselkom, R., Gornicki, P. Structure of a 10 gene encoding a cytosolic acetyl-CoA carboxylase of hexaploid wheat. Prcx Natl. Acad. Sci. U.S.A., 93:1870-1874 (1996). Post-Beittenmiller, D. J., Jaworski, J. G., Ohlrogge, J. B. In vivo pools of free and acylated acyl carrier proteins in spinach. Evidence for sites of regulation of fatty acid biosynthesis. J. Biol. Chem., 226:1858-1865 (1991). 15 Post-Beittenmiller, D., Roughan, P. G., Ohlrogge, J. B. Regulation of plant fatty acid biosynthesis: Analysis of acyl-coenzyme A and acyl-acyl carrier substate pools in spinach and pea chloroplast. Plant Physiol., 10:923-930 (1992). Rajhathy, T., Thomas, H. Cytogenetics of oats (Avena L.). Wilkes, A. 20 (ed.). Genetics Society of Canada Miscellaneous Publication No. 2. Genetics Society of Canada, Ottawa, Ontario (1974). Ripsin, C. M., Keenan, J. M., Jacobs, D. R., Elmer, P. J., Welch, R. R., Van Horn, L., Liu, K., Turnbull, W. H., Thye, F. W., Kestin, M., Hegsted, M., Davidson, D. M., Davidson, M. H., Dugan, L. D., Demark-Wahnefried, W., 25 Beling, S. Oat products and lipid lowering. J. Am. Med. Assoc., 267:3317-3325 (1992). Robertson, D. S. A possible technique for isolating genic DNA for quantitative traits in plants. J.Theor.Biol., 112:1-10 (1985). Rocheford, T. R. Chromosome Regions Associated with Control of 30 Maize Kernel Composition. Proceedings of the 49th Annual Sorghum Research Conference, pp. 239-249 (1994).

WO 99/67367 PCT/US99/14022 113 Roesler, K., Shintani, D., Savage, L., Boddupalli, S., Ohlrogge, J. Targeting of the Arabidopsis homomeric acetyl-coenzyme A carboxylase to plastids of rapeseeds. Plant PhysioL, 113:75-81 (1997). Roesler, K. R., Shorrosh, B. S., Ohlrogge, J. B. Structure and expression 5 of an Arabidopsis acetyl-coenzyme A carboxylase gene. Plant Physiol, 105:611-617 (1994). Roessler, P. G., Ohlrogge, J. B. Cloning and characterization of the gene that encodes acetyl-coenzyme A carboxylase in the alga Cyclotella cryptica. L Biol. Chem.,268:19254-19259 (1993). 10 Rooney, W. L., Jellen, E. N., Phillips, R. L., Rines, H. W., Kianian, S. F. Identification of homoeologous chromosomes in hexaploid oat (A. byzantina cv. Kanota) using monosomics and RFLP analysis. Theor. Apple. Genet., 89:329 335 (1994). Sambrook, J., Fritsch, E.F., Maniatis, T. Molecular cloning: A laboratory 15 manual, 2nd ed., Cold Spring Harbor Laboratory Press (1989). Siripoonwiwat, W., O'Donoughue, L.S., Wesenberg, D., Hoffman, D.L., Barbosa-Neto, J.F., Sorrells, M.E. Chromosomal regions associated with quantitative traits in oats. JQTL (http://probe.nalusda.gov:8000/otherdocs/jqtl/jqtl1996-03) (1996). 20 SAS Institute, Inc. SAS procedures guide, Version 6, 3rd ed. SAS Institute Inc., Cary, N.C. (1990). Schipper, H., Frey, K. J. Selection for groat-oil content in oats grown in field and greenhouse. Crop Sci., 31:661-665 (1991 a). Schipper, H. Frey, K. J. Observed gains for three recurrent selection 25 regimes for increased groat-oil content. Crop Sci., 31:1505-1510 (1991b). Schrickel, D. J. Oat production, value and use. In: Webster F. H. (ed.) Oat chemistry and technology. Assoc. Cereal Chem., St. Paul, Minnesota, pp. 1 11(1986). Schulte, W., Schell, J., Toepfer, R. A gene encoding acetyl-coenzyme A 30 carboxylase from Brassica napus. Plant Physiol., 1OL:793-794 (1994). Schulte, W., Topfer, R., Stracke, R., Schell, J., Martini, N. Multi functional acetyl-CoA carboxylase from Brassica napus is encoded by a multi- WO 99/67367 PCT/US99/14022 114 gene family: Indication for plastidic localization of at least one isoform. Pra. Natl. Acad. Sci. U.S.A., 94:3465-3470 (1997). Sears, E. R. The nature of mutation in hexaploid wheat. Symp. Biol. Hung., 12:73-82 (1972). 5 Shintani, D., Roesler, K., Shorrosh, B., Savage, L., Ohlrogge, J. Antisense expression and overexpression of biotin carboxylase in tobacco leaves. Plant Physiol.,114:881-886 (1997). Shorrosh, B. S., Dixon, R. A., Ohlorogge, J. B. Molecular cloning, characterization, and elicitation of acetyl-CoA carboxylase from alfalfa. Pxa 10 Natl. Acad. Sci. U.S.A., 91:4323-4327 (1994). Simcox, P. D., Gerland, W., De Luca, V., Canvin, D. T., Dennis, D. T. Respiratory pathways and fat synthesis in the developing castor oil seed. Can. J. BL, 52:1008-1014 (1979). Siripoonwiwat, W., O'Donoughue, L. S., Wesenberg, D., Hoffman, D. 15 L., Barbosa-Neto, J. F., Sorrells, M. E. Chromosomal regions associated with quantitative traits in oat. JQTL (http://probe.nalusda.gov:8000/otherdocs/jqtl/jqtl1996-03) (1996). Slabas, A. R., Fawcett, T. The biochemistry and molecular biology of plant lipid biosynthesis. Plant Mol. Biol., 19:169-191 (1992). 20 Slabas, A. R., Hellyer, A., Bambridge, H. E. The basic polypeptide subunit of rape leaf acetyl-CoA carboxylase is a 220 kDa protein. Biochem. Soc. Trans., 14:714-715 (1986). Somers, D. A., Keith, R. A., Egli, M. A., Marshall, L. C., Gengenbach, B. G., Gronwald, J. W., Wyse, D. L. Expression of the Accl gene-encoded 25 acetyl-coenzyme-A carboxylase in developing maize (Zea mays L.) kernels. Plant Physiol. ,101:1097-1101 (1993). Somers, D. A., Rines, H. W., Gu, W., Kaeppler, H. F., Bushnell, W. R. Fertile, transgenic oat plant. Bio/Technology, 10:1589-1594 (1992). Somers, D. A., Torbert, K. A., Pawlowski, W. P., Rines, H. W., Progress 30 in genetic engineering of oat. Proceedings of the 5th International Oat Conference and the 7th International Barley Genetics Symposium, Scoles, G., Rossnagel, B., eds. University Extension Press, University of Saskatchewan, pp. 230-234 (1996).

WO 99/67367 PCTUS99/14022 115 Takai, T., Yokayama, C., Wada, K., Tanabe, T. Primary structure of chicken liver acetyl-CoA carboxylase deduced from cDNA sequence. J. Biol. Chem., 263:2651-2657 (1988). Tanhuanpaa, P. K., Vikki, J. P., Vikki, H. J. Identification of a RAPD 5 marker for palmitic-acid concentration in the seed of spring turnip rape (Brassica rapa spp. oleifera). Theor. Apple. Genet., 91:477-480 (1995a). Tanhuanpaa, P. K., Vikki, J. P., Vikki, H. J. Association of a RAPD marker with linolenic acid concentration in the seed oil of rapeseed (Brassica napus L.). Genome, 38:414-416 (1995b). 10 Thro, A. M., Frey, K. J. Inheritance of groat-oil content and high-oil selection in oats (Avena sativa L.). Euphytic, 3_4:251-263 (1985). Tinker, N. A., Mather, D. E. Methods for QTL analysis with progeny replicated in multiple environments. JQTL (http://probe.nalusda.gov:8000/otherdocs/jqtl/jqtl1995-01) (1995a). 15 Tinker, N. A., Mather, D. E. MQTL: software for simplified composite interval mapping of QTL in multiple environments. QTL (http://probe.nalusda.gov:8000/otherdocs/jqtl/jqtl1995-02) (1995b). Tinker, N. A., Mather, D. E., Rossnagel, B. G., Kasha, K. J., Kleinhofs, A., Hayes, P. M., Falk, D. E., Ferguson, T., Shugar, L. P., Legge, W. G., Irvine, 20 R. B., Choo, T. M., Briggs, K. G., Ullrich, S. E., Franckowiak, J. D., Blake, T. K., Graf, R. J., Dofing, S. M., Saghai Maroof, M. A., Scoles, G. J., Hoffian, D., Dahleen, L. S., Kilian, A., Chen, F., Biyashev, R. M., Kudrna, D. A., Steffenson, B. J. Regions of the genome that affect agronomic performance in two-row barley. Crop Sci., 3.6:1053-1062 (1996). 25 Torbert, K. A., Rines, H. W., Somers, D. A. Use of paramomycin as a selective agent for oat transformation. Plant Cell Reports, 14:635-640 (1995). Torbert, K. A., Rines, H. W., Somers, D. A. Transformation of oat using mature embryo-derived tissue cultures. Crop Sci., 3_9:226-231 (1998). Turnham, E., Northcote, D. N. Changes in the activity of acetyl-CoA 30 carboxylase during rapeseed formation. Biochem. J., 212:223-229 (1983). Van Dee, K. L., Somers, D. A., Wyse, D. L. RFLP mapping of Acc1 in Zea mays L. Agronomy Abstracts, p.198 (1992).

WO 99/67367 PCT/US99/14022 116 Van Dee, K.L. Evidence for two independently segregating loci encoding acetyl-CoA carboxylase in Zea mays, M.S. Thesis, Univ. of Minnesota, 62 p. (1994). VanDeynze, A. E., Dubcovsky, J., Gill, K. S., Nelson, J. C., Sorrells, M. 5 E., Dvorak, J., Gill, B. S., Lagudah, E. S., McCouch, S. R., Appels, R. Molecular-genetic maps for group 1 chromosomes of Triticeae species and their relation to chromosomes in rice and oat. Genome, 3:45-59 (1995). Webster, F. H. Oat utilization: past, present and future. In: Webster F. H. (ed.) Oat chemistry and technology. Assoc. Cereal Chem., St. Paul, 10 Minnesota, pp. 413-427 (1986). Welch, R. W. The chemical composition of oats. In: The Oat Crop, Welch, R.W., ed. Chapman and Hall, pp. 279-320 (1995). Youngs, V. L. Oat lipids and lipid in related enzymes. In: Oats: Chemistry and Technology, Webster, F. H., ed. American Association of Cereal 15 Chemists, Inc. pp. 205-226 (1986). Youngs, V. L. Oat lipids and lipid-related enzymes. In: Webster, F. H. (ed.) Oat chemistry and technology. Assoc. Cereal Chem., St. Paul, Minnesota, pp. 205-226 (1986). Youngs, V. L., Peterson, D. M., Brown, C. M. Oats. In: Pomeranz, Y. 20 (ed.) Advances in cereal science and technology. Vol. 5. Am. Assoc. Cereal Chem., St. Paul, Minnesota, pp. 49-105 (1982). Yuzo, M., Takuji, S. Genbank accession #D24786 (rice) (1993). The complete disclosure of all patents, patent documents and publications 25 cited herein are incorporated herein by reference as if individually incorporated. The foregoing detailed description and examples have been given for clarity of understanding only. No unnecessary limitations are to be understood therefrom. The invention is not limited to the exact details shown and described for variations obvious to one skilled in the art will be included within the invention 30 defined by the claims.

Claims

1. An isolated and purified DNA molecule comprising a DNA segment encoding an oat acetyl CoA carboxylase, or a fragment thereof. 5

2. The DNA molecule of claim 1 which comprises a DNA segment comprising SEQ ID NO:23.

3. The DNA molecule of claim 1 or 2 further comprising a DNA molecule 10 encoding an amino terminal chloroplast transit peptide, wherein the chloroplast transit peptide is operably linked to the DNA encoding the oat acetyl CoA carboxylase.

4. The DNA molecule of claim 1 which encodes a polypeptide comprising 15 SEQ ID NO:24.

5. A method of imparting cyclohexanedione or aryloxyphenoxypropanoic acid herbicide tolerance to a plant tissue comprising: (a) introducing an expression cassette comprising a chimeric DNA 20 molecule encoding an oat acetyl CoA carboxylase comprising SEQ ID NO:24, operably linked to a promoter functional in a plant cell into cells of a susceptible plant so as to yield transformed plant cells; (b) regenerating the transformed plant cells to provide a differentiated 25 transformed plant; and (c) expressing the chimeric DNA molecule encoding the oat acetyl CoA carboxylase in the cells of the differentiated plant in an amount effective to render the plant tolerant to levels of a cyclohexanedione or an aryloxyphenoxypropanoic acid herbicide 30 which are toxic to a susceptible plant which does not comprise the expression cassette. WO 99/67367 PCT/US99/14022 118

6. The method of claim 5 wherein the chimeric DNA molecule encodes an acetyl CoA carboxylase, the activity of which is not inhibited by an amount of cyclohexanedione or aryloxyphenoxypropanoic acid herbicide that inhibits the activity of an acetyl CoA carboxylase which is sensitive 5 to inhibition by a cyclohexanedione or aryloxyphenoxypropanoic acid herbicide.

7. The method of claim 5 wherein the chimeric DNA molecule is expressed in the cells at about a 2' to about a 20-fold amount over that of a 10 cyclohexanedione or aryloxyphenoxypropanoic acid herbicide-sensitive acetyl CoA carboxylase.

8. The method of claim 5 wherein the chimeric DNA molecule comprises a DNA sequence comprising SEQ ID NO:23. 15

9. A transformed plant prepared by the method of claim 5.

10. A transformed seed of the transformed plant of claim 9. 20

11. A transformed monocot plant, which plant is tolerant to herbicides comprising: (a) an acetyl CoA carboxylase gene that encodes an acetyl CoA carboxylase the activity of which is inhibited by a cyclohexanedione or aryloxyphenoxypropanoic acid herbicide, 25 and (b) a recombinant DNA molecule encoding an oat acetyl CoA carboxylase operably linked to a promoter functional in a plant, wherein the recombinant DNA molecule is expressed in the plant in an amount effective to confer tolerance to the levels of a 30 cyclohexanedione herbicide or an aryloxyphenoxypropanoic acid herbicide which inhibit the activity of the acetyl CoA carboxylase of step (a). WO 99/67367 PCT/US99/14022 119

12. A method for altering the oil content in a plant comprising: (a) introducing an expression cassette comprising a recombinant DNA molecule encoding an oat acetyl CoA carboxylase operably linked to a promoter functional in a plant cell into the cells of a 5 plant so as to yield transformed plant cells; and (b) regenerating the transformed plant cells to provide a differentiated transformed plant, wherein the plant expresses the DNA molecule encoding the oat acetyl CoA carboxylase in cells of said plant so as to alter the oil content of the differentiated transformed plant 10 cells relative to the oil content of cells of a corresponding untransformed plant.

13. The method according to claim 12, wherein the recombinant DNA molecule is expressed at about a 2- to about a 20-fold increase over the 15 level of the native acetyl CoA carboxylase.

14. The method according to claim 13 wherein the oil content of the plant cells is increased about 1.2- to about 20-fold. 20

15. A transformed plant having an altered oil content in the plant cells comprising: a recombinant DNA molecule encoding an oat acetyl CoA carboxylase, wherein the recombinant DNA molecule is expressed in the cells of the plant in an amount of acetyl CoA carboxylase effective to alter the oil content of the plant cells relative to the oil content in the cells 25 of a corresponding untransformed plant.

16. The transformed plant of claim 15 wherein the transformed plant has an increase in oil content of about 1.2- to about 20-fold in its leaves, seeds, or fruit above that present in a corresponding untransformed plant. 30

17. A transformed plant prepared by the method of claim 12.

18. A seed of the transformed plant of claim 17. WO 99/67367 PCT/US99/14022 120

19. The seed of claim 18 wherein the oil content is increased in the endosperm.

20. A method of introducing an exogenous oat acetyl CoA carboxylase gene 5 into a host cell comprising: (a) introducing to host cells in vitro an expression cassette comprising a DNA molecule encoding an oat acetyl CoA carboxylase operably linked to a promoter functional in the host cell so as to yield transformed cells; and 10 (b) identifying transformed host cells which express the DNA molecule.

21. The method of claim 20 wherein the host cells are plant cells. 15

22. The method of claim 21 wherein the plant cells can be regenerated into a differentiated plant.

23. The method of claim 21 wherein the plant cells are monocot cells. 20

24. The method of claim 21 wherein the plant cells are dicot cells.

25. The method of claim 12, wherein the DNA molecule encoding the oat acetyl CoA carboxylase is introduced into plant cells by a method selected from the group consisting of electroporation, microinjection, 25 protoplast transformation, microprojectile bombardment, and liposomal encapsulation, Agrobacterium-mediated transformation.

26. A method for altering the oil content in a plant comprising: (a) introducing an expression cassette comprising a recombinant 30 DNA molecule encoding an antisense oat acetyl CoA carboxylase transcript operably linked to a promoter functional in a plant cell into the cells of a plant so as to yield transformed plant cells; and WO 99/67367 PCT/US99/14022 121 (b) regenerating the transformed plant cells to provide a differentiated transformed plant, wherein the transcript is expressed in the cells of the differentiated transformed plant so as to alter the oil content of the plant cells relative to the cells of a corresponding 5 untransformed plant.

27. A transformed plant having an altered oil content in the plant cells comprising: a recombinant DNA molecule encoding an antisense oat acetyl CoA carboxylase transcript wherein the transcript is expressed in 10 the cells of the plant so as to alter the oil content of the plant cells relative to the cells of a corresponding untransformed plant.

28. The method of claim 12 or 26 wherein the recombinant DNA molecule further comprises a DNA segment encoding a chloroplast transit peptide 15 which is operably linked to the DNA molecule encoding the acetyl CoA carboxylase.

29. The transformed plant of claim 15 or 27 wherein the recombinant DNA molecule further comprises a DNA segment encoding a chloroplast 20 transit peptide which is operably linked to the DNA molecule encoding the acetyl CoA carboxylase.

30. The method of claim 12 or 26 wherein the recombinant DNA molecule further comprises a DNA segment comprising a endosperm-specific 25 promoter.

31. The plant of claim 15 or 27 wherein the recombinant DNA molecule further comprises a DNA segment comprising a endosperm-specific promoter. 30

32. The method of claim 12 or 26 wherein the cells are monocot plant cells. WO 99/67367 PCT/US99/14022 122

33. The method of claim 32 wherein the cells are wheat, rice, maize, barley, millet, sorghum, sugarcane, rye, oat or timothy cells.

34. The method of claim 12 or 26 wherein the cells are dicot plant cells. 5

35. The method of claim 34 wherein the cells are soybean, rape, sunflower, tobacco, Arabidopsis, petunia, pea, Canola, bean, tomato, bean, potato, lettuce, spinach, alfalfa, cotton or carrot. 10

36. The method of claim 12, 20, or 25 wherein the DNA molecule encodes a polypeptide comprising SEQ ID NO:24.

37. The plant of claim 11, 15, or 27 wherein the DNA molecule encodes a polypeptide comprising SEQ ID NO:24. 15

38. An expression cassette comprising an endosperm-specific transcriptional control region operably linked to a DNA encoding a plant ACCase, or a biologically active subunit thereof. 20

39. The expression cassette of claim 38 wherein the endosperm-specific transcriptional control region comprises DylO.

40. The plant of claim 15 which has increased lipid content, lipid biosynthesis or lipid deposition in groat. 25

41. The plant of claim 27 which has decreased lipid content, lipid biosynthesis or lipid deposition in groat.

42. The plant of claim 15 which has increased storage of seed lipid in 30 endosperm.

43. The plant of claim 27 which has decreased storage of seed lipid in endosperm. WO 99/67367 PCT/US99/14022 123

44. A method to identify an oat plant having altered groat oil content, comprising: (a) contacting a probe comprising at least a portion of a nucleic acid sequence encoding oat acetyl CoA carboxylase with a nucleic 5 acid sample from a plant in an amount sufficient to form complexes, wherein the sample comprises genomic nucleic acid; and (b) detecting or determining the amount of complex formation. 10

45. A method to identify a plant having altered oil content, -glucan content, seed weight, starch content or protein content, comprising: (a) contacting a probe comprising at least a portion of a nucleic acid sequence encoding oat acetyl CoA carboxylase with a sample from a plant in an amount sufficient to form complexes, wherein 15 the sample comprises genomic nucleic acid; and (b) detecting or determining the amount of complex formation.

46. The method of claim 45 wherein the plant is an oat, maize or wheat plant. 20

47. A method for marker-assisted selection of plants having desired properties, comprising: (a) contacting a probe comprising at least a portion of a nucleic acid sequence encoding oat acetyl CoA carboxylase with a sample from a plant in an amount sufficient to form complexes, wherein 25 the sample comprises genomic nucleic acid; and (b) detecting or determining the amount of complex formation.

48. An isolated and purified nucleic acid segment associated with altered oil content or composition in monocots. 30

49. A method to map a restriction fragment length polymorphism in plants, comprising: WO 99/67367 PCTUS99/14022 124 (a) contacting a probe comprising at least a portion of a nucleic acid sequence encoding oat acetyl CoA carboxylase with a sample from a plant in an amount sufficient to form complexes, wherein the sample comprises genomic nucleic acid; and 5 (b) detecting or determining the amount of complex formation.

50. The method of claim 44, 45, 47 or 49 wherein the genomic nucleic acid is from oat, maize, rice or wheat.