AU3019800A - Cotton fiber transcriptional factors - Google Patents

Description

S&F Ref: 400590D1

AUSTRALIA

PATENTS ACT 1990 COMPLETE SPECIFICATION FOR A STANDARD PATENT

ORIGINAL

r

D

r r r Name and Address of Applicant Actual Inventor(s): Calgene, Inc.

1920 Fifth Street Davis California 95616 United States of America Kevin McBride David M Stalker Julie R Pear Luis Perez-Grau Spruson Ferguson St Martins Tower 31 Market Street Sydney NSW 2000 Cotton Fiber Transcriptional Factors Address for Service: Invention Title: a.

a. a 0 a a.

The following statement is a full description of this invention, including the best method of performing it known to me/us:- 5845c COTTON FIBER TRANSCRIPTIONAL FACTORS

INTRODUCTION

Technical Field This invention relates to methods of using in vitro constructed DNA transcription or expression cassettes capable of directing fiber-tissue transcription of a DNA sequence of interest i plants to produce fiber cells having an altered phenotype, ana to methods of providing for or modifying various characteristics of cotton fiber. The invention is exemplified by methods of using cotton fiber promoters for altering the phenotype of cotton fiber, and cotton fibers produced by the method.

S*20 Background In general, genetic engineering techniques have been directed to modifying the phenotype of individual prokaryotic and eukaryotic cells, especially in culture. Plant cells have proven more intransigent than other eukaryotic cells, due not only to a lack of suitable vector systems but also as a result of the different goals involved. For many applications, it is desirable to be able to control gene expression at a particular stage in the growth of a plant or in a particular plant part. For this purpose, regulatory sequences are required which afford the desired initiation of transcription in the appropriate cell types and/or at the appropriate time in the plant's development without having serious detrimental effects on plant development and productivity. It is therefore of interest to be able to isolate sequences which can be used to provide the desired regulation of transcription in a plant cell during the growing cycle of the host 10 plant.

One aspect of this interest is the ability to change the phenotype of particular cell types, such as differentiated epidermal cells that originate in fiber tissue, i.e. cotton fiber cells, so as to provide for altered or improved aspects of the 15 mature cell type. Cotton is a plant of great commercial significance. In addition to the use of cotton fiber in the production of textiles, other uses of cotton include food preparation with cotton seed oil and animal feed derived from cotton seed husks.

Despite :he imporzance of cotton as a crop, the breeding and genetic engineering of cotton fiber phenotypes has taken place at a relatively slow rate because of the absence of reliable promoters for use in selectively effecting changes in the phenotype of the fiber. In order to effect the desired phenotypic changes, transcription initiation regions capable of initiating transcription in fiber cells during development are desired.

Thus, an important goal of cotton bioengineering research is the acquisition of a reliable promoter which would permit expression of a protein selectively in cotton fiber to affect such qualities as fiber strength-, length, color and dyability.

Relevant Literature Cotton fiber-specific promoters are discussed in PCT publications WO 94/12014 and WO 95/08914, and John and Crow, Proc.

Natl. Acad. Sci. USA, 89:5769-5773, 1992. cDNA clones that are preferentially expressed in cotton fiber have been isolated. One 10 of the clones isolated corresponds to mRNA and protein that are highest during the late primary cell wall and early secondary cell wall synthesis stages. John and Crow, supra.

In animals, the ras superfamily is subdivided into the subfamilies ras which is involved in controlling cell growth and division, rab/YPT members which control secretory processes, and rho which is involved in control of cytoskeletal organization (Bourne et al., (1991) Nature 349: 117-127), and number of homologous genes have now been identified in plants (for a review, see Terryn e: a1., (1993) Plant Mol. Biol. 22: 143-152). None have been found for the important ras subfamily, all but one of the genes identified have been members of the rab/YPT1 subfamily, and there is only one recent report of the cloning of a rho gene in pea (Yang and Watson(1993) Proc. Natl. Acad. Sci. USA 8732-8736).

Little work has been done to characterize the functions of these genes in plants, although one recent report has shown that a small G protein from Arabidopsis can functionally complement a mutant form in yeast involved in vesicle trafficking, suggesting a similar function for the plant gene (Bednarek et al., (1994) Plant Physiol 104: 591-596).

In animals, two members of the rho subfamily, called Rac and Rho, have been shown to be involved in the regulation of actin organization (for a review, see Downward, (1992) Nature 359: 273- 274).

Raci has been shown to mediate growth factor-induced membrane ruffling by influencing microfilament alignment on the plasma 10 membrane (Ridley et al, (1992) Cell 70: 401-410), whereas RhoA regulates the formation of actin stress fibers associated with focal adhesions (Ridley and Hall, (1992) Cell 70: 389-399).

In yeast, the CDC42 gene codes for a rho-type protein which also regulates actin organization involved in the establishment of 15 cell polarity required for the localized deposition of chitin in the bud scar (Adams et al., (1990) J Cell Biol 111: 131-143.

Disruption of gene function, either by temperature shifts with a CDC42-temperature-sensitive mutant in yeast (Adams et al., 1990), or by micro-injec:ion into fibroblasts of mutant Rac or Rho proteins exibiting a dominant negative phenotype (Ridley et al., 1992; Ridley and Hall, 1992), leads to disorganization of the actin network.

In plants, control of cytoskeletal organization is poorly understood in spite of its importance for the regulation of patterns of cell division, expansion, and subsequent deposition of secondary cell wall polymers. The cotton fiber represents an excellent system for studying cytoskeletal organization. Cotton fibers are single cells in which cell elongation and secondary wall deposition can be studied as distinct events. These fibers develop synchronously within the boll following anthesis, and each fiber cell elongates for about 3 weeks, depositing a thin primary wall (Meinert and Delmer, (1984) Plant Physiol. 59: 1088-1097; Basra and Malik, (1984) Int Rev of Cytol 89: 65-113). At the time of transition to secondary wall cellulose synthesis, the fiber cells undergo a synchronous shift in the pattern of cortical microtubule and cell wall microfibril alignments, events which may 10 be regulated upstream by the organization of actin (Seagull, .(1990) Protoplasma 159: 44-59; and (1992) In: Proceedings of the Cotton Fiber Cellulose Conference, National Cotton Council of America, Memphis RN, pp 171-192.

Agrobacterium-mediated cotton transformation is described in Umbeck, United States Patents Nos. 5,004,863 and 5,159,135 and cotton transformation by particle bombardment is reported in WO 92/15675, published September 17, 1992. Transformation of Brassica has been described by Radke et al. (Theor. Appl. Genet.

*(1988) 75;685-694; Plan: Cell Reports (1992) 11:499-505.

SUMMARY OF THE INVENTION Novel DNA constructs and methods for their use are described which are capable of directing transcription of a gene of interest in cotton fiber, particularly early in fiber development and during secondary cell wall development. The novel constructs include a vector comprising a transcriptional and translational initiation region obtainable from a gene expressed in cotton fiber and methods of using constructs including the vector for altering fiber phenotype. Both the endogenous 3' regions and 5' regions may be important in directing efficient transcription and translation.

Three promoters are provided from genes involved in the regulation of cotton fiber development. One, Racl3, is from a protein in cotton which codes for an animal Rac protein homolog.

Racl3, shows highly-enhanced expression during fiber development.

This pattern of expression correlates well with the timing of 10 reorganization of the cytoskeleton, suggesting that the Racl3 cotton gene may, like its animal counterpart, be involved in the signal transduction pathway for cytoskeletal organization. Racl3 is a gene that is moderately expressed during fiber development turning on at 9 dpa and shutting down approximately 24 dpa. It is p 15 maximally expressed between 17-21 dpa developing fiber.

Another promoter from a cotton protein is designated 4-4.

The 4-4 mRNA accumulates in fiber cells at day 17 post anthesis and continues towards fiber maturity, which occurs at 60 days or so post anthesis. Data demonstrates that the 4-4 promoter remains very active at day 35 post anthesis.

Also provided is a promoter from a lipid transfer protein (hereinafter sometimes referred to as "Ltp") which is preferentially expressed in cotton fiber.

The methods of the present invention include transfecting a host plant cell of interest with a transcription or expression cassette comprising a cotton fiber promoter and generating a plant which is grown to produce fiber having the desired phenotype.

Constructs and methods of the subject invention thus find use in modulation of endogenous fiber products, as well as production of exogenous products and in modifying the phenotype of fiber and fiber products. The constructs also find use as molecular probes.

In particular, constructs and methods for use in gene expression in cotton embryo tissues are considered herein. By these methods, novel cotton plants and cotton plant parts, such as modified cotton fibers, may be obtained.

Also provided are constructs and methods of use relating to 10 modification of color phenotype in cotton fiber. Such constructs contain sequences for expression of genes involved in the production of colored compounds, such as anthocyanins, melanin or indigo, and also may contain sequences which provide for targeting oo of the gene products to particular locations in the plant cell, such as plastid organelles, or vacuoles. Plastid targeting is of particular interest for expression of genes involved in aromatic amino acid biosynthesis pathways, while vacuolar targeting is of particular interest where the precursors required in synthesis of the pigment are present in vacuoles.

Of particular interest are plants producing fibers which are color, that is, with pigment produced in the fiber by the plant during fiber development, as opposed to fibers which are harvested and dyed or otherwise pigmented by separate processing. Fibers from a plant producing such colored fiber may be used to produce colored yarns and/or fabric which have not been subjected to any dyeing process. While naturally colored cotton has been available from various domesticated and wild type cotton varieties, the instant application provides cotton fiber has a color produced by the expression of a genetically engineered protein.

Thus, the application provides constructs and methods of use relating to modification of color phenotype in cotton fiber. Such constructs contain sequences for expression of genes involved in the production of colored compounds, such as melanin or indigo, and also contain sequences which provide for targeting of the gene products to particular locations in the plant cell, such as plastid organelles, or vacuoles. Plastid targeting is of S 10 particular interest for expression of genes involved in the aromatic amino acid biosynthesis pathways, while vacuolar targeting is of particular interest where the precursors required in synthesis of the pigment are present in vacuoles.

DESCRIPTION OF THE DRAWINGS Figure 1 shows the DNA sequence encoding the structural Drotein from cDNA 4-4.

Figure 2 shows the sequence to the promoter construct pCGN5606 made using genomic DNA from 4-4-6 genomic clone.

Figure 3 shows the sequence to the 4-4 promoter construct DCGN5610.

Figure 4 shows the cDNA sequence encoding the Racl3 gene expressed in cotton fiber.

Figure 5 shows the sequence the promoter region from the racl3 gene.

Figure 6 shows a restriction map for pCGN4735.

Figure 7 shows the sequence of the Ltp promoter region from a cotton fiber specific lipid transfer protein gene.

Figure 8 shows the arrangement of a binary vectors pCGN5148 and pCGN5616 for plant transformation to express genes for melanin synthesis and indigo synthesis, respectively.

Figure 9 provides the results of color measurements taken from fibers of the control Coker 130 cotton used in transformation using color constructs.

Figure 10 shows the results of measurements made of color of 10 plants transformed by the pCGN5148 construct to express genes for melanin synthesis.

Figure 11 shows the results of measurements taken of the color of plants transformed by the pCGN5149 construct to express genes for melanin synthesis.

Figure 12 shows the results of measurements made of color of plants transformed to e>xress genes for indigo synthesis, usina construct pCGN5616.

Figure 13 shows control measurements made of naturally colored cotton plants which are produced by non-transgenic colored cotton plants.

DETAILED DESCRIPTION OF THE INVENTION In accordance with the subject invention, novel constructs and methods are described, which may be used provide for transcription of a nucleotide sequence of interest in cells of a plant host, preferentially in cotton fiber cells to produce cotton fiber having an altered color phenotype.

Cotton fiber is a differentiated single epidermal cell of the outer integument of the ovule. It has four distinct growth phases; initiation, elongation (primary cell wall synthesis), secondary cell wall synthesis, and maturation. Initiation of fiber development appears to be triggered by hormones. The primary cell wall is laid down during the elongation phase, lasting up to 25 days postanthesis (DPA). Synthesis of the secondary wall commences prior to the cessation of the elongation phase and continues to approximately 40 DPA, forming a wall of S 10 almost pure cellulose.

*..The constructs for use in such cells may include several e forms, depending upon the intended use of the construct. Thus, the constructs include vectors, transcriptional cassettes, expression cassettes and plasmids. The transcriptional and translational initiation region (also sometimes referred to as a "promoter,"), preferably comprises a transcriptional initiation regulatory region and a translational initiation regulatory region of untranslated 5' seouences, "ribosome binding sites," responsible for binding mRNAA to ribosomes and translational initiation. It is preferred that all of the transcriptional and translational functional elements of the initiation control region are derived from or obtainable from the same gene. In some embodiments, the promoter will be modified by the addition of sequences, such as enhancers, or deletions of nonessential and/or undesired sequences. By "obtainable" is intended a promoter having a DNA sequence sufficiently similar to that of a native promoter to provide for the desired specificity of transcription of a DNA sequence of interest. It includes natural and synthetic sequences as well as sequences which may be a combination of synthetic and natural sequences.

Cotton fiber transcriptional initiation regions chosen for cotton fiber modification may include the 4-4, racl3 and Ltp cotton fiber promoter regions provided herein.

A transcriptional cassette for transcription of a.nucleotide sequence of interest in cotton fiber will include in the direction of transcription, the cotton fiber transcriptional initiation 10 region, a DNA sequence of interest, and a transcriptional termination region functional in the plant cell. When the cassette provides for the transcription and translation of a DNA sequence of interest it is considered an expression cassette. One or more introns may be also be present.

Other sequences may also be present, including those encoding transit peptides and secretory leader sequences as desired.

Fiber-tissue transcription initiation regions of this invention are, preferably, not readily detectable in other plant :tissues. Transcription initiation regions capable of initiating transcription in other plant tissues and/or at other stages of fiber development, in addition to the foregoing, are acceptable insofar as such regions provide a significant expression level in cotton fiber at the defined periods of interest and do not negatively interfere with the plant as a whole, and, in particular, do not interfere with the development of fiber and/or fiber-related parts.

Downstream from, and under the regulatory control of, the cotton fiber transcriptional/translational initiation control region is a nucleotide sequence of interest which provides for modification of the phenotype of fiber. The nucleotide sequence may be any open reading frame encoding a polypeptide of interest, for example, an enzyme, or a sequence complementary to a genomic sequence, where the genomic sequence may be an open reading frame, an intron, a noncoding leader sequence, or any other sequence where the complementary sequence inhibits transcription, messenger S 10 RNA processing, for example, splicing, or translation. The nucleotide sequences of this invention may be synthetic, naturally derived, or combinations thereof. Depending upon the nature of the DNA sequence of interest, it may be desirable to synthesize the sequence with plant preferred codons. The plant preferred codons may be determined from the codons of highest frequency in the proteins expressed in the largest amount in the particular plant species of interest. Phenotypic modification can be *.achieved by modulating production either of an endogenous transcription or translation product, for example as to the amount, relative distribution, or the like, or an exogenous transcription or translation product, for example to provide for a novel function or products in a transgenic host cell or tissue.

Of particular interest are DNA sequences encoding expression products associated with the development of plant fiber, including genes involved in metabolism of cytokinins, auxins, ethylene, abscissic acid, and the like. Methods and compositions for modulating cytokinin expression are described in United States Patent No. 5,177,307, which disclosure is hereby incorporated by reference. Alternatively, various genes, from sources including other eukaryotic or prokaryotic cells, including bacteria, such as those from Agrobacterium tumefaciens T-DNA auxin and cytokinin biosynthetic gene products, for example, and mammals, for example interferons, may be used.

Other phenotypic modifications include modification of the color of cotton fibers. Of interest are genes involved in production of melanin and genes involved in the production of 10 indigo. Melanins are dark brown pigments found in animals, plants and microorganisms, any of which may serve as a source for sequences for insertion into the constructs of the present invention. Specific examples include the tyrosinase gene which can be cloned from Streptomyces antibioticus. The ORF438 encoded protein in S. antibioticus also is necessary for melanin production, and may provide a copper donor function. In addition, a tyrosinase gene can be isolated from any organism which makes melanin. The gene can be isolated from human hair, melanocytes or melanomas, cuttle fish and red roosters, among others. See, for example, EP Application No. 89118346.9 which discloses a process for producing melanins, their precursors and derivatives in microorganisms. Also, See, Bernan et al. Gene (1985) 37:101-110; and della-Cioppa et al. Bio/Technology (1990) 8:634-638.

Indigo may be obtained by use of genes encoding a monooxygenase such as xylene oxygenase which oxidizes toluene and xylene to (methyl) benzyl alcohol and also transforms indole to indigo. Cloning of the xylene oxygenase gene and the nucleotide and amino acid sequences are described in unexamined Japanese Patent Application Kokai:2-119777, published May 7, 1990. A dioxygenase such as naphthalene dioxygenase which also converts indole to indigo finds use; the naphthalene dioxygenase gene nahA is described in Science (1983) 222: 167. For cloning, nucleotide sequence in characterization of genes encoding naphthalene dioxygenase of Pseudomonas putida. See, Kurkela et al. Gene (1988) 73:355-362. A tryptophanase gene sequence can be used in conjunction with an oxygenase to increase the amount of indole 10 available for conversion to indigo. Sources of tryptophanase gene S. sequences include E. coli (see, for example, Deeley et al. (1982) J. Bacteriol. 151 :942-951).

Plastid targeting sequences (transit peptides) are available from a number of plant nuclear-encoded plastid proteins, such as the small subunit (SSU) of ribulose bisphosphate carboxylase, plant fatty acid biosynthesis related genes including acyl carrier protein (ACP), stearoyl-ACP desaturase, i-ketoacyl-ACP synthase and acyl-ACP thioesterase, or LHCPII genes. The encoding sequence for a transit peptide which provides for transport to plastids may include all or a portion of the encoding sequence for a particular transit peptide, and may also contain portions of the mature protein encoding sequence associated with a particular transit peptide. There are numerous examples in the art of transit peptides which may be used to deliver a target protein into a plastid organelle. The particular transit peptide encoding sequence used in the instant invention is not critical, as long as delivery to the plastid is obtained.

As an alternative to using transit peptides to target pigment synthesis proteins to plastid organelles, the desired constructs may be used to transform the plastid genome directly. In this instance, promoters capable of providing for transcription of genes in plant plastids are desired. Of particular interest is the use of a T7 promoter to provide for high levels of transcription. Since plastids do not contain an appropriate polymerase for transcription from the T7 promoter, T7 polymerase may be expressed from a nuclear construct and targeted to plastids S 10 using transit peptides as described above. (See McBride et al.

(1994) Proc. Nat. Acad. Sci. 91:7301-7305; see also copending US patent application entitled "Controlled Expression of Transgenic Constructs in Plant Plastids", serial no. 08/472,719, filed June 6, 1995, and copending US patent application SN 08/167,638, filed December 14, 1993 and PCT/US94/14574 filed December 12, 1994.) Tissue specific or developmentally regulated promoters may be Sooo useful for expression of the T7 polymerase in order to limit expression to the appropriate tissue or stage of development.

Targeting of melanin synthesis genes to vacuoles is also of interest in plant tissues which accumulate the tyrosine substrate involved in melanin synthesis in vacuoles. The protein signal for targeting to vacuoles may be provided from a plant gene which is normally transported across the rough endoplasmic reticulum, such as the 32 amino acid N-terminal region of the metallocarboxypeptidase inhibitor gene from tomato (Martineau et al. (1991) Mol. Gen. Genet. 228 :281-286). In addition to the signal sequence, vacuolar targeting constructs also encode a vacuolar localization signal (VLS) positioned at the carboxy terminus of the encoded protein. Appropriate signal sequences and VLS regions may be obtained from various other plant genes and may be similarly used in the constructs of this invention. Numerous vacuolar targetting peptides are known to the art, as are reviewed in Chrispeels et al., Cell (1992) 68:613-616.

The Maize Al gene which encodes a dihydroflavonol reductase, an enzyme of the anthocyanin pigmentation pathway is one such gene. In cells that express the Al gene, dihydrokempferol is converted to 2-8 alkylleucopelargonidin, which may be further metabolized to pelargonidin pigment by endogenous plant enzymes.

Other anthocyanin or flavonoid type pigments may also be of interest for modification of cotton cell fibers, and have been suggested for use in plant flowers (for a review of plant flower color, see van Tunen et al., Plant Biotechnology Series, Volume 2 (1990) Developmental Regulation of Plant Gene Expression,

D.

Grierson Anthocyanin is produced by a progression of steps from cellular phenylalanine pools. The R anc C1 genes are maize regulatory proteins which are active by positively affecting upstream steps in the anthocyanin biosynthesis from these pools.

The R gene is described in Perot and Cone (1989) Nucl. Acids Res., 17:8003, and the Cl gene is described in Paz-Ares et al (1987) EMBO, 6:3553-3558. Lloyd et al. (1992) Science, 258:1773-1775 discussed both genes.

Although cotton fibers in commercially grown varieties are primarily white in color, other naturally occurring cotton varieties have brown or reddish-brown fibers. Additionally, a cotton line containing green colored fibers has been identified.

Cotton lines providing such fibers are available from various sources, including the BC variety cottons (BC Cotton Inc., Box 8656, Bakersfield, CA 93389) and Fox Fibre cottons (Natural Cotton Colors, Inc., P.O. Box 791, Wasco, CA 93280).

The existence of such colored cotton lines suggests that the precursors required for the anthocyanin pigment pathways are present in cotton fibers cells, thus allowing further color phenotype modifications. Thus, the maize R and C1 genes could be 10 used in enhancing the levels of of anthocyanin produced in fiber cells. As the R and C1 proteins are proteins with a positive control at the regulatory level on anthocyanin pigment precursor biosynthesis, these proteins are expressed in the nucleus, and not

S

targetted to plastids or vacuoles.

For some applications, it is of interest to modify other aspects of the fiber. For example, it is of interest to modify various aspects of cotton fibers, such as strength or texture of a fiber. Thus, the appropriate gene may be inserted in the constructs of the invention, including genes for PHB biosynthesis (see, Peoples ec al. J. iol. Chem. (1989) 264: 15298-15303 and Ibid. 15293-15397; Saxena, Pian Molecular Biology (1990) 15:673- 683, which discloses cloning and sequencing of the cellulose synthase catalytic subunit gene; and Bowen et al. PNAS (1992) 89:519-523 which discloses chitin synthase genes of Saccharomyces cerevisiae and Candida albicans. Various constructs and methods are disclosed for the use of hormones to effect changes to fiber quality in copending US patent application entitled "Cotton Modification Using Ovary-Tissue Transcriptional factors", serial no. 08/397,652 filed February 2, 1995, the teachings of which are incorporated herein by reference.

Transcriptional cassettes may be used when the transcription of an anti-sense sequence is desired. When the expression of a polypeptide is desired, expression cassettes providing for transcription and translation of the DNA sequence of interest will be used. Various changes are of interest; these changes may include modulation (increase or decrease) of formation of S 1 0 particular saccharides, hormones, enzymes, or other biological parameters. These also include modifying the composition of the final fiber that is changing the ratio and/or amounts of water, solids, fiber or sugars. Other phenotypic properties of interest for modification include response to stress, organisms, herbicides, brushing, growth regulators, and the like. These results can be achieved by providing for reduction of expression of one or more endogenous products, particularly an enzyme or cofactor, either by producing a transcription product which is complementary (anti-sense) to the transcription product of a native gene, so as to inhibit the maturation and/or expression of the transcription product, or by providing for expression of a gene, either endogenous or exogenous, to be associated with the development of a plant fiber.

The termination region which is employed in the expression cassette will be primarily one of convenience, since the termination regions appear to be relatively interchangeable. The termination region may be native with the transcriptional initiation region, may be native with the DNA sequence of interest, may be derived from another source. The termination region may be naturally occurring, or wholly or partially synthetic. Convenient termination regions are available from the Ti-plasmid of A. tumefaciens, such as the octopine synthase and nopaline synthase termination regions. In some embodiments, it may be desired to use the 3' termination region native to the cotton fiber transcription initiation region used in a particular construct.

10 As described herein, in some instances additional nucleotide sequences will be present in the constructs to provide for targeting of a particular gene product to specific cellular locations. For example, where coding sequences for synthesis of aromatic colored pigments are used in a construct, particularly coding sequences for enzymes which have as their substrates aromatic compounds such tyrosine and indole, it is preferable to include sequences which provide for delivery of the enzyme into plastids, such as an SSU transit peptide sequence. Also, for a. synthesis of pigments derived from tyrosine, such as melanin, targeting to the vacuole may provide for enhanced color modifications.

For melanin production, the tyrosinase and ORF438 genes from Streptomyces ancibioticus (Berman et al. (1985) 37:101-110) are provided in cotton fiber cells for expression from a 4-4 and Racl3 promoter. In Streptomyces, the ORF438 and tyrosinase proteins are expressed from the same promoter region. For expression from constructs in a transgenic plant genome, the coding regions may be provided under the regulatory control of separate promoter regions. The promoter regions may be the same or different for the two genes. Alternatively, coordinate expression of the two genes from a single plant promoter may be desired. Constructs for expression of the tyrosinase and ORF438 gene products from 4-4 and rac promoter regions are described in detail in the following examples. Additional promoters may also be desired, for example plant viral promoters, such as CaMV 35S, can be used for constitutive expression of one of the desired gene products, with S 10 the other gene product being expressed in cotton fiber tissues from the 4-4 and rac promoter.

"Similarly, other constitutive promoters may also be useful in certain applications, for example the mas, Mac or DoubleMac, promoters described in United States Patent No. 5,106,739 and by Comai et al., Plant Mol. Biol. (1990) 15:373-381). When plants comprising multiple gene constructs are desired, for example plants expressing the melanin genes, 0RF438 and tyrosinase, the plants may be obtained by co-transformation with both constructs, or by transformation with individual constructs followed by plant breeding methods to obtain plants expressing both of the desired genes.

A variety of techniques are available and known to those skilled in the art for introduction of constructs into a plant cell host. These techniques include transfection with DNA employing A. tumefaciens or A. rhizogenes as the transfecting agent, protoplast fusion, injection, electroporation, particle acceleration, etc. For transformation with Agrobacterium, plasmids can be prepared in E. coli which contain DNA homologous with the Ti-plasmid, particularly T-DNA. The plasmid may or may not be capable of replication in Agrobacterium, that is, it may or may not have a broad spectrum prokaryotic replication system such as does, for example, pRK290, depending in part upon whether the transcription cassette is to be integrated into the Ti-plasmid or to be retained on an independent plasmid. The Agrobacterium host will contain a plasmid having the vir genes necessary for transfer of the T-DNA to the plant cell and may or may not have the

S.

10 complete T-DNA. At least the right border and frequently both the right and left borders of the T-DNA of the Ti- or Ri-plasmids will be joined as flanking regions to the transcription construct. The use of T-DNA for transformation of plant cells has received

S

extensive study and is amply described in EPA Serial No. 120,516, Hoekema, In: The Binary Plant Vector System Offset-drukkerij Kanters Alblasserdam, 1985, Chapter V, Knauf, et al., Genetic Analysis of Host Range Expression by Agrobacterium, In: Molecular Genetics of the Bacteria-Plant Interaction, Puhler, A.

*0 ed., Springer-Verlag, NY, 1983, p. 245, and An, et al., EMBO J.

(1985) 4:277-284.

For infection, particle acceleration and electroporation, a disarmed Ti-plasmid lacking particularly the tumor genes found in the T-DNA region) may be introduced into the plant cell. By means of a helper plasmid, the construct may be transferred to the A.

tumefaciens and the resulting transfected organism used for transfecting a plant cell; explants may be cultivated with transformed A. cumefaciens or A. rhizogenes to allow for transfer of the transcription cassette to the plant cells. Alternatively, to enhance integration into the plant genome, terminal repeats of transposons may be used as borders in conjunction with a transposase. In this situation, expression of the transposase should be inducible, so that once the transcription construct is integrated into the genome, it should be relatively stably integrated. Transgenic plant cells are then placed in an appropriate selective medium for selection of transgenic cells which are then grown to callus, shoots grown and plantlets S 10 generated from the shoot by growing in rooting medium.

O

To confirm the presence of the transgenes in transgenic cells and plants, a Southern blot analysis can be performed using .methods known to those skilled in the art. Expression products of the transgenes can be detected in any of a variety of ways, depending upon the nature of the product, and include immune assay, enzyme assay or visual inspection, for example to detect pigment formation in the appropriate plant part or cells. Once o 0 transgenic plants have been obtained, they may be grown to produce fiber having :he desired pheno-ype. The fibers may be harvested, and/or the seed colleczed. The seed may serve as a source for growing additional plants having the desired characteristics. The terms transgenic plants and transgenic cells include plants and cells derived from either transgenic plants or transgenic cells.

The various sequences provided herein may be used as molecular probes for the isolation of other sequences which may be useful in the present invention, for example, to obtain related transcriptional initiation regions from the same or different plant sources. Related transcriptional initiation regions obtainable from the sequences provided in this invention will show at least about 60% homology, and more preferred regions will demonstrate an even greater percentage of homology with the probes. Of particular importance is the ability to obtain related transcription initiation control regions having the timing and tissue parameters described herein. For example, using-the probe 4-4 and rac, at least 7 additional clones, have b-:n identified, but not further characterized. Thus, by employing the techniques 10 described in this application, and other techniques known in the art (such as Maniatis, et al., Molecular Cloning,- A Laboratory .Manual (Cold Spring Harbor, New York) 1982), other transcription initiation regions capable of directing cotton fiber transcription as described in this invention may be determined. The constructs 15 can also be used in conjunction with plant regeneration systems to obtain plant cells and plants; thus, the constructs may be used to modify the phenotype of fiber cells, to provide cotton fibers which are colored as the result of genetic engineering to eretofor unavailable hues and/or intensities.

Various variecies and lines of cotton may find use in the described methods. Cultivated cotton species include Gossypium hirsutum and G. babadense (extra-long stable, or Pima cotton), which evolved in the New World, and the Old World crops G.

herbaceum and G. arboreum.

Color phenotypes can be assessed by the use of a colorimeter, an instrument which is already used to provide objective measurements of the color of cotton samples. A colorimeter uses a combination of light sources and filters to make various estimates of a samples colors, sometimes referred to as tristimulus values.

In the past such estimtes have been used to calculate a value (Hunter's b, described below) indicating the degree of yellowness of a cotton sample. The yellowness and reflectance (from Rd, the degree of lightness or darkness of the samples) has been used to provide cotton color measurements for grading. Tests are typically conducted by exposing the face of a sample to a controlled light source. A typical color chart showing how the official grade standards relate to Rd and+ b measurements is shown in Cotton, RJ Kohel and CF Lewis, Editors #24 in AGRONOMY Series- American Soc. Agromony (see Fig. 12-6) Various colorimeter methods can be so used to quantify color and express it numerically. The Munsell method, devised by the 15 American artist Munsell, uses a classification system of paper color chips assorted according to their hue (Munsell Hue), lightness (Munsell Value), and saturation (Munsell Chroma) for visual comparison with a specimen color.

Other methods for expressing color numerically have been developed by an international organization concerned with light and color, the Commission Internationale de 1'Eclairage (CIE), having a Central Bureau located at Kegelgasse 27, A-1030 Vienna, AUSTRIA. The two most widely known of these methods are the Yxy color space, devised in 1931 based on the tristimulus value XYZ, as defined by CIE, and the L*a*b* color space, devised in 1976 to provide more uniform color differences in relation to visual differences. Color spaces* such as these are now used throughout the world for color communication. The Hunter Lab color space was developed in 1948 by R.S. Hunter as a uniform color space which could be read directly from a photoelectric colorimeter (tristimulus method).

The L*C*h color space uses the same diagram as the L*a*b* color space, but uses cylindrical coordinates instead of rectangular coordinates. In this color space, L* indicates lightness and is the same as the L* of the L*a*b* color space, C* is chroma, and h is the hue angle. The value of chroma C is 0 at 10 the center and increases according to the distance from the center. Hue angle is defined as starting at the +a axis of the L*a*b* space, and is expressed in degrees in a counterclockwise Srotation. Thus, relative to the L*a*b* space, 0' and 360' would be at the line, 90' would be 180' would be and 270' 15 would be All of the above methods can be used to obtain precise measurements of a cotton fiber color phenotype.

EXPERIMENTAL

The following examples are offered by way of illustration and not by limitation.

Example 1 cDNA libraries Tissue preparation for cDNA synthesis Leaf and root tissue were isolated from 8 inch tall greenhouse grown seedlings and immediately frozen in liquid nitrogen. Flowers were collected at the rapidly expanding 3 day preanthesis stage and also frozen. Seed was collected from 21 day postanthesis locules which had been removed from the boll and frozen entire in liquid nitrogen. Once frozen, the fiber was removed from the seed and the denuded seed used for RNA isolation.

All fibers were removed from the seed under liquid nitrogen and the fiber was ground to a powder prior to RNA isolation. Fibers were from bolls which had been tagged at anthesis.

DNA and RNA Manipulations 10 The lambda ZapII T cDNA library system of Stratagene was used for screening, and was prepared from cDNA derived from poly-A* mRNA isolated from fibers of Gossypium hirsutum cultivar Acala SJ- 2. The fibers were isolated from bolls harvested at approximately 21 dpa using field-grown plants in Israel.

15 Total PNA was isolated from 21 dpa seeds hirsutum cv Coker 130 from which the fiber had been removed) using the method of Hughes and Galau ((1988) Plant Mol Biol Reporter, 6:253-257.) All other RNAs were prepared according to Hall et al. ((1978), Proc Nail Acad Sci USA 75: 3196-3200), with the followino modifications. After the second 2M LiCI wash, the pellet was dissolved in 1/10 original volume of 10 mM Tris pH7.5 and brought to 35mM potassium acetate pH6.5 and 1/2 volume EtOH was added slowly. The mixture was placed on ice for 15 minutes and then centrifuged at 20,000 x g for 15 minutes at 4 0 C. The potassium acetate concentration was brought to 0.2M, 2 1/2 volumes EtOH added and the RNA placed at -20 0 C for several hours. The precipitate was centrifuged at 12,000 x g for 30 minutes at and the pellet was resuspended in diethylpyrocarbonate-treated water. Poly-A RNA was prepared from total mRNA utilizing an oligo(dT)-cellulose kit (Becton Dickenson) and following the manufacturer's protocol.

Cotton genomic DNA was prepared as follows. Four grams of young cotton leaf tissue (cv Coker 130) was ground to a powder in N2 and placed in an Oak Ridge tube with 0.4g polyvinylpyrolidone and 20mls extraction buffer (200mM Ches/NaOH ph9.1, 200mM NaC1, 100mMEDTA/NaOH pH9.0, 2% SDS, 0.5% Na deoxycholate, 2% Nonidet NP- 10 40, 20mM B-mercaptoethanol) was added to sample, gently mixed and incubated at 65 0 C in a shaking water bath for 10 minutes. 7.0 mis of 5M potassium acetate pH6.5 was added and carefully mixed.

Incubation was carried out on ice for 30 minutes with gentle mixing every 5 minutes. The sample was centrifuged for 20 minutes 15 at 21,000 x g and the supernatant was filtered through Miracloth into another tube and centrifuged as before. The supernatant was again filtered through Miracloth into 15 mls of room temperature isopropanol in an Oak Ridge tube. After gentle mixing, the sample was incubated a: room temperature for 10-60 minutes until the DNA precipitated. The DNA was spooled and allowed to air dry before being resuspended in 4 mis of TE on ice for 1 hour. CsCl was added to 0.97g/ml final concentration and 300 ul 10mg/ml ethidium bromide was also added before filling VTi80 quick seal tubes. The sample was centrifuged overnight at 225,000 x g overnight. The DNA was extracted with water saturated butanol and enough water was added to bring the volume to 4 mls before adding 2 volumes EtOH. The DNA was spooled, air dried and resuspended in 200 ul sterile water.

Northern and Southern Analysis For Northerns, 10ug of total RNA was isolated from various tissues, separated by electrophoresis in 1.2% agarose-formaldehyde gels and transfered onto Nytran Plus membranes (Schleicher and Schuell). Hybridization conditions consisted of a solution containing 50% formamide(v/v), 5xSSC, 0.1% SDS, 5mM EDTA, 10 Denhardts solution, 25mM sodium phosphate pH6.5 and 250 ug/ml carrier DNA. Washes were performed in 2xSSC, 0.1% SDS at 42 C 3 times for 30 minutes each time.

Cotton genomic DNA (12ug) was digested with various restriction endonucleases, electrophoresed in 0.9% agarose gels 15 and blotted onto Nytran Plus membranes. Hybridization and filter washing conditions for bozh the 3' specific and full-length cDNA insert probes were as described for Northern analysis.

Probes derived from 3'-untranslated regions were synthesized via oligonucleo:ide primers from the Racl3 cDNA, corresponding to bases 600-619 and 843-864 (Figure Each set of primers was used in a polymerase chain reaction to synthesize copies of 3'specific DNA sequences. These sequences were used as templates in the generation of single-stranded, 32 P-labeled probes off the antisense strand in a polymerase chain reaction. The full-length cDNA inserts for Racl3 were used as templates for double stranded, random primed probes using the Prime-It kit (Stratagene).

Example 2 Isolation of cDNA Clones from Cotton cDNA to the 4-4 clone was isolated from the cotton fiber library described above, and shown to express in fiber but not other tissues. This sequence was not related to any known protein. Only 400 kb of encoding sequence was present in this clone, so the library was rescreened using the cDNA to obtain full-length clones. The full-length encoding sequence is provided in Figure 1.

10 By comparing sequences of random cDNA clones against various sequence data banks via BLAST, a National Center for Biotechnology Information service, a clone, designated #105, was found to have an encoding sequence related to that of a reported lipid transfer protein.

Another clone was sequenced which showed high homology to animal Rac proteins. This clone, designated Rac, was not quite full-length, and the library was re-screened using this initial Rac DNA segment as probe. Of approximately 130,000 primary plaques screened, 56 screened positive; of these, 14 clones were isolated and seoqenced. Of these 14 clones, 12 showed identical sequence homology to :he original Rac clone and one of these cDNA clones encoded a full length cDNA and received the name Racl3.

Figure 4 shows the cDNA sequence encoding the Racl3 gene expressed in cotton fiber.

One other partial-length cDNA clone, designated Rac9, was clearly related, but distinct in DNA and amino acid sequence from Racl3. Re-screening of 150,000 plaques resulted in the isolation of 36 positive clones of which only two clones corresponded to the Rac9 sequence (both full-length clones), the remainder being Racl3. These results suggest that cotton contains genes for at least two distinct Rac proteins. Based upon the frequency of clone isolation, Racl3 is relatively highly-expressed and Rac9 less so in cotton fibers at 21 days post-anthesis (dpa), the age at which polyA mRNA was isolated for library construction.

Comparisons of the deduced amino acid sequence of Racl3 with other small G-proteins showed that the cotton Rac proteins are 10 very closely related to the Rhol protein sequence deduced from a cDNA clone isolated recently from pea (Yang and Watson, supra).

After the pea Rhol, mammalian Rac proteins show the highest homology with the cotton Rac proteins. Other proteins of the rho subfamily, such as the yeast CDC42 and human RhoA, are also clearly related to the cotton Rac genes. By contrast, the other small G-proteins of the Rab/YPT subfamily isolated from plants .such as the example shown of the tobacco RAB5 protein, as well as the human Ras proteins, are least homologous to the cotton Rac proteins of all the small G-proteins compared. The cotton and pea proteins, as well as the mammalian Racs, all have pi's above 9, whereas those of other rho and ras proteins are in the range of 5.0-6.5.

Example 3 Expression of Cotton Fiber Genes in Developina Fibers Expression of the Racl3 and 4-4 genes was assessed using mRNA prepared from various cotton tissues and from fibers at different stages of development. Blots were hybridized with probes derived from untranslated regions of Ltp, Racl3 and 4-4 genes. The gene for Racl3 exhibits highly-enhanced expression in fibers; virtually no detectable mRNA is present in leaves, roots, or flower parts, even under conditions of extended development time. Racl3 expression is detected in seeds at an age that corresponds to the highest expression levels observed in fiber tissue derived from seeds of this same age. The pattern of Racl3 expression in fibers is very dependent upon the developmental 10 stage. Expression is very low during the stage of primary wall o synthesis (0-14 dpa, see Meinert and Delmer, 1977), reaches a maximum during the transition to secondary wall synthesis (about 15-18 dpa), and declining during the stage of maximal secondary wall cellulose synthesis (about 24-28 dpa).

15 4-4 mRNA is begins to accumulate in fiber cells only at day 17 post anthesis and continues through at least day 35 post anthesis. Levels peak at day 21 and remain high. 4-4 mRNA is not o detected in other cotton tissues, and is not detected in fiber tissue before onset at 17 days post anthesis.

The #105 lipid transfer protein cDNA clone was used as a probe against cotton tissue and in a cotton fiber northern. The northern showed that the cotton fiber Ltp is highly expressed in cotton fiber. The mRNA that codes for this protein is expressed throughout fiber development at extremely high level. Northern blots indicate that this mRNA is expressed at 5 dpa and is continually expressed at a high level at 40 dpa.

Example 4 Genomic DNA cDNA for both the 4-4 and Racl3 was used to probe for genomic clones. For both, full length genomic DNA was obtained from a library made using the lambda dash 2 vector from Stratagene, which was used to construct a genomic DNA library from cotton variety Coker 130 (Gossypium hirsutum cv. coker 130), using DNA obtained from germinating seedlings.

The cotton genomic library was probed with a 3'-specific Ltp 10 probe and 6 genomic phage candidates were identified and purified.

Figure 7 provides an approximately 2 kb sequence of the Ltp 6* promoter region which is immediately 5' to the Ltp encoding region.

genomic phage clones from the cotton genomic library were identified using a 3'-specific probe for the Ltp mRNA. This was done to select the promoter from the Ltp gene that is maximally expressed in cotton fiber from the family of Ltp genes in cotton. The Ltp promoter is active throughout the fiber development period.

Example Preparation of 4-4 Promoter Constructs DCGN5606 The pCGN5606 promoter construct comprises the 4-4 cotton fiber expression cassette in a first version, version I (Figure The sequences from ntl to 65 and nt 5,494 to 5,547 correspond to fragments of the pBluescriptII polylinker where this cassette is cloned. Unique restriction enzyme sites present in these regions flanking the cassette allow the cloning of the fiber expression cassette into binary vectors including the pCGN 5138 and 1547 series.

The sequences from nt57 to 5,494 are contained in a lambda phage clone of a cotton Coker 130 genomic library. This lambda genomic clone was given the designation 4-4(6).

The region from nt 65 to nt 4,163 corresponds to the flanking region of the 4-4(6) gene. At nt 4,163 there is a NcoI 10 restriction site sequence that corresponds to the first codon of o the 4-4 (6)ORF.

The region from nucleotide 4,163 to 4,502 corresponds to part of the 4-4 (6)ORF. The sequence from nt 4,502 to 4,555 is a synthetic polylinker oligonucleotide that contains unique target 15 sites for the restriction enzymes EcoRI, Smai, Sail, NheI and t BglII. This frment from nt4,163 to 4,555 is a stuffer fragment and is left in place to facilitate the monitoring of cloning manipulations.

04 S. The genes to be expressed in cotton fiber cells using this cassette can be cloned between the NcoI restriction site and any of the polylinker sites. This operation will replace the stuffer fragment with the gene of interest. The region from nt 4,555to 5,494 corresponds to the 940 nucleotides downstream of the stop codon and constitute the 3' flanking region of the 4-4 gene.

There is a unique AscI restriction enzyme site at nt 5483.

DCGN5610 The pCGN5610 construct is a second version of a 4-4 cotton fiber expression cassette, version II, which is a modified version of pCGN5606. The two versions of the 4-4 cotton fiber expression cassette are designed to allow the cloning of tandem arrays of two fiber cassettes in one binary plasmid. The differences with respect to pCGN5606 are very minor and described below.

The XbaI restriction site in the region of nt 1 to 65 has been deleted by standard cloning manipulations.

The polylinker region is in the reverse orientation of pCGN5606.

There is a unique XbaI restriction enzyme site at nt5484. The sequences from ntl to 57 and nt 5,494 to 5,518 of pCGN5610 correspond to fragments of the pBluescriptII polylinker where this .o cassette is cloned. Unique restriction enzyme sites present in these regions allow the cloning of the fiber expression cassette into binary vectors of the pCGN 5138 and 1547 series.

The secuences from n:7 zo 5,494 are contained a lambda phage clone of a Coker 130 genomic library. This clone is described in my notebook as lambda genomic clone The *reion from n! 57 to n 4,155 corresponds to the 5' flankinc 20 region. A: nt 4,155 there is a NcoI restriction site sequence that corresponds to the first codon of the 4-4 ORF.

The region from nucleotide 4,156 to 4,500 corresponds to part of the 4-4 ORF. This fragment from nt4,156 to 4,550 is a stuffer fragment and is left in place to facilitate the monitoring of cloning manipulations. The sequence from nt 4,500 to 4,550 is a synthetic polylinker oligonucleotide containing unique target sites for the restriction enzymes BglII, NheI, SalI, SmaI and EcoRI.

The genes to be expressed in cotton fiber cells using this cassette can be cloned between the NcoI restriction site and any of the polylinker sites. This operation replaces the stuffer fragment with the gene of interest. The region from nt 4,550 to 5,494 corresponds to the 940 nucleotides downstream of .the stop codon and constitute the 3' flanking region of the 4-4 gene.

ExamDle 6 Preparation of Racl3 Promoter Constructs Genomic clone o From a genomic clone designated 15-1, mapping was done with restriction endonucleases. The largest fragment with the Racl3 coding region was identified. Theis was a Pst fragment, and when subcloned in the Bluescript T KS+ vector (BSKS+; Stratagene) was named pCGN4722. The insert had a length of 9.2 kb.

region of the Pst fragment with the Racl3 coding sequence was identified. DNA seouence was determined for approximately 1.7 20 kb 5' of the start codon and approximately 1.2 kb 3' of the stop codon. The entire Rac coding region (exons and introns) was conveniently flanked by Ndel sites.

pCGN4722 was digested with Xbal, and a 2.7 kb fragment was removed. Religation gave pCGN4730, which was then digested with Ndel, dropping out a 1.7 kb fragment containing the entire Rac coding region. Religation yielded pCGN4731.

A polylinker region was created using overlapping synthetic oligonucleotides which were PCR'ed using primers homologous to the and 3' ends of the resynthesized section. The resulting product was digested with EcooRl and Hind III and ligated into BSKS+ at the EcoRl and Hind III sites. The resulting plasmid was designated pCGN4733.

pCGN4731 and pCGN4633 were digested with Ndel and the Ndel fragment containing the synthesized polylinker region from pCGN4733 was dropped in the Ndel site of 4731, giving pCGN4734.

This last plasmid was digested with Sal and Xba, and so was pCGN5133. pCGN5133 was the 9.2 kb pst fragment in BSKS+ where the polylinker sites flanking the insert were altered to different sites for ease of manipulation. The fragment from pCGN4734 was then placed into the equivalent site of pCGN5143, giving pCGN4735.

A sequence for approximately 3 kb of the promoter construct pCGN4735 is provided in Figure 5. The resynthesized sequence :i falls between the Ndel sites located at bases 1706 and 1898 of the sequences. Thus, the sequence in Figure 5 includes approximately 1.7 kb 5' to the Ndel site 5' o the resynthesized polylinker 20 region. There is a roughly 2.5 kb sequence 5' from this sequence which is not provided in Figure 5, relative to the total 9.2 kb insert. The sequence of Figure 5 also includes approximately 1.1 kb 3' to the 3' Ndel site. Approximately 3 kb which is most 3' in the Racl3 insert is not provided in Figure 5. A map for pCGN4735 is provided in Figure 6.

Pigment Synthesis Genes Melanin A binary construct for plant transformation to express genes for melanin synthesis is prepared as follows. The melanin genes were originally isolated from the common soil bacterium Streptomyces antibioticus (Bernan et al. (1985) 34:101-110).

Melanin production is composed of a two gene system. The first gene, tyrA, encodes the catalytic unit responsible for the polymerization of the amino acid tyrosine, the primary substrate, and is termed tyrosinase. The second gene, ORF438, is responsible for binding copper and delivering copper to the tyrosinase and S: activating the enzyme. Expression of both the ORF438 and tyrA o genes ensures maximal tyrosinase activity.

The genes for both ORF438 and tyrA were fully re-synthesized 15 with respect to their DNA sequence. This was performed as the initial DNA sequence isolated from Streptomyces has a very high guanine and cytosine DNA content. Thus, the ORF438 and tryA genes were re-synthesized to appear more "plant-like" (reduced G+C content) with respec: to plant preferred codons encoding their S. 20 corresponding amino acids.

Indiao Indigo production involves conversion of the amino acid tryptophan, the primary substrate, into indole which is then converted into indoxyl. Molecules of indoxyl spontaneously convert to indigo in the presence of oxygen. A two gene system was used to affect indigo production in fiber cells. The first gene (tna) was obtained from the bacterium E. coli and encodes the enzyme tryptophanase. The designation tna stands for the gene encoding tryptophanase from E. coli, an enzyme which converts tryptophan to indole (Stewart et al., (1986) J Bacteriol 166:217- 223).

The pig designation is used for the encoding sequence to the protein for indigo production from Rhodococcus, which produces indigo from indole (Hart et al., (1990) J Gen Microbiol 136:1357- 1363). Both tna and pig were obtained by PCR. Tryptophanase is responsible for the conversion of tryptophan to indole, while the second gene (pig) encodes an indole oxygenase enzyme responsible S: for the conversion of indole to indoxyl. Both these bacterial genes were utilized in their native form.

15 Examole 8 Constructs for Tarcetina Pigment Synthesis Genes For plastid targeting, the constructs contain a fragment of the tobacco ribulose bisphosphate carboxylase small subunit gene encoding the transit peptide and 12 amino acids of the mature 20 protein (Tssu) positioned in reading frame with the appropriate encoding sequence.

For vacuolar targeting of the melanin synthesis genes, constructs include a fragment of the metallocarboxypeptidase inhibitor gene, encoding the entire 32 amino acid N-terminus signal peptide of that protein plus 6 amino acids of the mature protein (CPI+6) (Martineau et al., supra), positioned in reading frame with the appropriate encoding sequences. In addition to the signal peptide, a sequence encoding a vacuolar localization signal (VLS) is inserted 3' of the protein encoding sequence.

Constructs which contain encoding sequences for bacterial genes involved in biosynthesis of pigmented compounds and sequences for directing transport of the encoded proteins into plastids or vacuoles are prepared as follows.

Melanin The re-synthesized ORF438 and tyrA genes were treated in two distinct ways depending on which compartment in the fiber cell the final protein products would be localized. One chimeric gene/plant binary construct (designated pCGN5148) contained the genes targeted to the fiber cell plastids. To do this, 12 amino acids of a gene for the small subunit of carboxylase (SSU) plus 15 the original 54 amino acid SSU transit peptide were fused to the amino termini of both the ORF438 and tyrA gene products respectively. These peptide sequences allow the ORF438 and tyrA gene products (proteins) to be efficiently targeted to the plastid. This targeting was initiated as the plastid is the site 20 of tyrosine production within the fiber cell.

The second chimeric gene/plant binary construct (designated pCGN5149) contained the ORF438 and tyrA genes targeted to the vacuole within the fiber cell. Based on information from other biological systems, it was postulated that the fiber cell vacuole may contain a high concentration of tyrosine for melanin polymerization. Both the ORF438 and tryA genes contain the 29 amino acid signal peptide from a tomato carboxypeptidase inhibitor (CPI) protein as amino terminal gene fusions to direct these proteins to the endoplasmic reticulum (ER) secretory system of the fiber cell.

In addition, the tyrA gene has an 8 amino acid vacuolar targeting peptide (VTP) from CPI fused at the carboxy terminus so that the mature copper-activated tyrosinase will eventually be targeted to the vacuole of the fiber cell. Both the ORF438 and tyrA proteins also had potential glycosylation sites removed via site-directed mutagenesis of the ORF438 and tyrA genes respectively. Potential plant cell glycosylation of these proteins upon their expression in fiber cells could result in tyrosinase inactivation, hence removal of potential glycosylation sites was deemed necessary.

15 Indiao The only modification to the indigo genes was the fusion of the tobacco SSU transit peptide encoding DNA sequences onto the o amino terminal region of both the tna and pig genes to affect the localization of both the tryptophanase and indole oxygenase 20 proteins to the fiber cell plastid. These are the same exact gene fusions that were made for the plastid-directed proteins for melanin production in construct 5148. The tna and pig gene products were targeted to the fiber cell plastid as that is the primary site of tryptophan synthesis.

Example 9 Expression Constructs Melanin The modified genes for both the plastid and vacuolar targeted ORF438 and tyrosinase proteins were placed into a fiber expression cassette to be "switched" on during development of the cotton fiber cell. The "switch" (promoter) utilized for the melanin constructs was 4-4. The modified ORF438 and tyrA genes were cloned into the 4-4 promoter cassette and these chimeric genes then inserted into a binary plasmid to create plasmids pCGN5148 and pCGN5149, containing the modified genes for plastid and vacuolar targeted ORF438 and tyrosinase proteins, respectively.

These binary plasmids also contain genetic determinants for their stable maintenance in E. coli and Agrobacterium and also contain a r chimeric gene for plant cell expression of the bacterial kanamycin resistance gene. This kanamycin resistance marker allows for the 15 selection of transformed versus non-transformed cotton cells when plant hypocotyl or leaf segments are infected with Agrobacterium containing the binary plasmids.

S* A block diagram of the plasmid pCGN5149, having vacuolor targetting seounces, is shown in Figure 8. Plasmid pCGN5148 (not 20 shown) is constructed the same as 5149, only pCGN5148 has plastidtargetting sequences Indico As with the melanin genes, the plastid-directed tna and pig genes were placed in the fiber-specific 4-4 promoter cassette and these chimeric genes subsequently inserted into a binary plasmid to create plasmid pCGN5616. A block diagram of plasmid pCGN5616 is shown in Figure 8.

10 e oo Anthocvanin A construct has been prepared for the expression of the maize R and CI genes in developing cotton fiber. These genes are known to be responsible for the production of Anthocyanin pigments by acting in a regulatory manner to turn on the chalcone pathway for production of anthocyanins (red spectrum colors). The R and CI genes were placed under the control of the Racl3 promoter cassette. A binary plasmid designated pCGN4745 (not shown), contains both the R and CI genes each under control of the Racl3 promoter.

Example Cotton Transformation Explant Preparation Coker 315 seeds are surface disinfected by placing in Clorox sodiumr hypochlorite solution) for 20 minutes and rinsing 3 times in sterile distilled water. Following surface sterilization, seeds are germinated in 25 x 150 sterile tubes containing 25 mls 1/2 x MS salts: 1/2 x B5 vitamins: 1.5% glucose: 0.3% gelrite. Seedlings are germinated in the dark at 28 0 C for 7 days. On the seventh day seedlings are placed in the light at 28±2 0

C.

Cocultivation and Plant Regeneration Single colonies of A. tumefaciens strain 2760 containing .binary plasmids pCGN2917 and pCGN2926 are transferred to 5 ml of MG/L broth and grown overnight at 30 0 C. Bacteria cultures are diluted to 1 x 108 cells/ml with MG/L just prior to cocultivation.

Hypocotyls are excised from eight day old seedlings, cut into 0.7 cm sections and placed onto tobacco feeder plates (Horsch et al. 1985). Feeder plates are prepared one day before use by plating 1.0 ml tobacco suspension culture onto a petri plate containing Callus Initiation Medium CIM without antibiotics (MS 10 salts: B5 vitamins: 3 glucose: 0.1 mg/L 2,4-D: 0.1 mg/L kinetin: 0.3% gelrite, pH adjusted to 5.8 prior to autoclaving). A sterile filter paper disc (Whatman was placed on top of the feeder cells prior to use. After all sections are prepared, each section was dipped into an A. tumefaciens culture, blotted on sterile paper towels and returned to the tobacco feeder plates.

Following two days of cocultivation on the feeder plates, e hypocotyl sections are placed on fresh Callus Initiation Medium containing 75 mg/L kanamycin and 500 mg/L carbenicillin. Tissue g* was incubated ar 28±2°C, 30uE 16:8 light:dark period for 4 weeks.

At four weeks the entire exolant was transferred to fresh callus initiation medium containing antibiotics. After two weeks on the second pass, the callus was removed from the explants and split between Callus Initiation Medium and Regeneration Medium (MS salts: 40mM KN03: 10 mM NH4C1:B5 vitamins:3% glucose:0.3% gelrite:400 mg/L carb:75 mg/L kanamycin).

Embryogenic callus was identified 2-6 months following initiation and was subcultured onto fresh regeneration medium.

Embryos are selected for germination, placed in static liquid Embryo Pulsing Medium (Stewart and Hsu medium: 0.01 mg/l NAA: 0.01 mg/L kinetin: 0.2 mg/L GA3) and incubated overnight at 30 0 C. The embryos are blotted on paper towels and placed into Magenta boxes containing 40 mis of Stewart and Hsu medium solidified with Gelrite. Germinating embryos are maintained at 28±2 0 C 50 uE m-2s- 1 16:8 photoperiod. Rooted plantlets are transferred to soil and established in the greenhouse.

Cotton growth conditions in growth chambers are as follows: 10 16 hour photoperiod, temperature of approximately 80-850, light intensity of approximately 5001lEinsteins. Cotton growth conditions in greenhouses are as follows: 14-16 hour photoperiod with light intensity of at least 400pEinsteins, day temperature 90-95 0 F, night temperature 70-75°F, relative humidity to approximately Plant Analysis Flowers from greenhouse grown Tl plants are tagged at o anthesis in the creenhouse. Scuares (cotton flower buds), flowers, bolls etc. are harvested from these plants at various stages of development and assayed for enzyme activity.

GUS

fluorometric and histochemical assays are performed on hand cut sections as described in co-pending application filed for Martineau et al., supra. For fiber color characteristics, plants are visually inspected, or northern or western analysis can be performed, if necessary.

Example 11 Expression of Transgenic Pigment Synthesis Genes Melanin Plants that exhibited resistance to the kanamycin selectable marker via a leaf assay and corresponding Western analysis were considered transformed. Transgenic fiber was collected from individual plant transformants at different stages of fiber development and analyze in two ways. One was to analyze fiber at 10 a single developmental time point for each transgenic cotton plant to compare tyrosinase expression between transgenic events. The other was to screen developing fiber from selected plants to analyze the timing of tyrosinase expression under the control of the fiber-specific 4-4 promoter, by Western blots using antisera prepared against purified tyrosinase protein.

For the plastid-targeted construct pCGN5148 9 of 13 events screened for tyrosinase expression were positive, while 13 of the 16 transformed vacuolar-targeted construct pCGN5149 events which were screened were positive. Expression level in the fiber in tyrosinase positive plants is approximately 0.1-0.5% fiber cell protein. Clearly, the cotton fiber cells comprising the DNA color constructs DNA produce the necessary proteins required for synthesis of a pigment.

Visually, the lint from the tyrosinase positive events exhibits color to varying degrees, while plants that do not express the enzyme do not exhibit any color. Colorimeter measurements of cotton fiber taken from control Coker 130 plants and plants from various events transformed with pCGN5148 are provided in Figures 9 and 10, respectively.

Fiber from pCGN5148 (plastid-directed) plants demonstrates a bluish-green color phenotype. One event, 5148-50-2-1 included cotton fiber cells (linters) which were colored and which had an negative a* value less than 8.0, as measured on the L*a*b* color space. Coker 130 cotton fiber cells do not typically demonstrate a negative a* value.

These colored cotton cells also had a color located on the 10 L*C*h color space with a relatively high hue angle value h, greater than 135'. Normal Coker 130 fibers have a similar value which is not greater than about 90' as measured by this method.

Results of colorimeter measurements of cotton fiber taken from plants transformed with pCGN5149 are provided in Figure 11.

Fiber from plants expressing tyrosinase from construct pCGN5149 (vacuolar-targetted) tends to have a light brown phenotype.

Indico Resistance to the kanamycin selectable marker via leaf assay and Western analysis was again the criterion for designating a plant as transformed by pCGN5616. Transgenic fiber was collected from individual plant transformants at different stages of fiber development. The transgenic developing fiber is screened from selected plants to analyze the timing of tna and pig gene expression under the control of the fiber-specific 4-4 promoter and fiber is also analyzed at a single developmental time point for each transgenic cotton plant for comparison of both tryptophanase and indole oxygenase expression between transgenic events, by using Western blots with antisera prepared against the tryptophanase and indole oxygenase proteins.

For the indigo events, 15 of 24 screened plants were positive for expression of both the tryptophanase and indole oxygenase enzymes. Expression levels in the fiber of these proteins is between 0.05-0.5% fiber cell protein. Approximately half of these transformants are expressing both genes in the fiber resulting in a very faint light blue color phenotype. Visually, there is a 10 faint blue color in the majority of these positive events, particularly in 20-30 dpa fiber in the unopened boll. Results of colorimeter measurements of cotton fiber taken from various events of plants transformed with pCGN5616 are provided in Figure 12.

Many of these events had relatively low a* values (less than 2) with elevated b* values (greater than 10), as measured on the LT'ab* color space. Similarly, several 5149 events also measured with an a* value less than 2 while maintaining a b* value greater than BC Cotton Colorimeter measurements taken on naturally colored fiber from four separate BC cotton lines is provided in Figure 13.

The above results demonstrate that the color phenotype of a transgenic cotton fiber cell can be altered by expressing pigment synthesis genes. The transgenic cotton fiber cells include both a pigment synthesizing protein, and pigment produced by the pigment synthesizing protein. As shown from the results of Figures 9 through 13, expression of a pigment gene of interest can result in cotton fiber cells in which the synthesis of pigments combined with appropriate targeting sequences results in modification of color phenotype in the selected plant tissue, yielding colored cotton fiber by expression from a genetically engineered construct.

All publications and patent applications cited in this 10 specification are herein incorporated by reference as if each ***individual publication or patent application are specifically and individually indicated to be incorporated by reference.

Although the foregoing invention has been described in some detail, by way of illustration and example for purposes of clarity 15 and understanding, it will be readily apparent to those of .:ee oordinar skill in the art that certain changes and modifications may be made thereto, without departing from the spirit or scope of the appended claims.

Claims (42)

1. A DNA construct comprising as operably joined components in the direction of transcription, a cotton fiber transcriptional factor and an open reading frame encoding a protein of interest, wherein said transcriptional factor is selected from the group consisting of the Ltp, the 4-4 and the rac promoter sequences. 10 2. The DNA construct according to Claim 1, further comprising a transport signal encoding sequence from a plant nuclear-encoded gene.

3. The DNA construct according to Claim 2, wherein said transport signal encoding sequence comprises a plastid transit peptid.

4. The DNA construct according to Claim 1, wherein said transport signal encoding sequence encodes a signal peptide which provides for transport across the rough endoplasmic reticulum.

5. The DNA construct according to Claim 4, wherein said sequence further comprises, 3' to said open reading frame, a vacuolar localization signal.

6. The DNA construct of Claim 1 wherein said pigment is melanin or indigo.

7. The DNA construct of Claim 6 wherein said open reading frame is from a bacterial gene.

8. The DNA construct of Claim 7 wherein said bacterial gene is selected from the group consisting of ORF438, tyrA, anthocyanin R gene, anthocyanin C1 gene, pig, and tna.

9. A plant cell comprising a DNA construct of Claim 1.

10. A cotton plant cell according to Claim 9.

11. A cotton fiber cell according to Claim

12. A plant comprising a cell of any one of Claims 9- 11.

13. A method of modifying fiber phenotype in a cotton 10 plant, said method comprising: transforming a plant cell with DNA comprising a *,construct for expression of a protein in a pigment biosynthesis pathway, wherein said construct comprises as operably joined components: a transcriptional initiation region functional in cells of said cotton plant, an open reading frame encoding a protein of interest, and a transcriotional :ermination region functional in cells of said cotton plant, wherein said plan: cell comprises a substrate of said protein; and growing said plant cell to produce a cotton plant, wherein said protein reacts with said substrate to produce said pigment.

14. The method of Claim 13 wherein said construct further comprises a transport signal encoding sequence from a plant nuclear-encoded gene. The method of Claim 13 wherein said transport signal encoding sequence encodes a signal peptide which provides for transport across the rough endoplasmic reticulum.

16. The method of Claim 13 wherein said DNA comprises constructs for expression of two proteins in a pigment 10 biosynthesis pathway, wherein each of said constructs comprises components i) through iv), and wherein said two o.o proteins are not encoded by the same gene.

17. The method of Claim 16 wherein said pigment is melanin and said proteins are encoded by tyrA and ORF438. 15 18. The method of Claim 16 wherein said pigment is indigo and said proteins are cna and pig.

19. The method of Claim 16 wherein said pigment is anythocyanin and said constructs comprise the anthocyanin R and Cl regulatory genes.

20. The method of Claim 13 wherein plant cell is a cotton fiber cell, and wherein said transcriptional region is a fiber tissue transcription iniation region.

21. The method of Claim 20 wherein said transcriptional region is selected from the group consisting of the Ltp, the 4-4 and the rac promoter sequences

22. A recombinant DNA construct comprising the cotton tissue transcriptional sequence shown in Figure 2.

23. A recombinant DNA construct comprising the cotton tissue transcriptional sequence shown in Figure

24. An isolated DNA encoding sequence of Figure 1. An isolated DNA encoding sequence of Figure 4.

26. The method of Claim 13 wherein said protein of interest is involved in the synthesis of a plant hormone.

27. An isolated DNA sequence comprising the cotton lipid transfer protein encoding sequence of Figure 7.

28. A cotton fiber cell comprising a DNA sequence, wherein 10 said DNA sequence comprises as operably joined components in the direction of transcription, a cotton fiber transcriptional factor and an open reading frame encoding a protein required for C synthesis of a pigment.

29. A cotton fiber cell according to Claim 27 comprising pigment produced by said pigment synthesizing protein.

30. A cotton fiber cell according to Claim 27 wherein said DNA sequence further comprises a transport signal encoding a sequence from a plant nuclear-encoded gene.

31. A ccr:cn fiber cell according to Claim 29, wherein said transpor: signal encoding sequence comprises a plastid transit peptid.

32. A cotton fiber cell according to Claim 29, wherein said transport signal encoding sequence encodes a signal peptide which provides for transport across the rough endoplasmic reticulum.

33. A cotton fiber cell according to Claim 31, wherein said sequence further comprises, 3' to said open reading frame, a vacuolar localization signal.

34. A cotton fiber cell according to Claim 27 wherein said transcriptional factor is selected from the group consisting of the cotton fiber lipid transfer promoter sequence, the 4-4 promoter sequence and the rac promoter sequence.

35. A cotton fiber cell according to Claim 27 wherein said pigment is melanin or indigo.

36. A cotton fiber cell according to Claim 27 wherein said open reading frame is from a bacterial gene.

37. A cotton fiber cell according to Claim 35 wherein said 10 bacterial gene is selected from the group consisting of ORF438, tyrA, anthocyanin R gene, anthocyanin C1 gene, pig, and tna.

38. A cotton fiber cell comprising melanin.

39. A cotton fiber cell comprising indigo. A cotton fiber cell which is colored by genetic engineering and which has a negative a* value less than 1.0 as measured on the Laa*b' color space. to 41. The cotton fiber cell of Claim 39 wherein said negative a* value is less than a

42. The cotton fber cell of Claim 40 wherein said negative a" value is less than a -E.0.

43. A cotton fiber cell which is colored by genetic engineering and which has an a' value less than 2 and the b* value greater than 10 as measured on the L*a*b* color space.

44. A cotton fiber cell which is colored by genetic engineering and which has a hue angle value h of greater than 100' as measured on the L*C'h color space. 54 The cotton fibre cell of claim 43 wherein said h value is greater than a 135*.

46. A cotton plant comprising one or more cells according to any one of claims 28 to

47. A DNA construct comprising a cotton fiber transcriptional factor operably linked to an open reading frame encoding a protein of interest, substantially as hereinbefore described with reference to any one of the Examples.

48. A method of modifying phenotype in a cotton plant, substantially as hereinbefore described with reference to any one of the Examples.

49. A cotton fibre cell which is coloured by genetic engineering, substantially as hereinbefore described with reference to any one of the Examples. Dated 28 April 2000 Calgene, Inc. S Patent Attorneys for the Applicant/Nominated Person SPRUSON FERGUSON o* o ER:\LBA]0313