Plant promoter sequences and methods of use for same

Abstract

The invention discloses novel promoter sequences capable of expressing genes in plant cells. The promoters include engineered versions of the maize ubiquitin promoter to increase expression levels beyond those observed with the native ubiquitin promoter and alter the tissue preference. expression constructs, vectors, transgenic plants and methods are also disclosed.

Description

This application is a continuation of U.S. application Ser. No. 09/590,558, filed Jun. 9, 2000, now abandoned.

FIELD OF THE INVENTION

This invention relates generally to the field of plant molecular biology and in particular to engineered promoter sequences and their combined arrangement within a promoter region such that expression of an expression construct is enhanced in a plant cell.

BACKGROUND OF THE INVENTION

Gene expression encompasses a number of steps originating from the DNA template, ultimately to the final protein or protein product. Control and regulation of gene expression can occur through numerous mechanisms. The initiation of transcription of a gene is generally thought of as the predominant control of gene expression. Transcriptional controls (or promoters) are generally short sequences embedded in the 5′-flanking or upstream region of a transcribed gene. There are promoter sequences which affect gene expression in response to environmental stimuli, nutrient availability, or adverse conditions including heat shock, anaerobiosis or the presence of heavy metals. There are also DNA sequences which control gene expression during development, or in a tissue, or in an organ specific fashion, and, of course there are constitutive promoters.

Promoters contain the signals for RNA polymerase to begin transcription so that protein synthesis can proceed. DNA binding, nuclear, localized proteins interact specifically with these cognate promoter DNA sequences to promote the formation of the transcriptional complex and eventually initiate the gene expression process. The entire region containing all the ancillary elements affecting regulation or absolute levels of transcription may be comprised of less than 100 base pairs or as much as 1 kilobase pairs.

One of the most common sequence motifs present in the promoters of genes is the “TATA” element which resides upstream of the start of transcription. Promoters are also typically comprised of components which include a TATA box consensus sequence at about 35 base pairs 5′ relative to the transcription start site or cap site which is defined as +1. The TATA motif is the site where the TATA-binding-protein (TBP) as part of a complex of several polypeptides (TFIID complex) binds and productively interacts (directly or indirectly) with factors bound to other sequence elements of the promoter. This TFIID complex in turn recruits the RNA polymerase II complex to be positioned for the start of transcription generally 25 to 30 base pairs downstream of the TATA element and promotes elongation thus producing RNA molecules.

In most instances sequence elements other than the TATA motif are required for accurate transcription. Such elements are often located upstream of the TATA motif and a subset may have homology to the consensus sequence CCAAT.

Promoters are usually positioned 5′ or upstream relative to the start of the coding region of the corresponding gene, and the entire region containing all the ancillary elements affecting regulation or absolute levels of transcription may be comprised of less than 100 base pairs or as much as 1 kilobase pair.

A number of promoters which are active in plant cells have been described in the literature. These include nopaline synthase (NOS) and octopine synthase (OCS) promoters (which are carried on tumor inducing plasmids of Agrobacterium tumefaciens) The cauliflower mosaic virus (CaMV) 19S and 35S promoters, the light-inducible promoter from the small subunit of ribulose bisphosphate carboxylase (ssRUBICSO, a very abundant plant polypeptide), the alcohol dehydrogenase (AdhI and AdhII) promoters from maize, and the sucrose synthase promoter. All of these promoters have been used to create various types of DNA constructs which have been expressed in plants. (See for example PCT publication WO84/02913 Rogers, et al). Perhaps the most commonly used promoter is the 35S promoter of Cauliflower Mosaic Virus. The (CaMV) 35S promoter is a dicot virus promoter, however, it directs expression of genes introduced into protoplasts of both dicots and monocots. The 35S promoter is a very strong promoter and this accounts for its widespread use for high level expression of traits in transgenic plants. The CaMV35S promoter however has also demonstrated relatively low activity in several agriculturally significant graminaceous plants such as wheat.

The promoters of the maize genes encoding alcohol dehydrogenase, AdhI and AdhII, have also been widely used in plant cell transformations. Both genes are induced after the onset of anaerobiosis. Maize AdhI has been cloned and sequenced as has been AdhII. Formation of an AdhI chimeric gene, Adh-CAT comprising the AdhI promoter linked to the chloramphenicol acetyltransferase (CAT) coding sequences and nopaline synthase (NOS) 3′ signal caused CAT expression at approximately 4-fold higher levels at low oxygen concentrations than under control conditions. Sequence elements necessary for anaerobic induction of the ADH-CAT chimeric have also been identified. The existence of anaerobic regulatory element (ARE) between positions −140 and −99 of the maize AdhI promoter composed of at least two sequence elements at positions −133 to −124 and positions −113 to −99 both of which have found to be necessary and are sufficient for low oxygen expression of ADH-CAT gene activity. The Adh promoter however responds to anaerobiosis and is not a constitutive promoter, drastically limiting its effectiveness.

Yet another important promoter in plants is the maize ubiquitin promoter which is described in U.S. Pat. No. 5,510,474, to Quail et al. the disclosure of which is incorporated herein by reference

The 5′ heat shock consensus sequence is underlined. The 3′ heat shock consensus sequence is overlined. As can be seen, the overlap is a CTCGA 5-mer. According to the invention, novel promoters are designed which do not include two overlapping heat shock elements. Variants included, deletion of both heat shock elements, deletion of the 3′ element, deletion of the 5′ element, and removal of the overlap so that the two elements are adjacent.

A chart depicting the engineering in the heat shock region is below:

TABLE 1
Engineering of Ubi-1 promoter HSE
DNA		HSE	Trans-
con-		engineer-	genic
struct	DNA sequence1	ing	lines
PGN7062	CTGGACCCCTCTCGAGAGTTCCGCT	wild type	GSB
	(SEQ ID NO: 1)
PGN7547	-------------------------	HSEs	GSC
		deleted
PGN7565	CTGGACCCCTCTCGA----------	3′HSE	GSD
	(SEQ ID NO: 2)	deleted
PGN7583	----------CTCGAGAGTTCCGCT	5′HSE	GSE
	(SEQ ID NO: 3)	deleted
PGN7600	CTGGACCCCTCTCGACTCGAGAGTTC	HSEs	GSF
	CGCT (SEQ ID NO: 4)	adjacent
PGN8926	3x(GACACGTAGAATGAGTCATCAC)	HSEs	GSG
	(SEQ ID NO: 5)	replaced
		by Ps1
		trimer
1The 5′ HSE is in bold type and the 3′ HSE is underlined.

In yet another embodiment a transcription binding factor can be added in the engineered heat shock element region, to add in transcription of the sequences following the promoter. Such factors are known to those of skill in the art and include but are not limited to: the prolactin seed specific binding factor: (dePater, S., et al. (1994), “A 22-bp fragment of the pea lectin promoter containing essential TGAC-like motifs confers seed-specific gene expression”, Plant Cell 5:877-886dePater, S., et al. (1996), “The 22 bp W1 element in the pea lectin promoter is necessary and, as a multimer, sufficient for high gene expression in tobacco seeds”, Plant Mol. Biol. 32:515-523), and the basic domain/leucine zipper proteins TGA1a and Opaque-2 can bind this sequence in vitro (dePater, S., et al. (1994), “bZIP proteins bind to a palindromic sequence without and ACGT core located in a seed-specific element of the pea lectin promoter”, Plant J. 6:133-140). A table of transcription factors which may be used according to the invention follows:

TABLE A
			5′	3′
			extent	extent
Species	Factor	Target Gene	of site	of site	Site Sequence
Arabidopsis	EBP	Pathogenesis-related	−207	−192	atGGCTctta (SEQ ID NO: 6)
thaliana		protein 1b
Arabidopsis	HY5	Ribulose-1,	−241	−230	CTTCCACGTGGCA
thaliana		5-biphosphate			(SEQ ID NO: 7)
		carboxylase
Hordeum	BLZ-1	B-hordein	−252	−220	acatgtaaagtgaataagGTGAGTCA
vulgare					(SEQ ID NO: 8)
Hordeum	Gamyb	High-pI	−149	−128	ggccgaTAACAAACtccggccg
vulgare		alpha-amylas			(SEQ ID NO: 9)
an Oryza	RF2a	Rice tungro	−53	−39	CCAGTGTGCCCCTGG
sativa virus		bacilliform virus			(SEQ ID NO: 10)
		promoter
Phaseolus	ROM1	Phytohemagglutinin	−207	−199	GCCACGTCA
vulgare
Pisum	GT-1	Ribulose-1,	−257	−245	GATTTACACT (SEQ ID NO: 11)
sativum		5-biphosphate
		carboxylase
Triticum	SPA	Low molecular weight	−256	−241	taaGGTGAGTCATata
aestivum		glutenin-1D1			(SEQ ID NO: 12)
Zea mays	Dof2	C4-type	−774	−765	ATACTTTTC (SEQ ID NO: 13)
		phosphoenolpyruvate
		carboxylase
Zea mays	Opaque-2	22-kD Zein	−305	−288	tgTCATTCCACGTAGAtg
					(SEQ ID NO: 14)

Transgenic Techniques Overview

Likewise, by means of the present invention, agronomic genes in combination with the promoters of the invention can be expressed in transformed plants. Production of a genetically engineered plant tissue either expressing or inhibiting expression of a structural gene combines the teachings of the present disclosure with a variety of techniques and expedients known in the art. In most instances, alternate expedients exist for each stage of the overall process. The choice of expedients depends on the variables such as the plasmid vector system chosen for the cloning and introduction of the recombinant DNA molecule, the plant species to be engineered, the particular structural gene, promoter elements and upstream elements used. Persons skilled in the art are able to select and use appropriate alternatives to achieve functionality. Culture conditions for expressing desired structural genes and cultured cells are known in the art. Also as known in the art, a number of both monocotyledonous and dicotyledonous plant species are transformable and regenerable such that whole plants containing and expressing desired genes under regulatory control of the promoter molecules according to the invention may be obtained. As is known to those of skill in the art, expression in transformed plants may be tissue specific and/or specific to certain developmental stages. Truncated promoter selection and structural gene selection are other parameters which may be optimized to achieve desired plant expression or inhibition as is known to those of skill in the art and taught herein.

The following is a non-limiting general overview of Molecular biology techniques which may be used in performing the methods of the invention.

Structural Gene

Likewise, by means of the present invention, heterologous nucleotide sequences can be expressed in transformed plants. More particularly, plants can be genetically engineered to express various phenotypes of agronomic interest.

Exemplary genes include but are not limited to: plant disease resistance genes, (Martin et al., Science 262: 1432 (1993) (tomato Pto gene for resistance to Pseudomonas syringae pv. tomato encodes a protein kinase)); a Bacillus thuringiensis protein, (Geiser et al., Gene 48: 109 (1986); a lectin, (Van Damme et al., Plant Molec. Biol. 24: 25 (1994)); a vitamin-binding protein, (such as avidin. see PCT application US93/06487); an enzyme inhibitor, (Abe et al., J. Biol. Chem. 262: 16793 (1987)); an insect-specific hormone or pheromone, (see, for example, Hammock et al., Nature 344: 458 (1990)); an insect-specific peptide or neuropeptide, (Regan, J. Biol. Chem. 269: 9 (1994)); an insect-specific venom, (Pang et al., Gene 116: 165 (1992); an enzyme responsible for an hyperaccumulation of a monterpene; an enzyme involved in the engineering, including the post-translational engineering, of a biologically active molecule; for example, a glycolytic enzyme, a proteolytic enzyme; (See PCT application WO 93/02197); a molecule that stimulates signal transduction, (for example, Botella et al., Plant Molec. Biol. 24: 757 (1994)); a hydrophobic moment peptide, (PCT application WO 95/16776); a membrane permease, (Jaynes et al., Plant Sci. 89: 43 (1993)); a viral-invasive protein or a complex toxin derived therefrom, (Beachy et al., Ann. Rev. Phytopathol.28: 451 (1990)); (Taylor et al., Abstract #497, SEVENTH IN'L SYMPOSIUM ON MOLECULAR PLANT-MICROBE INTERACTIONS (Edinburgh, Scotland, 1994)); a virus-specific antibody, (Tavladoraki et al., Nature 366: 469 (1993)); a developmental-arrestive protein produced in nature by a pathogen or a parasite, (Lamb et al., Bio/Technology 10: 1436 (1992)); a developmental-arrestive protein produced in nature by a plant, (Logemann et al., Bio/Technology 10: 305 (1992)); a herbicide that inhibits the growing point or meristem, such as an imidazalinone or a sulfonylurea, (Lee et al.,EMBO J. 7: 1241 (1988)); Glyphosate (resistance imparted by mutant 5-enolpyruvl-3-phosphikimate synthase (EPSP) and aroA genes, respectively) (U.S. Pat. No. 4,940,835); a herbicide that inhibits photosynthesis, such as a triazine (psbA and gs+ genes) and a benzonitrile (nitrilase gene). (Przibilla et al., Plant Cell 3: 169 (1991)); Engineered fatty acid metabolism, for example, by transforming a plant with an antisense gene of stearoyl-ACP desaturase to increase stearic acid content of the plant. See Knultzon et al., Proc. Natl. Acad. Sci. USA 89: 2624 (1992); decreased phytate content, (Van Hartingsveldt et al., Gene 127: 87 (1993)); engineered carbohydrate composition, for example, by transforming plants with a gene coding for an enzyme that alters the branching pattern of starch. (See Shiroza et al., J. Bacteriol. 170: 810 (1988)); genes that controls cell proliferation and growth of the embryo and/or endosperm such as cell cycle regulators (Bogre L et al., “Regulation of cell division and the cytoskeleton by mitogen-activated protein kinases in higher plants.” Results Probl Cell Differ 27:95-117 (2000).

Promoters

The promoters disclosed herein may be used in conjunction with naturally occurring flanking coding or transcribed sequences of the desired heterologous nucleotide sequence or structural gene or with any other coding or transcribed sequence that is critical to structural gene formation and/or function.

It may also be desirable to include some intron sequences in the promoter constructs since the inclusion of intron sequences in the coding region may result in enhanced expression and specificity. Thus, it may be advantageous to join the DNA sequences to be expressed to a promoter sequence that contains the first intron and exon sequences of a polypeptide which is unique to cells/tissues of a plant critical to seed specific Structural formation and/or function.

Additionally, regions of one promoter may be joined to regions from a different promoter in order to obtain the desired promoter activity resulting in a chimeric promoter. Synthetic promoters which regulate gene expression may also be used.

The expression system may be further optimized by employing supplemental elements such as transcription terminators and/or enhancer elements.

Other Regulatory Elements

In addition to a promoter sequence, an expression cassette or construct should also contain a transcription termination region downstream of the structural gene to provide for efficient termination. The termination region or polyadenylation signal may be obtained from the same gene as the promoter sequence or may be obtained from different gene. Polyadenylation sequences include, but are not limited to the Agrobacterium octopine synthase signal (Gielen et al., EMBO J. (1984) 3:835-846) or the nopaline synthase signal (Depicker et al., Mol. and Appl. Genet. (1982) 1:561-573), or pin II the proteinase inhibitor II gene from potato.

Marker Genes

Recombinant DNA molecules containing any of the DNA sequences and promoters described herein may additionally contain selection marker genes which encode a selection gene product which confer on a plant cell resistance to a chemical agent or physiological stress, or confer a distinguishable phenotypic characteristic to the cells such that plant cells transformed with the recombinant DNA molecule may be easily selected using a selective agent. One such selection marker gene is neomycin phosphotransferase (NPT II) which confers resistance to kanamycin and the antibiotic G-418. Cells transformed with this selection marker gene may be selected for by assaying for the presence in vitro of phosphorylation of kanamycin using techniques described in the literature or by testing for the presence of the mRNA coding for the NPT II gene by Northern blot analysis of RNA from the tissue of the transformed plant. Polymerase chain reactions are also used to identify the presence of a transgene or expression using reverse transcriptase PCR amplification to monitor expression and PCR on genomic DNA. Other commonly used selection markers include the ampicillin resistance gene, the tetracycline resistance gene and the hygromycin resistance gene. Transformed plant cells thus selected can be induced to differentiate into plant structures which will eventually yield whole plants. It is to be understood that a selection marker gene may also be native to a plant.

Transformation

In accordance with the present invention, a transgenic plant is produced that contains a DNA molecule, comprised of elements as described above, integrated into its genome so that the plant expresses a heterologous gene-encoding DNA sequence. In order to create such a transgenic plant, the expression vectors containing the gene 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 introduced into intact tissues. 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, Glick et al (eds) pp. 67-68 (CRC Press 1993) and by Phillips et al, “Cell/Tissue Culture and In Vitro Manipulation” in Corn and Corn Improvement 3d Edit. Sprague et al (eds) pp. 345-387 (American Soc. Of Agronomy 1988). The selectable marker incorporated in the DNA molecule allows for selection of transformants.

Methods for introducing expression vectors into plant tissue available to one skilled in the art are varied and will depend on the plant selected. Procedures for transforming a wide variety of plant species are well known and described throughout the literature. See, for example, Miki et al, supra; Klein et al, Bio/Technology 10:268 (1992); and Weisinger et al., Ann. Rev. Genet. 22: 421-477 (1988). For example, the DNA construct may be introduced into the genomic DNA of the plant cell using techniques such as microprojectile-mediated delivery, Klein et al., Nature 327: 70-73 (1987); electroporation, Fromm et al., Proc. Natl. Acad. Sci. 82: 5824 (1985); polyethylene glycol (PEG) precipitation, Paszkowski et al., Embo J. 3: 2717-2722 (1984); direct gene transfer, WO 85/01856 and EP No. 0 275 069; in vivo protoplast transformation, U.S. Pat. No. 4,684,611; and microinjection of plant cell protoplasts or embryogenic callus. Crossway, Mol. Gen. Genetics 202:179-185 (1985). Co-cultivation of plant tissue with Agrobacterium tumefaciens is another option, where the DNA constructs are placed into a binary vector system. Ishida et al., “High Efficiency Transformation of Maize (Zea mays L.) Mediated by Agrobacterium tumefaciens” Nature Biotechnology 14:745-750 (1996). The virulence functions of the Agrobacterium tumefaciens host will direct the insertion of the construct into the plant cell DNA when the cell is infected by the bacteria. See, for example Horsch et al., Science 233: 496-498 (1984), and Fraley et al., Proc. Natl. Acad. Sci. 80: 4803 (1983).

Standard methods for transformation of canola are described by Moloney et al. “High Efficiency Transformation of Brassica napus Using Agrobacterium Vectors” Plant Cell Reports 8:238-242 (1989). Corn transformation is described by Fromm et al, Bio/Technology 8:833 (1990) and Gordon-Kamm et al, supra. Agrobacterium is primarily used in dicots, but certain monocots such as maize can be transformed by Agrobacterium. U.S. Pat. No. 5,550,318. Rice transformation is described by Hiei et al., “Efficient Transformation of Rice (Oryza sativs L.) Mediated by Agrobacterium and Sequence Analysis of the Boundaries of the T-DNA” The Plant Journal 6(2): 271-282 (1994), Christou et al, Trends in Biotechnology 10:239 (1992) and Lee et al, Proc. Nat'l Acad. Sci. USA 88:6389 (1991). Wheat can be transformed by techniques similar to those used for transforming corn or rice. Sorghum transformation is described by Casas et al, supra and by Wan et al, Plant Physiology 104:37 (1994). Soybean transformation is described in a number of publications, including U.S. Pat. No. 5,015,580.

In one preferred method, the Agrobacterium transformation methods of Ishida supra and also described in U.S. Pat. No. 5,591,616, are generally followed, with engineering that the inventors have found improve the number of transformants obtained. The Ishida method uses the A188 variety of maize that produces Type I callus in culture. In one preferred embodiment the High II maize line is used which initiates Type II embryogenic callus in culture. While Ishida recommends selection on phosphinothricin when using the bar or PAT gene for selection, another preferred embodiment provides for use of bialaphos instead.

The bacterial strain used in the Ishida protocol is LBA4404 with the 40 kb super binary plasmid containing three vir loci from the hypervirulent A281 strain. The plasmid has resistance to tetracycline. The cloning vector cointegrates with the super binary plasmid. Since the cloning vector has an E. coli specific replication origin, it cannot survive in Agrobacterium without cointegrating with the super binary plasmid. Since the LBA4404 strain is not highly virulent, and has limited application without the super binary plasmid, the inventors have found in yet another embodiment that the EHA101 strain is preferred. It is a disarmed helper strain derived from the hypervirulent A281 strain. The cointegrated super binary/cloning vector from the LBA4404 parent is isolated and electroporated into EHA 101, selecting for spectinomycin resistance. The plasmid is isolated to assure that the EHA101 contains the plasmid.

Further, the Ishida protocol as described provides for growing fresh culture of the Agrobacterium on plates, scraping the bacteria from the plates, and resuspending in the co-culture medium as stated in the '616 patent for incubation with the maize embryos. This medium includes 4.3 g MS salts, 0.5 mg nicotinic acid, 0.5 mg pyridoxine hydrochloride, 1.0 ml thiamine hydrochloride, casamino acids, 1.5 mg 2,4-D, 68.5g sucrose and 36 g glucose, all at a pH of 5.8. In a further preferred method, the bacteria are grown overnight in a 1 ml culture, then a fresh 10 ml culture re-inoculated the next day when transformation is to occur. The bacteria grow into log phase, and are harvested at a density of no more than OD600=0.5 and is preferably between 0.2 and 0.5. The bacteria are then centrifuged to remove the media and resuspended in the co-culture medium. Since Hi II is used, medium preferred for Hi II is used. This medium is described in considerable detail by Armstrong, C. I. and Green C. E. “Establishment and maintenance of friable, embryogenic maize callus and involvement of L-proline” Planta (1985) 154:207-214. The resuspension medium is the same as that described above. All further Hi II media are as described in Armstrong et al. The result is redifferentiation of the plant cells and regeneration into a plant. Redifferentiation is sometimes referred to as dedifferentiation, but the former term more accurately describes the process where the cell begins with a form and identity, is placed on a medium in which it loses that identity, and becomes “reprogrammed” to have a new identity. Thus the scutellum cells become embryogenic callus.

It is often desirable to have the DNA sequence in homozygous state which may require more than one transformation event to create a parental line, requiring transformation with a first and second recombinant DNA molecule both of which encode the same gene product. It is further contemplated in some of the embodiments of the process of the invention that a plant cell be transformed with a recombinant DNA molecule containing at least two DNA sequences or be transformed with more than one recombinant DNA molecule. The DNA sequences or recombinant DNA molecules in such embodiments may be physically linked, by being in the same vector, or physically separate on different vectors. A cell may be simultaneously transformed with more than one vector provided that each vector has a unique selection marker gene. Alternatively, a cell may be transformed with more than one vector sequentially allowing an intermediate regeneration step after transformation with the first vector. Further, it may be possible to perform a sexual cross between individual plants or plant lines containing different DNA sequences or recombinant DNA molecules preferably the DNA sequences or the recombinant molecules are linked or located on the same chromosome, and then selecting from the progeny of the cross plants containing both DNA sequences or recombinant DNA molecules.

Expression of recombinant DNA molecules containing the DNA sequences and promoters described herein in transformed plant cells may be monitored using Northern blot techniques and/or Southern blot techniques or PCR-based methods known to those of skill in the art.

A large number of plants have been shown capable of regeneration from transformed individual cells to obtain transgenic whole plants. Corn has long been a successful plant transformation recipient. Fromm, et al., Bio Technology, 8:33 (1990). Others are as follows. For example, regeneration has been shown for dicots as follows: apple, Malus pumila (James et al., Plant Cell Reports (1989) 7:658); blackberry, Rubus, Blackberry/raspberry hybrid, Rubus, red raspberry, Rubus (Graham et al., Plant Cell, Tissue and Organ Culture (1990) 20:35); carrot, Daucus carota (Thomas et al., Plant Cell Reports (1989) 8:354; Wurtele and Bulka, Plant Science (1989) 61:253); cauliflower, Brassica oleracea (Srivastava et al., Plant Cell Reports (1988) 7:504); celery, Apium graveolens (Catlin et al., Plant Cell Reports (1988) 7:100); cucumber, Cucumis sativus (Trulson et al., Theor. Appl. Genet. (1986) 73:11); eggplant, Solanum melonoena (Guri and Sink, J. Plant Physiol. (1988) 133:52) lettuce, Lactuca sativa (Michelmore et al., Plant Cell Reports (1987) 6:439); potato, Solanum tuberosum (Sheerman and Bevan, Plant Cell Reports (1988) 7:13); rape, Brassica napus (Radke et al., Theor. Appl. Genet. (1988) 75:685; Moloney et al., Plant Cell Reports (1989) 8:238); soybean (wild), Glycine canescens (Rech et al., Plant Cell Reports (1989) 8:33); strawberry, Fragaria ×ananassa (Nehra et al., Plant Cell Reports (1990) 9:10; tomato, Lycopersicon esculentum (McCormick et al., Plant Cell Reports (1986) 5:81); walnut, Juglans regia (McGranahan et al., Plant Cell Reports (1990) 8:512); melon, Cucumis melo (Fang et al., 86th Annual Meeting of the American Society for Horticultural Science Hort. Science (1989) 24:89); grape, Vitis vinifera (Colby et al., Symposium on Plant Gene Transfer, UCLA Symposia on Molecular and Cellular Biology J Cell Biochem Suppl (1989) 13D:255; mango, Mangifera indica (Mathews, et al., symposium on Plant Gene Transfer, UCLA Symposia on Molecular and Cellular Biology J Cell Biochem Suppl (1989) 13D:264); and for the following monocots: rice, Oryza sativa (Shimamoto et al., Nature (1989) 338:274); rye, Secale cereale (de la Pena et al., Nature (1987) 325:274); maize, (Rhodes et al., Science (1988) 240:204).

In addition, regeneration of whole plants from cells (not necessarily transformed) has been observed in apricot, Prunus armeniaca (Pieterse, Plant Cell Tissue and Organ Culture (1989) 19:175); asparagus, Asparagus officinalis (Elmer et al., J. Amer. Soc. Hort. Sci. (1989) 114:1019); Banana, hybrid Musa (Escalant and Teisson, Plant Cell Reports (1989) 7:665); bean, Phaseolus vulgaris (McClean and Grafton, Plant Science (1989) 60:117); cherry, hybrid Prunus (Ochatt et al., Plant Cell Reports (1988) 7:393); grape, Vitis vinifera (Matsuta and Hirabayashi, Plant Cell Reports, (1989) 7:684; mango, Mangifera indica (DeWald et al., J Amer Soc Hort Sci (1989) 114:712); melon, Cucumis melo (Moreno et al., Plant Sci letters (1985) 34:195); ochra, Abelmoschus esculentus (Roy and Mangat, Plant Science (1989) 60:77; Dirks and van Buggenum, Plant Cell Reports (1989) 7:626); onion, hybrid Allium (Lu et al., Plant Cell Reports (1989) 7:696); orange, Citrus sinensis (Hidaka and Kajikura, Scientia Horiculturae (1988) 34:85); papaya, Carrica papaya (Litz and Conover, Plant Sci Letters (1982) 26:153); peach, Prunus persica and plum, Prunus domestica (Mante et al., Plant Cell Tissue and Organ Culture (989) 19:1); pear, Pyrus communis (Chevreau et al., Plant Cell Reports (1988) 7:688; Ochatt and Power, Plant Cell Reports (1989) 7:587); pineapple, Ananas comosus (DeWald et al., Plant Cell Reports (1988) 7:535); watermelon, Citrullus vulgaris (Srivastava et al., Plant Cell Reports (1989) 8:300); wheat, Triticum aestivum (Redway et al., Plant Cell Reports (1990) 8:714).

The regenerated plants are transferred to standard soil conditions and cultivated in a conventional manner. After the expression or inhibition cassette is stably incorporated into regenerated transgenic plants, it can be transferred to other plants by sexual crossing. Any of a number of standard breeding techniques can be used, depending upon the species to be crossed.

It may be useful to generate a number of individual transformed plants with any recombinant construct in order to recover plants free from any position effects. It may also be preferable to select plants that contain more than one copy of the introduced recombinant DNA molecule such that high levels of expression of the recombinant molecule are obtained.

According to a preferred embodiment, the transgenic plant provided for commercial production of foreign protein is maize. In another preferred embodiment, the biomass of interest is seed. For the relatively small number of transgenic plants that show higher levels of expression, a genetic map can be generated, primarily via conventional Restriction Fragment Length Polymorphisms (RFLP), Polymerase Chain Reaction (PCR) analysis, and Simple Sequence Repeats (SSR) which identifies the approximate chromosomal location of the integrated DNA molecule. For exemplary methodologies in this regard, see Glick and Thompson, METHODS IN PLANT MOLECULAR BIOLOGY AND BIOTECHNOLOGY 269-284 (CRC Press, Boca Raton,1993) . Map information concerning chromosomal location is useful for proprietary protection of a subject transgenic plant. If unauthorized propagation is undertaken and crosses made with other germplasm, the map of the integration region can be compared to similar maps for suspect plants, to determine if the latter have a common parentage with the subject plant. Map comparisons would involve hybridizations, RFLP, PCR, SSR and sequencing, all of which are conventional techniques.

As indicated above, it may be desirable to produce plant lines which are homozygous for a particular gene. In some species this is accomplished rather easily by the use of anther culture or isolated microspore culture. This is especially true for the oil seed crop Brassica napus (Keller and Armstrong, Z. flanzenzucht 80:100-108, 1978). By using these techniques, it is possible to produce a haploid line that carries the inserted gene and then to double the chromosome number either spontaneously or by the use of cochicine. This gives rise to a plant that is homozygous for the inserted gene, which can be easily assayed for if the inserted gene carries with it a suitable selection marker gene for detection of plants carrying that gene. Alternatively, plants may be self-fertilized, leading to the production of a mixture of seed that consists of, in the simplest case, three types, homozygous (25%), heterozygous (50%) and null (25%) for the inserted gene. Although it is relatively easy to score null plants from those that contain the gene, it is possible in practice to score the homozygous from heterozygous plants by Southern blot analysis in which careful attention is paid to the loading of exactly equivalent amounts of DNA from the mixed population, and scoring heterozygotes by the intensity of the signal from a probe specific for the inserted gene. It is advisable to verify the results of the southern blot analysis by allowing each independent transformant to self-fertilize, since additional evidence for homozygosity can be obtained by the simple fact that if the plant was homozygous for the inserted gene, all of the subsequent plants from the selfed seed will contain the gene, while if the plant was heterozygous for the gene, the generation grown from the selfed seed will contain null plants. Therefore, with simple selfing one can easily select homozygous plant lines that can also be confirmed by southern blot analysis.

Creation of homozygous parental lines makes possible the production of hybrid plants and seeds which will contain a engineered protein component. Transgenic homozygous parental lines are maintained with each parent containing either the first or second recombinant DNA sequence operably linked to a promoter. Also incorporated in this scheme are the advantages of growing a hybrid crop, including the combining of more valuable traits and hybrid vigor.

The following examples serve to better illustrate the invention described herein and are not intended to limit the invention in any way. All references cited herein are hereby expressly incorporated to this document in their entirety by reference.

EXAMPLES

Methods

Construction of Ubi-1 Promoter Variants

The DNA construct PHP8904 (Pioneer Hi-Bred; Johnston, Iowa), contains the GUS reporter gene positioned 3′ to approximately 0.9 kb of 5′ flanking sequence of maize Ubi-1, plus the Ubi-1 5′ untranslated leader sequence and first intron. The potato proteinase inhibitor II transcription terminator region is present 3′ of GUS. PHP8904 also carries right and left border sequences of an Agrobacterium tumefaciens Ti plasmid, bacterial antibiotic resistance and origin of replication sequences, and the bar gene of Streptomyces hygroscopicus, conferring resistance to the herbicide bialaphos. The construct PGN7062 is essentially identical to 8904, except that the GUS reporter gene includes sequences encoding six C-terminal histidine residues. All subsequent constructs are similar to PGN7062 but have engineering in Ubi-1 5′ flanking sequences (Table 1). For each Ubi-1 5′ flanking sequence variant, a series of oligonucleotides were generated that together span the putative heat shock elements. These oligonucleotides were assembled and the sequences amplified by the polymerase chain reaction. The DNA fragments were introduced into the cloning vector pCR2.1 (Invitrogen; Carlsbad, Calif.). SalI-BglII restriction enzyme generated DNA fragments spanning the engineered HSEs were isolated from the pCR2.1 based plasmids and were transferred into an intermediate PGEM (Promega Corporation; Madison, Wis.) based plasmid, PGN5796, so replacing corresponding wild type Ubi-1 5′ flanking sequence. HindIII-NheI restriction enzyme generated DNA fragments, spanning the entire Ubi-1 5′ flanking sequence and 5′ untranslated region plus part of the first intron, were then transferred into PGN7062, so replacing corresponding wild type Ubi-1 sequence.

TABLE 1
Engineered Ubi-1 promoter HSE
DNA			Trans-
con-		descrip-	genic
struct	DNA sequence1	tion	lines
PGN7062	CTGGACCCCTCTCGAGAGTTCCGCT	wild	GSB
	(SEQ ID NO: 1)	type
PGN7547	---------------------------	HSEs	GSC
		deleted
PGN7565	CTGGACCCCTCTCGA----------	3′ HSE	GSD
	(SEQ ID NO: 2)	deleted
PGN7583	----------CTCGAGAGTTCCGCT	5′ HSE	GSE
	(SEQ ID NO: 3)	deleted
PGN7600	CTGGACCCCTCTCGACTCGAGAGTTCC	HSEs	GSF
	GCT (SEQ ID NO: 4)	adjacent
PGN8926	3x(GACACGTAGAATGACTCATCAC)	HSEs	GSG
	(SEQ ID NO: 5)	replaced
		by Ps1
		trimer
1The 5′ HSE is in bold type and the 3′ HSE is underlined.

Transient Transformation

Transient transformations using Agrobacterium tumefaciens were performed using sonication-assisted Agrobacterium transformation as described by Trick and Finer (Trick, H. N. et al. (1997) SAAT: Sonication assisted Agrobacterium-mediated transformation”, Transgenic Res. 6:329-336). Ten immature zygotic embryos per tube were sonicated in the presence of Agrobacterium tumefaciens EHA 101 (pSB111) at an O. D._{600 nm} of 0.5 for 30 s, were placed onto co-cultivation medium and were incubated for 5 days. Embryos were stained for 24 hours with 5 mgml⁻¹ X-gluC (5-bromo-4-chloro-3-indolyl-β-D-glucoronic acid: cyclohexyl ammonium salt) (Inalco; Milan, Italy) dissolved in Jefferson's buffer (Jefferson, R. A. (1987), “Assaying chimeric genes in plants: the GUS gene fusion system”, Plant Molec. Biol. Reporter 5:387-405). They were subsequently transferred to 70% ethanol.

Transformation, Tissue Culture and Plant Growth

The procedure for stable transformation followed a engineered version of Ishida et al. (Ishida, Y. et al. (1996), “High efficiency transformation of maize (Zea mays L.) mediated by Agrobacterium tumefaciens”, Nature Biotech 14:745-750) and Armstrong and Green (Armstrong, C. L., et al. (1985) “Establishment and maintenance of friable, embryogenic maize callus and the involvement of L-proline”, Planta 164:207-214). Transformation and regeneration media are described in Table 2. Immature zygotic embryos were isolated from Hi-II maize kernels at 12 days after pollination and were transformed with Agrobacterium tumefaciens strain EHA 101 containing the engineered Ubi-1 variant constructs. For Agrobacterium infection, bacteria were grown overnight in YEP liquid medium supplemented with antibiotic. Agrobacterium were then re-inoculated into YEP supplemented with 100 mgl⁻¹ kanamycin and 100 mgl⁻¹ spectinomicin and were grown to an OD_{550 nm} of 0.4-0.6. The Agrobacterium culture was centrifuged to remove the media and the pellet was resuspended in inoculation medium. Immature zygotic embryos were washed with inoculation medium and immersed in the Agrobacterium solution, vortexed for 30 s, incubated for 5 minutes and plated and co-cultivated for 4 days on solid co-culture medium. Embryos were transferred for three days onto non-selective medium supplemented with 100 mg/l carbenicillin, and then subcultured to Bialaphos selection medium and subsequently subcultured every two weeks. Embryogenic tissue (events) proliferating on selection media were excised and cultured on the same medium for proliferation for four weeks and were then subcultured onto regeneration medium for three weeks to allow embryo formation. Embryos were picked and transferred to germination medium for one to two weeks with light at 28° C. Plants that regenerated were transferred to tubes for root and shoot elongation. Multiple T0 plants were regenerated from embryogenic tissues that were selected on Bialaphos and these were transferred to a greenhouse. T0 plants were crossed with elite inbred lines to produce T1 seeds. For analysis of T1 leaves, T1 seeds were germinated in a greenhouse and were leaf painted with a 1% active ingredient of Finale® for selection of transformed plants. Leaf samples were collected three weeks after germination.

Preparation of Plant Extracts

For seed extracts, individual dry seeds were pulverized using a hammer and extracted with 500 μl of lysis buffer (50 mM sodium phosphate pH 7.0, 1 mM EDTA, 10 mM β-mercaptoethanol). Samples were placed in extraction tubes, each with a ball bearing added, in a Beckman rack and were homogenized in a high speed shaker for 20 s. Samples were centrifuged, and the supernatants recovered and stored on ice prior to analysis. For leaf extracts, small portions of the ends of leaves were cut off and pulverized under liquid nitrogen. Weights were recorded and extractions completed using lysis buffer at 10 μl per mg of sample. For callus extracts, samples were extracted using lysis buffer at 1 μl per mg of callus tissue. Protein concentrations were determined according Bradford (1976).

β-glucuronidase Activity Assay

GUS assays were performed as described by Jefferson (1987, supra). Total soluble protein (1 μg) was incubated in 100 μl of lysis buffer and the reaction initiated with 25 μl of 5 mM 4-methylumbelliferyl β-D-glucuronide (MUG, Sigma M-9130). The reaction was incubated for up to 20 minutes at 37° C. At specific time points 25 μl volumes of the reaction mixture were transferred to a Dynatech Microfluor reading plate that had 175 μl of stop buffer (0.2M Na₂CO₃) per well. Fluorescence was measured at an excitation wavelength of 360 nm and an emission wavelength of 460 nm on a Microplate Fluorometer (Cambridge Technologies 7625). GUS protein levels were then calculated by comparison to a standard curve of 1 -100 μM 4-methylumbelliferyl (MU, Sigma M-1508).

Results

Conserved Ubi-1 Promoter Sequences are not Required for Transient Expression in Maize Embryos

To investigate whether engineered versions of the maize Ubi-1 promoter would facilitate high levels of constitutive expression, we generated a series of fusions of native or engineered Ubi-1 sequences to the GUS reporter gene for introduction into plants. The putative HSEs of the Ubi-1 promoter were removed, their relative spacing was altered or they were substituted with a trimer of a seed specific element from the promoter of the pea lectin gene Ps1 (Table 1).

The promoter variants were first assessed in a transient transformation system. The DNA constructs were introduced into zygotic embryos of maize and GUS activity was detected qualitatively by histochemical staining.

TABLE 2

                        Mean GUS expression score
Transient transformants (relative value)

GSB                     2.1
GSC                     2.0
GSD                     1.7
GSE                     2.5
GSF                     2.1
GSG                     2.2
promoterless GUS        0.0
no vector               0.0

*Score system: 3 = high, 2 = medium; 1 = low; 0 = nothing

In all cases, GUS is synthesized, indicating that none of the engineering to the Ubi-1 promoter knock out expression. However, embryos transiently transformed with PGN8926 produce much less GUS than those transformed with the other constructs.

Engineering of Conserved Ubi-1 Promoter Sequences can Increase Expression in Stable Transformed Lines of Maize

To more accurately assess the engineered Ubi-1 promoter variants, stable transformed lines were developed. The series of GUS fusions were introduced into zygotic embryos of maize to generate stable transformation events. Multiple seedlings were regenerated from embryogenic callus tissue of each event to give transformed lines, and seedlings matured and flowered to generate T1 seeds. GUS activity was determined in embryogenic callus tissue, leaves of seedlings regenerated from tissue culture and T1 seeds. The native Ubi-1 5′ flanking sequence and the promoter variants of Ubi-1 all drive GUS expression in each tissue type, but levels of GUS are much lower in embryogenic callus and in leaf tissue than in seeds.

Among plants derived from any specific transformation event, considerable variation in the level of GUS expression exists in leaves of regenerated seedlings and in T1 seeds. In addition, GUS expression in embryogenic callus tissue, leaves of regenerated seedlings and T1 seeds varies between different transformation events.

However, focusing on T1 seed, which is the preferred site of expression for the commercial production of foreign proteins in corn, there are significant differences in mean levels of expression between transformed lines carrying the engineered promoters. GSD and GSG lines have levels of GUS expression similar to the control GSB line, but surprisingly, GSC, GSE and GSF lines have elevated expression levels (FIG. 1C). A ranking of GUS expression levels in T1 seed between lines transformed with the promoter variants is similar whether mean or highest recorded expression levels are considered (FIG. 2).

Ubi-1 Promoter Variants Drive Constitutive Expression but Have Tissue Preferences in the Kernel

Maize Ubi-1 is constitutively expressed and the Ubi-1 promoter can drive the constitutive expression of reporter genes in transgenic plants (Christensen, A. H., et al. (1992), “Maize polyubiquitin genes: structure, thermal perturbation of expression and transcript splicing, and promoter activity following transfer to protoplasts by electroporation”, Plant Mol. Biol. 18:675-689; Takimoto, I., et al. (1994), “Non-systemic expression of a stress-responsive maize polyubiquitin gene (Ubi-1) in transgenic rice plants”, Plant Mol. Biol. 26:1007-1012; Christensen, A. H., et al. (1996), “Ubiquitin promoter-based vectors for high-level expression of selectable and/or screenable marker genes in monocotyledonous plants”, Transgenic Res. 5:213-218). The engineered Ubi-1 promoter variants generated here drive expression of GUS in embryogenic callus tissue and leaves of seedlings regenerated from tissue culture. To examine whether the promoter variants cause constitutive expression in plants germinated from seed, a selection of transformed lines that express GUS at a high level in T1 seeds were analyzed. GUS activity was determined in leaf tissue of developing T1 seedlings and was compared to the activity that had been recorded for T1 seeds (FIG. 3). GUS was detected in leaves of transformed lines carrying all engineered Ubi-1 promoter variants, but expression was much lower than in seeds. Due to small selected sample sizes, there is considerable variation in the expression data among lines carrying each engineered promoter variant. However, the ranking of GUS expression in leaf tissue among the variants reflects the ranking in seed, except that in GSC lines expression ranks higher in seed than in leaf tissue.

The activity of the Ubi-1 promoter variants was also assessed visually in various tissues. Selected T1 kernels were either directly analyzed by cutting into sections and staining for GUS activity, or were germinated to generate seedlings of which root and leaf tissues were analyzed. GUS activity is observed in leaves and roots with all transformed lines, and in both tissue types is highest for GSE (5′ HSE deleted) and GSF (HSE's adjacent) lines. In kernels of GSB lines GUS expression is higher in the embryo than in the endosperm. In transformed lines carrying the Ubi-1 promoter variants the distribution of GUS activity is more uniform across the seed, indicating increased expression in the endosperm compared to lines transformed with the native promoter sequence.

TABLE 3
Root and leaf qualitative data
	Construct	Mean root	Mean leaf
	GSB	2.0	2.0
	GSC	2.3	2.7
	GSD	3.0	2.8
	GSE	3.7	3.6
	GSF	4.0	3.7
	GSG	3.0	2.0
	*Score on a scale of 0 to 4 (−, +/−, +, ++, +++)

Tissue specific expression within the seed was further investigated by dissecting apart embryos and endosperm, and then determining expression levels separately. GSB (wild type) lines have a strong tissue type bins in the expression of GUS, with over 90% of the total activity in the embryo. GSD(3′ HSE deleted), GSE (5′ HSE deleted) and GSF lines show a lesser degree of embryo preferred expression, GSD (3′ HSE deleted) lines have a similar level of GUS in each tissue and GSG (HSE's replaced by PsI trimer) lines have much more GUS in the endosperm. IN fact, with GSG (HSE's replaced by PsI trimer) lines the activity of the engineered Ubi-1 promoter is similar in the embryo and endosperm, but since the endosperm is about 7.5-fold larger than the embryo, most of the GUS is in the embryo.

	TABLE 4
			Proportion
	Transformants	Seed fraction	of GUS
	GSB	embryo	0.92
		endosperm	0.08
	GSC	embryo	0.89
		endosperm	0.11
	GSD	embryo	0.47
		endosperm	0.53
	GSE	embryo	0.83
		endosperm	0.17
	GSF	embryo	0.21
		endosperm	0.79
	GSG	embryo	0.15
		endosperm	0.85

As can be seen, the expression for GSC, GSD, GSE, GSF and GSG all had altered ratios of embryo/endosperm expression. GSD had almost 50/50 and GSG had the ratio reversed with endosperm expression preferred.

Discussion

Several maize Ubi-1 promoter sequences with engineering to the putative HSEs were used to drive GUS expression in transgenic corn seed. Surprisingly, deletion or engineering of the elements does not significantly reduce expression of a reporter gene. Rather, with some Ubi-1 promoter variants, expression of GUS is increased. Deletion of both putative HSEs or of the 5′ element alone significantly increases expression, as does placing the elements adjacent so that they no longer overlap. Thus, engineering to the 5′ putative HSE increase the level of expression in seed. In the case of re-positioning the elements to remove overlap, the affect may be to inadvertently diminish the activity of the 5′ putative HSE by altering immediately adjacent sequence. Since removal or engineering of the 5′ element appears to increase expression of a reporter gene in seed, the element may restrict expression under standard growth conditions in the context of the native Ubi-1 promoter. Surprisingly, replacement of the putative HSEs with a trimer of a 22 base pair sequence from the promoter of the pea lectin gene, Ps1, does not lead to increased expression. Although the Ps1 derived element does not include a HSE consensus sequence, it does include a five out of seven base pair match to the sequence GACCCCT within the 5′ putative HSE of the Ubi-1 promoter, and this sequence may substitute for the 5′ element.

The wild type Ubi-1 sequence analyzed here drives constitutive expression of GUS. Expression is observed in leaf and root tissue and is particularly high in seed tissue. Within the kernel expression is seen in both embryo and endosperm tissues, but is preferred in the embryo. This seed and specifically embryo-preferred expression is in agreement with previous work using Ubi-1 promoter sequences in stable transformed lines (Hood, E. E., et al. (1997), “Commercial production of avidin from transgenic maize: characterization of transformant, production, processing, extraction and purification”, Mol. Breed. 3:291-306; Witcher et al., 1998, supra; Zhong et al., 1999, supra) and in embryos transiently transformed by microparticle bombardment Like the wild type sequence all of the Ubi-1 promoter variants examined here cause constitutive expression, with GUS being synthesized in leaf, root and especially seed tissue. However, within the kernel, there are notable differences in the balance of expression between embryo and endosperm tissue. None of the Ubi-1 promoter variants are as strongly embryo biased as the wild type sequence, indicating that within the kernel, the putative HSEs favor expression in embryo tissue.

Replacing the putative HSEs with a trimer of the Ps1 promoter element results in similar promoter activity in embryo and endosperm tissue, thus because of the relative tissue mass, a greater accumulation of transgene product in the endosperm. When fused to a minimal promoter, the Ps1 trimer confers seed-preferred expression in tobacco (dePater, S., et al. (1994), “A 22-bp fragment of the pea lectin promoter containing essential TGAC-like motifs confers seed-specific gene expression”, Plant Cell 5:877-886dePater, S., et al. (1996), “The 22 bp W1 element in the pea lectin promoter is necessary and, as a multimer, sufficient for high gene expression in tobacco seeds”, Plant Mol. Biol. 32:515-523), and the basic domain/leucine zipper proteins TGA1a and Opaque-2 can bind this sequence in vitro (dePater, S., et al. (1994), “bZIP proteins bind to a palindromic sequence without and ACGT core located in a seed-specific element of the pea lectin promoter”, Plant J. 6:133-140). Opaque-2 is a well characterized transcription factor of maize endosperm, and may be binding to the Ps1 trimer introduced into the Ubi-1 promoter, so facilitating expression in the endosperm. Since the overall level of transgene expression in the seed is similar in lines transformed with native Ubi-1 sequences, or with a promoter in which a Ps1 trimer replaces the HSEs, the Ps1 trimer must act to reduce expression in the embryo, as well as to increase expression in the endosperm.

Claims

1. An engineered isolated ubiquitin promoter sequence capable of directing expression of a nucleotide sequence in a plant cell, said engineered isolated ubiquitin promoter sequence comprising:

a heat shock region, wherein said heat shock region has the sequence as set forth in SEQ ID NO: 4.

2. A method for causing expression of a heterologous structural gene or open reading frame in a plant cell, said method comprising:

introducing to a plant cell an expression construct comprising an engineered isolated ubiquitin promoter sequence operably linked to said heterologous structural gene or open reading frame, wherein said engineered isolated ubiquitin promoter sequence comprises a heat shock region, wherein said heat shock region has the sequence as set forth in SEQ ID NO: 4.

3. An isolated modified ubiquitin promoter sequence comprising:

bases 1-1990 of SEQ ID NO:15 which is capable of directing expression of an operably linked sequence in a plant cell, said isolated sequence having been modified so as to delete SEQ ID NO:1.