TRANSGENES

BACKGROUND OF THE INVENTION

The present invention relates to a transgene that contains a polynucleotide segment encoding at least a portion of a signal sequence from a floury2 gene. For example, a transgene within the invention also can contain a second segment coding for an agronomically high-value protein, such that the transgene expresses a fusion protein comprised of the signal-sequence moiety and amino acid sequence of the high-value protein. The present invention also relates to cereal plants that contain a transgene, such as millet or sorghum plants containing a maize or other heterologous flσury2 gene.

Between 50% and 60% of the protein in maize kernels consists of a mixture of prolamin storage proteins known as "zeins," which are essentially devoid of lysine. This makes the seed nutritionally inferior for monogastric animals. The lysine deficiency of maize spurred extensive efforts to identify mutants with higher levels of this essential amino acid. The maize floury2 (f22) mutant was first described by Emerson et al . , and was reported to result from a semidominant mutation that causes a soft, starchy endosperm. CORNELL UNIVERSITY

AGRICULTURAL EXPERIMENTAL STATION REPORT 180 (1935) .

(The contents of this document and all others mentioned herein are incorporated by reference.) The f!2 allele occurs on the short arm of chromosome 4, and for many years it served as a useful genetic marker for the short arm of chromosome .

In 1964, l2 and another mutation in maize, σpaqrue2 (o2) , took on special interest when it was reported that both of these mutations lead to a substantial increase in lysine content of maize seeds. Nelson et al . , Science

150: 1468-70 (1965), and Mertz et al . , Science 145: 279-80 (1964) . Kernels in normal maize genotypes average around 0.20 to 0.25% lysine, while kernels from o2 and f!2 mutations have lysine contents of 0.3 to 0.35%. But the soft starchy endosperm associated with the fl2 and o2 phenotypes causes the kernels to be susceptible to mechanical damage, which creates a higher susceptibility to insect and fungal damage. Consequently, neither mutant gained widespread commercial application. For many years, o2 and fl2 were considered to be defects of genes regulating zein synthesis. This conclusion was based on the significant and fairly specific effect these mutations have on storage protein synthesis. Both σ2 and f!2 reduce zein synthesis by about 50% of the wild type level, with the o2 mutation specifically affecting the 22-kDa α-zeins, and the f!2 mutation equally affecting synthesis of all classes of zeins.

Other distinctive biochemical differences have been reported for o2 and £12 mutants. Protein bodies in both o2 and f ^* 12 are smaller than normal, but f!2-encoded protein bodies are asymmetrical and misshapen compared to the spherical protein bodies of normal and o2 endosperm. Lending →. Larkins, Plant Cell 1: 123-133 (1989). The o2 mutation is recessive, while the £12 mutation is semidominant, with the severity of the phenotype correlated to dosage of the mutant allele.

The hypothesis that o2 is a zein regulatory gene was confirmed eventually, following its tagging with a transposable element. This led to the molecular cloning of 02 and the demonstration that it encodes a leucine zipper-type transcription factor that binds the promoters of certain α-zein genes and controls expression of the 22-kDa family of α-zein genes. Schmidt et al . , Science 238: 960-63 (1987) . Numerous attempts to tag f22 by a similar strategy were unsuccessful, and the basis of the £12 defect remained unknown.

The failure to discern the nature of the f!2 defect continued despite fairly extensive study of this mutant. Several studies noted the appearance of an unusual α-zein protein in £12 with a molecular weight of 24 kDa, higher than normal zein. Lee et al . , Biochem. Genet . 14: 641-50 (1976); Soave et al . , Maydica 23: 145-52 (1978); Galante et al . , Mol . Gen . Genet . 192: 316-21 (1983) . The level of this protein was found to be dependent on the dosage of the f!2 allele. A high concentration of b-70, the maize homologue of the BiP chaperonin, also was reported to be associated with protein bodies in f!2. Zhang & Boston, Protoplasma 171: 142-52 (1993) . BiP is a member of the hsp-70 protein family that binds malfolded polypeptide chains. The level of b-70 is affected by the dosage of fl2 double mutants, as is the degree to which the protein bodies become misshapen. It also was reported that in o2 /£12 double mutants, the unusual 24- kDa α-zein was not synthesized and the morphology of the protein bodies was similar to that in o2. Thus, the o2 gene was reported to be epistatic to £12 . Lopes et al . , Mol . Gen . Genet . 245: 537-47 (1994) .

Lopes et al . reported three α-zeins proteins in addition to the abnormal 24-kDa protein in the storage protein fraction of £12 , with molecular weights ranging from about 25 to 27 kDa. They also detected a restriction fragment length polymorphism (RFLP) linked to the f22 locus with a 22-kDa α-zein probe. They hypothesized that the characteristics of f22 might be a response to the accumulation of the defective 24-kDa protein, but were unable to prove that the accumulation of this protein was responsible for the f!2 phenotype.

SUMMARY OP THE INVENTION

It is an object of the present invention to provide a fusion protein of a 21 amino acid signal sequence from f!2 with a desired protein.

It is a further object of the invention to provide plants that contain an exogenous DNA sequence compri£3ing this fusion protein, in which expression of the desired protein is increased in seeds of the plant. It also is an object of the present invention to provide a method of increasing the content of essential amino acids in an animal feed without supplementation.

In accomplishing these and other objectives, there has been provided, in accordance with one aspect of the present invention, a cereal plant that contains a transgene comprised of (i) a first polynucleotide segment comprising a nucleotide sequence that encodes the amino acid sequence MATKILALIj-AL-LALLVSATNV and (ii) a second polynucleotide segment coding for a protein. In one preferred embodiment, polynucleotide segment (ii) has a high content of an amino acid selected from the group consisting of methionine, lysine, tryptophan and threonine, such that the amount of said amino acid in seeds of said cereal plant is increased as compared to seeds from otherwise identical plants that are not transformed. In another preferred embodiment, the first and second polynucleotide segments are operably linked to a promoter, such as the £12 promoter, so that said cereal plant expresses both segments under the control of said promoter. Pursuant to other embodiments, the cereal plant can be a maize plant, where segment (ii) is not native to maize, or can be rice, wheat, barley, mille or sorghum, for example.

In accordance with other aspects of the present invention, there also is provided seed produced by a plant as described above and a feed product comprising meal obtained from such seed.

Pursuant to another aspect of the present invention, a transgene is provided that comprises (i) a fiist polynucleotide segment comprising the nucleotide sequence coding for the amino acid sequence MATKIJ--AL---AL---ALLVSATNV and (ii) a second polynucleotide segment coding for a protein. In a preferred embodiment, the transgene

additionally comprises a sequence of £12 selected from one or both of nucleotides 761-3824 and 4613-8335 of Figure 1, described in greater detail below. In addition, a transgene is provided that comprises the f!2 promoter operably linked to a polynucleotide segment.

Other objects, features and advantages of the present invention will become apparent from the following detailed description. It should be understood, however, that the detailed description and the specific examples, while indicating preferred embodiments of the invention, are given by way of illustration only, since various changes and modifications within the spirit and scope of the invention will become apparent to those skilled in the art from this detailed description.

BRIEF DESCRIPTION OF THE DRAWING

Figures 1A-1H together depict the nucleotide sequence of a clone of £12 . Positions 1-760 and 8,336-10,539 are vector sequences, and positions 761-8,335 are the complete nucleotide sequence of f22. Figure 2 shows the nucleotide sequence and deduced amino acid sequence of the coding region of f!2, including the signal sequence. Numbers on the left correspond to position of the first amino acid of each line beginning with -21 to reflect the signal peptide, so that the -1 position is occupied by the C-terminal residue of the signal peptide and the +1 position is occupied by the first amino acid of the predicted mature polypeptide.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

It has been discovered that a gene encoding a 22-kDa α-zein protein, and not a regulatory gene, is responsible for the fl2 mutant phenotype discussed above. In

particular, it has been found that the 24-kDa protein identified in £12 mutants comprises the amino acid sequence of a 22-kDa α-zein plus an uncleaved, 21-amino acid signal peptide. The complete nucleotide sequence of the £12 is shown in Figure l, while Figure 2 depicts the nucleotide sequence and deduced amino acid sequence of the coding region of the 24-kDa protein. The deduced amino acid sequence of the signal peptide begins at position -21 in Figure 2, so that the -l position is occupied by the C-terminal residue of the signal peptide and the +1 position is occupied by the first amino acid of the predicted mature polypeptide. The sequence of the signal peptide is MATKILALI^ALLALLVSATNV. A comparison of this deduced N-terminal amino acid sequence of the 24-kDa α-zein protein with other α-zeins has revealed an alanine to valine substitution at the C-terminal position of the signal peptide, a histidine insertion within the seventh α-helical repeat, and an alanine to threonine substitution with the same α-helical repeat of the protein. When an alanine codon is substituted for the valine codon of the mutant α-zein gene, the in vivo translated protein product is processed correctly in the presence of maize microsomes. The signal peptide targets the α-zein protein tc the lumen of the rough endoplasmic reticulum (RER) . The signal peptide is retained on the 24-kDa α-zein precursor; that is, the 24-kDa α-zein is not processed in f!2 endosperm. The 24-kDa α-zein is believed to remain anchored to the RER membrane, disrupting the normal biogenesis of protein bodies. In normal protein body development, zein proteins are retained within the ER where they coalesce into spherical bodies in which α-zeins are localized to the interior of a shell of cross-linked β- and -zein. Attachment of an α-zein to the RER membrane inhibits its movement into the interior of the protein body. More particularly, interaction of the RER-attached α-zein with the shell of β- and 7-zeins

disrupts the spatial organization of developing protein bodies by forming multiple foci for α-zein aggregation near the surface of the ER membrane, which the irregular budding that occurs in f22 protein bodies. The retention of the signal peptide on the 22-kDa α-zein also provides an explanation for the overexpression of BiP in £12 , since this would affect the normal folding of the protein.

The conversion of valine to alanine in the signal peptide of this α-zein provides an explanation for its retention on the protein, and for many of the phenotypic effects of the f!2 mutation. According to von Heijne's "-3- rule" for signal peptides, the -1 position is critical for recognition by signal peptidase and is generally occupied by an uncharged amino acid with a small side chain. See von Heijne, Eur. J. Biochem. 133: 17-27 (1983) .

A point mutation in the signal peptide is consistent with the fact that only one f22 allele has been identified. A point mutation also helps to explain the difficulty in conventional approaches to tag fl2 by transposon mutagenesis.

The complete fl2 sequence contains 7575 base pairs, nucleotides 761-8,335 of Figure 1. The f22 coding region (open reading frame) , including the stop codon, comprises nucleotides 3,825-4,613. This sequence has been transformed into maize. Transgenic seed that contained the gene expressed the f\22, 24-kDa zein, and seed segregating which did not have the f22 protein did not have the gene.

Since the coding region matches the sequence of f!2, 24-kDa protein, it is understood that the sequence shown in Figure 1 includes the promoter for f!2. Nucleotides between nucleotides 761 and about 3,824 in the sequence of Figure l encode the f!2 promoter. Several motifs common to 22-kDa zein promoters are found in this region of the f!2 sequence of Figure 1. For example, located upstream of the start of initiation is a sequence

5 ' -GTCATTCCAC-3 ' . The first nucleotide is at -300 rfith respect to the start of initiation. This corresponds to part of the sequence recognized by the 02 gene product, also located 300 bp upstream of the start of initiation, as disclosed in Figure 5 of Schmidt et al . , Plant ell 4:689 (1992) .

Similarly, Morton et al . refer to a prolamin-specific 5' -TGTAAAG-3 ' motif common to all zein genes of maLze, commonly referred to as the "-300 box" by virtue of its location 300 bp from the start of translation. "Regulation of Seed Storage Protein Gene Expression, " in SEED DEVELOPMENTAND GERMINATION (Kigel andGaliii, eds.) , New York: Marcel Dekker, Inc. (1995) . A corresponding sequence is found at nucleotide 3500 in the fI2 sequence of Figure 1. Morton et al. also disclose a 5 ' -CATGCATG-3 ' element common to many seed-specific genes. This sequence is similar to the sequence 5' -CATGCGTG-3' of £12, which begins at nucleotide 3517 in Figure 1. The retention of the 24-kDa protein on the RER and its accumulation in the endosperm leads to a concomitant decrease in the levels of other zein proteins and, hence, to a decreased level of total storage proteins. The reduction in total storage protein leads to the soft, starchy phenotype of f22 and the reduction of the zein fraction as a percentage of total storage protein le»ads to an apparent increase in lysine content, since the other storage proteins in maize, such as globulins and albumins, have higher lysine contents. The discovery of the nature of the f!2 defect provides the basis for seed progeny genetica-lly engineered to express various phenotypes of agronomic interest. That is, a fusion protein of the 21 amino εicid signal sequence from f22 with a desired polypeptide dan be used to provide increased expression of the desired polypeptide in seeds of a host plant. The region from 3,888-4,613 is replaced with the coding region (open reading frame) of the desired polypeptide.

Alternatively, f22 sequences upstream and downstream of the coding region of f22 can be used to drive expression of a desired polypeptide. In this case, the region from 3,825-4,613 in 22 is replaced by the coding region of the desired polypeptide.

A wide variety of polypeptides can be fused to the f!2 signal sequence and/or to the regions upstream or downstream of the f!2 coding region, in order to achieve higher levels of expression and/or accumulation of the polypeptide than might otherwise be attained. The polypeptides may or may not be native to the plant in which they are being expressed. Illustrative of such polypeptides are short polypeptides of less than about 40 amino acids, polypeptides which are either unusually hydrophilic or hydrophobic or amphipathic, polypeptides which have unique solubility characteristics, polypeptides which have a unique three-dimensional structure, polypeptides which have motifs which include binding domains, polypeptides which have either very few or many disulfide bonds, polypeptides which have a high content of charged amino acids, polypeptides which have a particularly high content of specific amino acids, e . g. methionine, lysine, tryptophan or threonine, and polypeptides which have been altered from their native structure such that they might otherwise not be stable. DNA coding for the polypeptide may be modified to reflect preferred codon usage in the particular crop that is the target of the transformation.

A fusion protein of the signal peptide with a protein other than a zein protein should not affect the total amount of zein storage proteins that are accumulated by the plant. Inasmuch as it is the decrease in storage proteins in fl2 that is reported to give rise to the starchy endosperm characteristic of the f22 mutant, a phenotype like that of the f22 mutant is not expected when plants are transformed with a fusion protein of the f!2 signal peptide and a protein other than a zein protein.

Alternatively, the entire fl2 gene may be transformed into a plant that produces seeds that are mechanically stronger than maize. For example, sorghum and millet produce smaller and rounder seeds which are less affected by shear forces and, hence, to by any kernel weakening associated with £12 expression. Expression of I2 in these crops can provide seeds with improved digestibility without unacceptable levels of kernel weakening. It is expected that rice, wheat and barley likewise could accommodate £12 expression in accordance with the present invention.

In accordance with the present invention, a DNA molecule comprising a transformation/expression vector is engineered to include the 21 amino acid signal sequence from f22 and/or regions upstream or downstream of the f22 coding region, and either the remainder of the f22 gene or a DNA segment encoding a high-value protein, as described above. A copy of the sequence of £12 or of the 21 amino acid signal sequence from f\12 coupled to a desired high-value protein is placed into an expression vector by standard methods. The selection of an appropriate expression vector will depend upon the method of introducing the expression vector into host cells.

A typical expression vector contains: prokaryoti DNA elements coding for a bacterial replication origin and an antibiotic resistance gene to provide for the growth and selection of the expression vector in the bacterial host; a cloning site for insertion of the exogenous DNA sequence; eukaryotic DNA elements that control initicttion of transcription of the exogenous DNA sequence, such as a promoter and an optional enhancer; and DNA elements that control the processing of transcripts, such as a transcription termination-polyadenylation sequence. The vector also could contain additional sequences that, are necessary to allow for the eventual integration of the vector into a chromosome. For a general description of plant expression vectors, see Gruber et al . , "Vectors for

Plant Transformation, " in METHODS IN PLANT MOLECULAR BIOLOGY AND BIOTECHNOLOGY 89-119 (CRC Press, 1993) .

Expression of the gene sequence is under the control of a promoter. Examples of suitable promoters are the promotor for the small subunit of ribulose-1,5-bis- phosphate carboxylase, promoters from tumor-inducing plasmids of Agro-bacterium tu-mefacier-s, such as the nopaline synthase and octopine synthase promoters, and viral promoters such as the cauliflower mosaic virus (CaMV) 19S and 35S promoters or the figwort mosaic virus 35S promoter. The promoter can be constitutive or inducible.

Especially preferred is a "seed tissue-preferred" or "seed tissue-specific" promoters, that is, promoters that drive high expression of the heterologous DNA segment in seed tissue where control of genes that are involved in seed metabolism is desired, and little or no expression in other parts of the plant. Manufacture of the protein encoded by the heterologous DNA segment in other parts of the plant needlessly expends the plant's energy. Examples of known seed tissue-preferred or seed tissue- specific promoters include the seed-directed promoters from the zein genes of maize endosperm. Pedersen et al . , Cell 29: 1015 (1982) . The £12 promoter is particularly preferred.

In addition to a suitable promoter, one or more enhancers are useful in the invention to increase transcription of the introduced DNA segment. The enhancer or enhancer-like element can be inserted into the promoter to provide higher levels of transcription. Examples of such enhancers include, inter alia, viral enhancers like those within the 35S promoter, as shown by Odell et al., Plant Mol . Biol . 10: 263-72 (1988), and an enhancer from an opine gene as described by Fromm et al., Plant Cell 1: 977 (1989) .

Select-able marker genes, in physical proximity to the introduced DNA segment, are used to allow transformed cells to be recovered by either positive genetic

selection or screening. The selectable marker genes also allow for maintaining selection pressure on a transgenic plant population, to ensure that the introduced DNA segment, and its controlling promoters and enhancers, are retained by the transgenic plant.

Many of the commonly used positive selectable marker genes for plant transformation have been isolated from bacteria and code for enzymes that metabolically detoxify a selective chemical agent which may be an antibiotic or a herbicide. Other positive selection marker genes encode an altered target which is insensitive to the inhibitor.

A preferred selection marker gene for plant transformation is the BAR or PAT gene, which is used with the selecting agent bialaphos. Spencer et al . , THSOR. APPL. GENET., Berlin: Springer International, vol. 79, pp 625-631, 1990. Another useful selection marker gene is the neomycin phosphotransferase II (nptJJ) gene, isolated from Tn5, which confers resistance to kanamycin when placed under the control of plant regulatory signals. Fraley et al . , Proc . Na '2 Acad. Sci . USA 80: 4803 (1983) . The hygromycin phosphotransferase gene, which confers resistance to the antibiotic hygromycin, is a further example of a useful selectable marker. Vanden Elzen et al., Plan Mol . Biol . 5: 299 (1985). Additional positive selectable markers genes of bacterial origin that confer resistance to antibiotics include gentamicin acetyl transferase, streptomycin phosphotransferase, aminoglycoside-3' -adenyl transferase and the bleomycin resistance determinant. Hayford et al . , Plant Physiol . 86: 1216 (1988); Jones et al . , Mol . Gen . Genet . 210: 86 (1987); Svab et al . , Plant Mol . Biol . 14: 197 (1990); Hille et al., loc. ci t . 7: 171 (1986).

Other positive selectable marker genes for plant transformation are not of bacterial origin. These genes include mouse dihydrofolate reductase, plant

5-enolpyruvylshikimate-3-phosphate synthase and plant acetolactate synthase. Eichholtz et al., Somatic Cell

Mol . Genet . 13: 67 (1987); Shah et al . , Science 233: 478 (1986); Charest et al . , Plant Cell Rep. 8: 643 (1990).

Another class of useful marker genes for plant transformation with the DNA sequence requires screening of presumptively transformed plant cells rather than direct genetic selection of transformed cells for resistance to a toxic substance such as an antibiotic. These genes are particularly useful to quantitate or visualize the spatial pattern of expression of the DNA sequence in specific tissues and are frequently referred to as reporter genes because they can be fused to a gene or gene regulatory sequence for the investigation of gene expression. Commonly used genes for screening presumptively transformed cells include 0-glucuronidase (GUS) , 3-galactosidase, luciferase, and chloramphenicol acetyltransferase. Jefferson, Plant Mol . Biol . Rep. 5: 387 (1987); Teeri e a ., EMBO J. 8: 343 (1989); Koncz et al . , Proc . Nat ' l Acad. Sci . USA 84: 131 (1987) ; De Block et al . , EMBO J. 3: 1681 (1984) . Another approach to the identification of relatively rare transformation events has been use of a gene that encodes a dominant constitutive regulator of the Zea mays anthocyanin pigmentation pathway. Ludwig et al . , Science 247: 449

(1990) . In order to create an expression vector containing the gene or DNA segment of interest, an expression cassette first is made by inserting a cloned f!2 gene, or a DNA segment comprising the f!2 signal sequence fused to a desired high-value protein as described above, into a plasmid under the control of a regulatory sequence. The resulting expression cassette can be ligated back to itself to produce an expression cassette with a tandem repeat of the cloned gene. A further ligation can be performed to generate a construct that contains four tandem copies of the gene.

One or more copies of the expression cassette containing the introduced DNA segment corresponding to the f!2 gene or to the DNA segment comprising the f!2

signal sequence fused to a desired protein is transferred to an expression vector. In a preferred embodiment, the vector also contains a gene encoding a selection marker which is functionally linked to promoters that control transcription initiation.

To create a transgenic plant, an expression vector containing the f!2 gene or the DNA segment comprising the £12 signal sequence fused to a desired protein can be introduced into protoplasts; into intact tissues, such as immature embryos and meristems; into callus cultures or into isolated cells. Preferably, expression vectors are inserted into intact tissues, such as explants derived from hypocotyl or cotyledonary nodes of a germinated seed. (In this regard, an e plant is a piece of tissue that is taken from a donor plant and is capable of producing callus in culture. Hypocotyl tissue is that portion of the stem of a plant embryo or seedling below the cotyledons and above the root. A cotyledon is an embryonic leaf, and a cotyledonary node is that part: of the seedling between the embryonic axis and the cotyledons which botanically defines the division of the hypocotyl and the epicotyl, or embryonic shoot,) General methods of culturing plant tissues are provided, for example, by Miki et al . , "Procedures for Introducing Foreign DNA into Plants." in METHODS IN PLANT MOLECULAR BIOLOGY AND BIOTECHNOLOGY 67-88 (CRC Press 1993) .

Preferably, the f!2 gene or DNA segment comprising the f!2 signal sequence fused to a desired high-vεtlue protein is transformed into embryogenic maize callus by particle bombardment. Transgenic maize plants are produced by bombardment of embryogenically responsive immature embryos with tungsten particles associated with DNA plasmids. The plasmids consist of a selectable and an unselectable marker gene. The present invention is described further by reference to the following, illustrative examples.

Example l. Characterization o£ 24 -kDa protein o£ £12

Wild-type maize, W64A+, was crossed with W64Afl2 to give an F2 population, according to Lopes et al . , 1994, the contents of which are incorporated herein by reference. The progeny were sorted into floury, semi- floury and normal kernels, and DNA was isolated from 30 individuals in each phenotypic class. To identify DNA polymorphisms linked with the fl2 mutation, a bulked segregant analysis was used. Michelmore et al . , Proc. Nat ' Acad. USA 88: 9828-32 (1991). After extensive screening with different DΝA probe/restriction enzyme combinations, an RFLP was detected in SstI-digested DΝA with a 22-kDa α-zein cDΝA. A 7.7 kb DΝA fragment was present in the homozygous f!2 bulked sample and the heterozygous semi-floury bulk, but not in the normal bulk.

The 7.7 kb Sstl fragment was isolated from genomic

DΝA of W64Afl2, and the resulting clone, pCC515, was mapped by restriction enzyme digestion, as described in Coleman et al . , Proc . Nat ' l Acad . Sci . USA 92: 6828-31

(1995) , the contents of which are incorporated herein by reference. Clone pCC515 was found to contain a single

22-kDa α-zein coding sequence, which was obtained as a

1.6 kb EcoRl fragment. Upon nucleotide sequence analysis, the deduced amino acid sequence was found to correspond to a 22-kDa α-zein. The protein contains 262 amino acids, including a 21 amino acid signal peptide.

Comparison of the deduced amino acid sequence of the signal peptide with the signal peptides of other α-zeins reveals an alanine to valine substitution at the C-terminal (-1) residue of the signal peptide, insertion of a histidine following the seventh residue in the seventh ct-helical repeat and an alanine to thoureonine substitution in the same α-helical repeat. Ν-terminal sequence analysis of the purified 24-kDa protein from f!2 endosperm showed an identical match for the first 45 amino acid residues between pCC515 and the 24-kDa polypeptide. The signal peptide that targets the protein

into the lumen of the rough endoplasmic reticulu was found to be attached to the protein.

Example 2. Transformation of maize with the fl2 gene

To prove that pCC515 contains the f!2 gene, the gene was transformed into embryogenic maize callus by particle bombardment. Transgenic maize plants were produced by bombardment of embryogenically responsive immature embryos with tungsten particles associated with DNA plasmids. The plasmids consist of a selectable and an unselectable marker gene. Preparation of tissue

Immature embryos of maize variety High Type II were the target for particle bombardment-mediated transformation. This genotype is the F- of two purebred genetic lines, parents A and B, derived from the cross of two know maize inbreds, A1Θ8 and B73. Both parents were selected for high competence of somatic embryogenesis, according to Armstrong et al . , Maize Genetics Coop. News 65: 92 (1991). The High Type II genotype does not possess the native mutant f!2 gene.

Ears from F- plants were selfed or sibbed, and embryos were aseptically dissected from developing caryopses when the scutellum first became opaque. This stage occurred about 9-13 days post-pollination, and most generally about 10 days post-pollination, depending on growth conditions. The embryos were about 0.75 to 1.5 millimeters long. Ears were surface sterilized with 20-50% Clorox for 30 minutes, followed by three rinses with sterile distilled water. Immature embryos were cultured with the scutellum oriented upward, on embryogenic induction medium comprised of N6 basal salts, Eriksson vitamins, 0.5 rrιg/1 thiamine HCL, 30 gm/1 sucrose, 2.88 gm/1 L-proline, 1 mg/1 2,4-dichlorophenoxyacetic acid, 2 gm/1 Gelrite, and 8.5 mg/1 AgNo ₃ . Chu et al . , Sci . Sin . 18: 659

(1975); Eriksson, Physiol . Plant 18: 976 (1965). The medium was sterilized by autoclaving at 121 ^β C for 15

minutes and dispensed into 100 X 25 mm Petri dishes. AgN0 ₃ is filter-sterilized and added to the medium after autoclaving. The tissues were cultured in complete darkness at 28°C. After about 3 to 7 days, most usually about 4 days, the scutellum of the embryo had swelled to about double its original size and the protuberances at the coleorhizal surface of the scutellum indicated the inception of embryogenic tissue. Up to 100% of the embryos displayed this response, but most commonly, the embryogenic response frequency was about 80%.

When the embryogenic response was observed, the embryos were transferred to a medium comprised of induction medium modified to contain 120 gm/1 sucrose. The embryos were oriented with the coleorhizal pole, the embryogenically responsive tissue, upwards from the culture medium. Ten embryos per Petri dish were located in the center of a Petri dish in an area about 2 cm in diameter. The embryos were maintained on this medium for 3-16 hour, preferably 4 hours, in complete darkness at 28°C just prior to bombardment with particles associated with plasmid DNAs containing the selectable and unselectable marker genes.

To effect particle bombardment of embryos, the particle-DNA agglomerates were accelerated using a DuPont PDS-1000 particle acceleration device. The particle-DNA agglomeration was briefly sonicated and 10 μl were deposited on macrocarriers and the ethanol was allowed to evaporate. The macrocarrier was accelerated onto a stainless-steel stopping screen by the rupture of a polymer diaphragm (rupture disk) . Rupture was effected by pressurized helium. The velocity of particle-DNA acceleration was determined based on the rupture disk breaking pressure. Rupture disk pressures of 200 to 1800 psi were used, with 650 to 1100 psi being preferred, and about 900 psi being most highly preferred. Multiple disks were used to effect a range of rupture pressures. The shelf containing the plate with embryos was placed 5.1 cm below the bottom of the macrocarrier

platform (shelf #3) . To effect particle bombardment of cultured immature embryos, a rupture disk and a macrocarrier with dried particle-DNA agglomerates were installed in the device. The He pressure delivered to the device was adjusted to 200 psi above the rupture disk breaking pressure. A Petri dish with the target embryos was placed into the vacuum chamber and located in the projected path of accelerated particles. A vacuum was created in the chamber, preferably about 28 in Hg. After operation of the device, the vacuum was released and the Petri dish was removed.

Bombarded embryos remained on the osmotically- adjusted medium during bombardment, and 1 to 4 days subsequently. The embryos were transferred to selection medium comprised of N6 basal salts, Eriksson vitamins, 0.5 mg/1 thiamine HCL, 30 gm/1 sucrose, 1 mg/1 2,4- dichlorophenoxyacetic acid, 2 gm/1 Gelrite, 0.85 mg/1 Ag N0 ₃ and 3 mg/1 bialaphos (Herbiace, Meiji) . Bialaphos was added filter-sterilized. The embryos were subcultured to fresh selection medium at 10 to 14 day intervals. After about 7 weeks, embryogenic tissue, putatively transformed for both selectable and unselected marker genes, proliferated from about 7% of the bombarded embryos. Putative transgenic tissue was rescued, and that tissue derived from individual embryos was considered to be an event and was propagated independently on selection medium. Two cycles of clonal propagation were achieved by visual selection for the smallest contiguous fragments of organized embryogenic tissue.

A sample of tissue from each event was processed to recover DNA. The DNA was restricted with a restriction endonuclease and probed with primer sequences designed to amplify DNA sequences overlapping the f!2 and non- £12 portion of the plasmid. Embryogenic tissue with amplifiable sequence was advanced to plant regeneration. For regeneration of transgenic plants, embryogenic tissue was subcultured to a medium comprising MS salts

and vitamins (Murashige & Skoog, Physiol . Plant 15: 473 (1962)), 100 mg/1 myo-inositol, 60 gm/1 sucrose, 3 gm/1 Gelrite, 0.5 mg/1 zeatin, 1 mg/1 indole-3-acetic acid, 26.4 ng/1 ±cis-trans-abscissic acid, and 3 mg/1 bialaphos in 100 X 25 mm Petri dishes, and was incubated in darkness at 28°C until the development of well-formed, matured somatic embryos could be seen. This required about 14 days. Well-formed somatic embryos were opaque and cream-colored, and were comprised of an identifiable scutellum and coleoptile. The embryos were individually subcultured to a germination medium comprising MS salts and vitamins, 100 mg/1 myo-inositol, 40 gm/1 sucrose and 1.5 gm/1 Gelrite in 100 X 25 mm Petri dishes and incubated under a 16 hour light:8 hour dark photoperiod and 40 μeinsteinsm^sec ^"1 from cool-white fluorescent tubes. After about 7 days, the somatic embryos had germinated and produced a well-defined shoot and root. The individual plants were subcultured to germination medium in 125 X 25 mm glass tubes to allow further plant development. The plants were maintained under a 16 hour light:8 hour dark photoperiod and 40 μeinsteinsm ² sec ^"1 from cool-white fluorescent tubes. After about 7 days, the plants were well-established and were transplanted to horticultural soil, hardened off, and potted into commercial greenhouse soil mixture and grown to sexual maturity in a greenhouse. An elite inbred line was used as a male to pollinate regenerated transgenic plants. Preparation of particles Fifteen mg of tungsten particles (General Electric) , 0.5 to 1.8 μ, preferably 1 to 1.8 μ , and most preferably

1 μ, were added to 2 ml of concentrated nitric acid.

This suspension was sonicated at 0°C for 20 minutes

(Branson Sonifier Model 450, 40% output, constant duty cycle) . Tungsten particles were pelleted by centrifugation at 10000 rpm (Biofuge) for one minute, and the supernatant was removed. Two milliliters of sterile distilled water were added to the pellet, and brief sonication was used to resuspend the particles. The

suspension was pelleted, one milliliter of absolute ethanol was added to the pellet, and brief sonication was used to resuspend the particles. Rinsing, pelleting, and resuspending of the particles was performed two more times with sterile distilled water, and finally the particles were resuspended in two milliliters of sterile distilled water. The particles were subdivided into 250-μl aliquots and stored frozen.

Preparation of particle-plasmid DNA association The stock of tungsten particles was sonicated briefly in a water bath sonicator (Branson Sonifier Model 450, 20% output, constant duty cycle) and 50 μl was transferred to a microfuge tube. Equimolar amounts of selectable and unselectable plasmid DNA were added to the particles for a final DNA amount of 0.1 to 10 μg in 10 μl total volume, and briefly sonicated. Preferably, l μg total DNA was used. Specifically, 3.5 μl of DP4810 (ubi _p :ubiint: :BAR: :pinll,, 5.6 kbp) plus 6.5 μl of DP6645 (fl2 _p : :fl2: :fl2,, 10.2 kbp), both at 0.1 μg/μl in TE buffer, were added to the particle suspension. Fifty microliters of sterile aqueous 2.5 M CaCl ₂ were added, and the mixture was briefly sonicated and vortexed. Twenty microliters of sterile aqueous 0.1 M spermidine were added and the mixture was briefly sonicated and vortexed. The mixture was incubated at room temperature for 20 minutes with intermittent brief sonication. The particle suspension was centrifuged, and the supernatant was removed. Two hundred fifty microliters of absolute ethanol were added to the pellet, followed by brief sonication. The suspension was pelleted, the supernatant was removed, and 60 μl of absolute ethanol were added. The suspension was sonicated briefly before loading the particle-DNA agglomeration onto macrocarriers.

Example 3. Extraction and characterization of protein -from transgenic seed

Embryos were hand-dissected from dry, mature kernels sampled from fully developed ears and endosperms were

pulverized to a fine meal with a ball mill. Alpha-zeins were extracted overnight in 70% (v/v) ethanol with constant shaking at 37°C. After centrifugation for 15 minutes at 12,000 rpm, the supernatant was collected, vacuum dried, and stored at 4°C until use. Total zeins and non-zein proteins were isolated according to Wallace et al . , Plant Physiol . 92: 191-96 (1990).

SDS-polyacrylamide gels (10 and 12.5%, w/v) and gradient gels (7.5-18%, w/v) were prepared according to Laemmli, Nature 227: 680-85 (1970), but the TRIS concentrations used in the resolving gel and running buffer were doubled. Protein samples were diluted in Laemmli sample buffer and boiled for 3 minutes before loading. Gradient and 12.5% gels were run at room temperature at a constant current until the dye front migrated throught the stacking gel, and then at 250 mA through the resolving gel. Gels were stained with Coomassie overnight, and destained in 40% (v/v) methanol and 10% (v/v) acetic acid for at least 8 hours. Immunoblotting analyses were used specifically to detect α-zeins in protein extracts. Protein extracts were separated by SDS-PAGE as described above, transferred to nitrocellulose filters, and treated with a rabbit anti- α-zein polyclonal antibody. Lending et al . , Protoplasms 143: 51-62 (1988). Goat anti-rabbit alkaline phsophatase conjugate was used for indirect detection of α-zein, as described by Knecht et al . , Anal . Biochem. 136: 180-84 (1984) . The 24 kDa α-zein was detected in the seed from the transgenic plants.