AU2003254410A1 - Flavonoid 3',5'hydroxylase gene sequences and uses therefor - Google Patents

Description

WO 2004/020637 PCT/AU2003/001111 Flavonoid 3',5' Hydroxylase gene sequences and uses therefor BACKGROUND OF THE INVENTION 5 HELD OF THE INVENTION The present invention relates generally to a genetic sequence encoding a polypeptide having flavonoid 3', 5'-hydroxylase (F3'5'H) activity and to the use of the genetic sequence and/or its corresponding polypeptide thereof inter alla to manipulate color in flowers or 10 parts. thereof or in other plant tissue. More particularly, the F3'5'H has the ability to modulate dihydrokaempfcrol (DHK) metabolism as well as the metabolism of other substrates such as dihydroquercetin (DHQ), naringenin and eriodictyol. Even more particularly, the present invention provides a genetic sequence encoding a polypeptide having F3'5rH activity when expressed in rose or gerbera or botanically related plants. The 15 instant invention further ,relates to antisense and sense molecules or RNAi-inducing molecules corresponding to all or part of the subject genetic sequence or a transcript thereof as well as to genetically modified plants as well as cut flowers, parts and reproductive tissue from such plants. The present invention further relates to promoters which operate efficiently in plants such as rose, gerbera or botanically related plants. 20 DESCRIPTION OF PRIOR ART Reference to any prior art in this specification is not, and should not be taken as, an acknowledgment or any form of suggestion that this prior art forms part of the common 25 general knowledge in any country. Bibliographic details of references provided in the subject specification are listed at the end of the specification.

WO 2004/020637 PCT/AU2003/001111 The flower or ornamental plant industry strives to develop new and different varieties of flowers and/or plants. An effective way to create such novel varieties is through the manipulation of flower color. Classical breeding techniques have been used with some success to produce a wide range of colors for almost all of the commercial varieties of 5 flowers and/or plants available today. This approach has been limited, however, by the constraints of a particular species' gene pool and for this reason it is rare for a single species to have the full spectrum of colored varieties. For example, the development of novel colored varieties of plants or plant parts such as flowers, foliage and stems would offer a significant opportunity in both the cut flower and ornamental markets. In the flower 10 or ornamental plant industry, the development of novel colored varieties of major flowering species such as rose, chrysanthemum, tulip, lily, carnation, gerbera, orchid, lisianthus, begonia, torenia, geranium, petunia, nierembergia, pelargonium, iris, impatiens and cyclamen would be of great interest. A more specific example would be the development of a blue rose or gerbera for the cut flower market. 15 In addition, the development of novel colored varieties of plant parts such as vegetables, fruits and seeds would offer significant opportunities in agriculture. For example, novel colored seeds would be useful as proprietary tags for plants. Furthermore modifications to flavonoids common to berries or fruits including grapes and apples and their juices 20 including wine have the potential to impart altered style characteristics of value to such fruit and byproduct industries. Flower color is predominantly due to three types of pigment: flavonoids, carotenoids and betalains. Of the three, the flavonoids are the most common and contribute a range of 25 colors from yellow to red to blue. The flavonoid molecules that make the major contribution to flower color are the anthocyanins, which are glycosylated derivatives of cyanidin and its methylated derivative peonidin, delphinidin or delphinidin-based molecules and its methylated derivatives petunidin and malvidin and pelargonidin. Anthocyanins are-localised in the vacuole of the epidermal cells of petals or the vacuole of 30 the sub epidermal cells of leaves.

WO 2004/020637 PCT/AU2003/001111 .-3.

The flavonoid pigments are secondary metabolites of the phenylpropanoid pathway. The biosynthetic pathway for the flavonoid pigments flavonoidd pathway) is well established, (Holton and Cornish, Plant Cell 7: 1071-1083, 1995; Mol et al, Trends Plant Se. 3: 212 217, 1998; Winkel-Shirley, Plant Physiol. 126: 485-493, 2001a; and Winkel-Shirley, Plant 5 Physiol. 127: 1399-1404, 2001b) and is shown in Figures 1A and B. Three reactions and enzymes are involved in the conversion of phenylalanine to p-coumaroyl-CoA, one of the first key substrates in the flavonoid pathway. The enzymes are phenylalanine ammonia lyase (PAL), cinnamate 4-hydroxylase (C4H) and 4-coumarate: CoA ligase (4CL). The first committed step in the pathway involves the condensation of three molecules of 10 malonyl-CoA (provided by the action of acetyl CoA carboxylase (ACC) on acetyl CoA and C0 2 ) with one molecule of p-coumraroyl-CoA. This reaction is catalysed by. the enzyme chalcone synthase (CHS). The product of this reaction, 21,4,4',6, tetrahydroxy chalcone, is normally rapidly isomerized by the enzyme chalcone flavanone isomerase (CHI) to produce naringenin. Naxingenin is subsequently hydroxylated at the 3 position of 15 the central ring by flavanone 3-hydroxylase (F3H) to produce dihydrokaempferol (DHK). The pattern of hydroxylation of the B-ring of dihydrokaempferol (DHK) plays a key role in determining petal color. The B-ring can be hydroxylated at either the 3', or both the 3' and 5' positions, to produce dihydroquercetin (DHQ) or dihydromyricetin (DHM), 20 respectively. Two key enzymes involved in this part of the pathway are flavonoid 3' hydroxylase and flavonoid 3', 5'-hydroxylase, both of the cytochrome P450 class of enzymes. Cytochrome P450 enzymes are widespread in nature and genes have been isolated and sequenced from vertebrates, insects, yeasts, fungi, bacteria and plants. 25 Flavonoid 3'-hydroxylase (F3'H) is a key enzyme in the flavonoid pathway leading to the cyanidin- based pigments which, in many plant species (for example Rosa spp., Dianthus spp., Petunia spp., begonia, cyclamen, impatiens, morning glory and chrysanthemum), contribute to red and pink flower color. 30 Flavonoid 3', 5'-hydroxylase (F3'5'H) is a key enzyme in the flavonoid pathway leading to the delphinidin- based pigments which, in many plant species (for example, Petunia spp., WO 2004/020637 PCT/AU2003/001111 -4 Viola app., Lisianthus spp., Gentiana spp., Sollya spp., Salvia spp., Clitoria &pp., Kennedia spp., Campanula spp., Lavandula spp., Verbena spp., Torenia app,, Delphinium spp., Solanum Fpp., Cineraria spp., Vilis spp., Babiana stricia, Pinus app., Picea spp., Larix spp., Phaseolus spp,, Vaccinium spp., Cyclamen spp., Iris spp., Pelargonium sp., 5 Liparicae, Geranium spp., Pisum spp., Lathyrus spp., Catharanthus spp., Malvia spp., Mucuna spp., Vicla spp., Saintpaula spp., Lagerstroemia spp., bouchina spp., Plumbago spp., Hypocalyptus spp., Rhododendron spp., Linum app., Macroptilum app., Hibiscus spp., Hydrangea spp., Cymbidium spp., Millettia spp., Hedysarum spp., Lespedeza spp., Asparagus spp. Anuigonon spp., Pisum spp., Freesia spp., Brunella spp., Clarkia spp., 10 etc.), contribute to purple and blue flower color. Many plant species such as roses, gerberas, chrysanthemums and carnations, do not produce delphinidin-based pigments because they lack a F3'5'H activity. The next step in the pathway, leading to the production of the colored anthocyanins from 15 the dihydroflavonols (DHK, DHQ, DHM), involves dihydroflavonol-4-reductase (DFR) leading to the production of the leucoanthocyanidins. The leucoanthocyanidins are subsequently converted to the anthocyanidins, pelargonidin, cyanidin and delphinidin or delphinidin-based molecules. These flavonoid molecules are unstable under nonnal physiological conditions and glycosylation at the 3-position, through the action of 20 glycosyltransferases, stabilizes the anthocyanidin molecule thus allowing accumulation of the anthocyanins. In general, the glycosyltransferases transfer the sugar moieties from UDP sugars to the flavonoid molecules and show high specificities for the position of glycosylation and relatively low specificities for the acceptor substrates (Seitz and Hinderer, Anthocyanins. In: Cell Culture and Somatic Cell Genetics of Plants. Constabel, 25 F. and Vasil, I.K. (eds.), Academic Press, New York, USA, 5: 49-76, 1988). Anthocyanins can occur as 3-monosides, 3-biosides and 3-triosides as well as 3, 5-diglycosides and 3, 7 diglycosides associated with the sugars glucose, galactose, rhamnose, arabinose and xylose (Strack and Wray, In: The Flavonoids - Advances in Research since 1986. Harborne, J.B. (ed), Chapman and Hall, London, UK, 1-22, 1993). 30 WO 2004/020637 PCT/AU2003/001111 -5 Glycosyltransferases involved in the stabilization of the anthocyanidin molecule include UDP glucose: flavonoid 3-glucosyltransferase (3GT), which transfers a glucose moiety from LDP glucose to the 3-0-position of the anthocyanidin molecule to produce anthocyanidin 3-O-glucoside. 5 In petunia and pansy (amongst others), anthocyanidin 3-0-glucoside are generally glycosylated by another glycosyltransferase, TJDP rhamnose: anthocyanidin 3-glucoside rhamnosyltransferase (3RT), which adds a rhamnose group to the 3-0-bound glucose of the anthocyanin molecule to produce the anthocyanidin 3-rutinosides, and once acylated, 10 can be further modified by UDP: glucose authocyanin 5 glucosyltransferase (5GT). However, in roses (amongst others), the anthocyanidin 3-0-glucosides are generally glycosylated by another glycosyltransferase, UJDP: glucose anthocyanin 5 glucosyltransferase (5GT) to produce anthocyanidin 3, 5 diglucosides. 15 Many anthocyanidin glycosides exist in the form of acylated derivatives. The acyl groups that modify the anthocyanidin glycosides can be divided into two major classes based upon their structure. The aliphatic acyl groups include malonic acid or sucoinic acid and the aromatic class include the hydroxy cinnamic acids such as p-cournaric acid, caffeic acid and ferulic acid and the benzoic acids such asp-hydroxybenzoic acid. 20 Methylation at the 3' and 5' positions of the B-ring of anthocyanidin glycosides can also occur. Methylation of cyanidin-based pigmnents leads to the production of peonidin. Methylation' of the 3' position of delphinidin-based pigments results in the production of petunidin, whilst methylation of the 3' and 5' positions results in malvidin production. 25 Methylation ofmalvidin can also occur at the 5-0 and 7-0 positions to produce capensinin (5-0-methyl malvidin) and 5, 7-di-O-methyl malvidin. In addition to the above modifications, pH of the vacuole or compartment where pigments are localised and copigmentation with other flavonoids such as flavonols and flavones can 30 affect petal color. Flavonols and flavones can also be aromatically acylated (Brouillard and WO 2004/020637 PCT/AU2003/001111 -6 Dangles, In: The Flavonoids -Advances in Research since 1986. Harborne, J.B. (ed), Chapman and Hall, London, UK, 1-22, 1993). The ability to control F3'5'H activity, or other enzymes involved in the flavonoid pathway, 5 in flowering plants would provide a means of manipulating the color of plant parts such as petals, fruit, leaves, sepals, seeds etc. Different colored versions of a single cultivar could thereby be generated and in some instances a single species would be able to produce a broader spectrum of colors. 10 Two nucleotide sequences (referred to herein as SEQ ID NO:l and SEQ ID NO:3) encoding petunia F3'5'Hs have been cloned (see International Patent Application No. PCT/AU92/00334 and Holton et aL, Nature, 366: 276-279, 1993a). These sequences were efficient in modulating 3', 5' hydroxylation of flavonoids in petunia (see International Patent Application No. PCT/AU92/00334 incorporated herein by reference and Holton et 15 al., 1993a, supra), tobacco (see Intemational Patent Application No. PCT/AU92/00334 incorporated herein by reference) and carnations (see International Patent Application No. PCT/AU96/00296 incorporated herein by reference). Surprisingly, however, inclusion of these sequences in standard expression cassettes, did not lead to the production of intact or fall-length transcripts as detectable by RNA or Northern blot analysis and consequently 3', 20 5'-hydroxylated flavonoids were not produced in roses. There is a need, therefore, to identify further genetic sequences encoding F3'5'Hs which efficiently accumulate and are then able to modulate 3', 5' hydroxylation of flavonoids such as anthocyanins in roses and other key commercial plant species.

WO 2004/020637 PCT/AU2003/001111 -7 SUMMARY OF THE INVENTION Throughout this speciftication, unless the context requires otherwise, the word "comprise", or variations such as "comprises" or "comprising", will be understood to imply the 5 inclusion of a stated element or integer or group of elements or integers but not the exclusion of any other element or integer or group of elements or integers. Nucleotide and amino acid sequences are referred to by a sequence identifier number (SEQ ID NO:). The SEQ ID NOs: correspond numerically to the sequence identifiers <400>1 10 (SEQ ID NO:l), <400>2 (SEQ ID NO:2), etc. Genetic sequences encoding a F3'5'H have been identified and cloned from a number of plant species. The F3'5'H genetic sequences when expressed in rose petal tissue results in detectable level of delphinidin or delphinidin-based molecules as determined by a 15 chromatographic technique such as thin layer chromatography (TLC) or high performance liquid chromatography (HPLC). Alternatively, or in addition, expression of the genetic sequences in rose petal tissue results in a sufficient level and length of transcript which is capable of being translated to F3'5'H. This is conveniently measured as delphinidin or delphinidin-based molecules, detectable using a chromatographic technique such as TLC 20 or HPLC. The genetic sequences of the present invention permit the modulation of expression of genes encoding this enzyme by, for example, de novo expression, over expression, suppression, antisense inhibition, ribozyme activity, RNAi-induction or methylation-induction. The ability to control F3'5'H synthesis in plants and more specifically in roses or gerberas permits modulation of the composition of individual 25 anthocyanins as well as alteration of relative levels of flavonols and anthocyanins, thereby enabling the manipulation of color of tissues and/or organs of plants such as petals, leaves, seeds, sepals, fruits etc. Accordingly, one aspect of the present invention provides an isolated nucleic acid 30 molecule comprising a sequence of nucleotides encoding or complementary to a sequence encoding a flavonoid 3', S' hydroxylase (F3'5'H) or a polypeptide having F3'5'H activity WO 2004/020637 PCT/AU2003/001111 -8 wherein expression of said nucleic acid molecule in a, rose petal tissue results in detectable levels of delphinidin or delphinidin-based molecules as measured by a chromatographic technique. 5 Another aspect of the present invention is directed to an isolated nucleic acid molecule comprising a sequence of nucleotides encoding or complementary to a sequence encoding a F3'5'H or a polypeptide having F3'5'H activity wherein expression of said nucleic acid molecule in a rose petal tissue results in a sufficient level and length of transcript which is translated to said F3'5'H as determined by detectable levels of delphinidin or delphinidin 10 based molecules as measured by a chromatographic technique. The isolated nucleic acid molecule of the present invention, therefore, encodes a F3'5'H which is capable of more efficient conversion of DHK to DHM in roses than is the F3'5'Hl encoded by the nucleotide sequence set forth in SEQ ID NO: I and SEQ ID NO:3 as 15 measured by delphinidin production in rose petals. Efficiency as used herein relates to the capability of the F3'5'H enzyme to convert its substrate DHK or DHQ into DHM in a rose cell (or any cell of a commercially important plant such as gerbera). This conversion provides the plant with a substrate (DHM) for other 20 enzymes of the flavonoid pathway which are present in the palnt to further modify the substrate. This modification may include for example, glycosylation; acylation, rhamnosylation and/or methylation, to produce various anthocyanins which contribute to the production of a range of colors. The modulation of 3',5'-hydroxylated anthocyanins in rose is thereby enabled. Efficiency is conveniently assessed by one or more parameters 25 selected from: extent of F3'5'H transcription, as determined by the amount of intact F3'5'H mRNA produced (as detected by Northern blot analysis); extent of translation of the F3'5'H mRNA, as determined by the amount of translation product produced; extent of F3'5'H enzyme activity as determined by the production of anthocyanin derivatives of DHQ or DIM including delphinidin or delphinidin-based pigments (as detected by TLC or HPLC); 30 the extent of effect on flower color.

WO 2004/020637 PCT/AU2003/001111 It has also been surprisingly determined that certain combinations of promoter and F3'5'H gene sequences that were functional in carnation and petunia were not functional in rose. Surprisingly, only a particular subset of promoter and F3'5'H gene sequence combinations resulted in 3', 5'-hydroxylated flavonoids in rose flowers. These included F3'5'H sequences 5 isolated from Viola spp., Salvia spp. Lavandula spp. and Sollya spp. Furthermore, the Viola P3'5'H (or pansy F3'5'H) sequences were found to result in the highest accumulation of 3', 5'-hydroxylated flavonoids in rose. The novel promoter and F3'5'H gene sequence combinations can be employed inter alia to modulate the color or flavour or other characteristics of plants or plant parts such as but not limited to flowers, fruits, nuts, roots, 10 stems, leaves or seeds. Thus, the present invention represents a new approach to developing plant varieties having altered color characteristics. Other uses include, for example, the production of novel extracts of F3'5'H transformed plants wherein the extract has use, for example, as a flavouring or food additive or health product or beverage or juice or coloring. Beverages may include but are not limited to wines, spirits, teas, coffee, 15 milk and dairy products. In a preferred embodiment, therefore, the present invention provides an isolated nucleic acid molecule comprising a sequence of nucleotides encoding, or complementary to a sequence encoding pansy F3'5'H, salvia F3'5'1, lavender F3'5'H, kennedia F3'5'H or 20- sollya F3'5'H or a functional derivative of the enzyme. The nucleotide sequences encoding the pansy F3'5'H (SEQ ID NOs:9 and 11), salvia F3'5'H (SEQ ID NOs:13 and 15), sollya F3'5'H (SEQ ID NO:17), lavender F3'5'H (SEQ ID NO:31) and kennedia F3'5'H (SEQ ID NO:26) are defined by sequence identifiers 25 indicated in parentheses. A summary of the sequence identifiers is shown in Table 1. Accordingly, another aspect of the present invention provides a nucleic acid molecule comprising a nucleotide sequence or complementary nucleotide sequence substantially as set forth in SEQ ID NO:9 (pansy) or SEQ ID NO:1l (pansy) or SEQ ID NO:13 (salvia) or 30 SEQ ID NO:15 (salvia) or SEQ ID NO:17 (sollya) or SEQ ID NO:31 (lavender) or SEQ ID NO:26 (kennedia) or having at least about 50% similarity thereto or capable of WO 2004/020637 PCT/AU2003/001111 -10 hybridizing to the sequence set forth in SEQ ID NO:9 or SEQ ID NO:11 or SEQ ID NO:13 or SEQ ID NO:15 or SEQ ID NO:17 or SEQ ID NO:31 or SEQ ID NO:26 under low stringency conditions. 5 The amino acid sequences of the preferred F3'5H enzymes are set forth in SEQ ID NO:10 (pansy) or SEQ ID NO:12 (pansy) or SEQ ID NO:14 (salvia) or SEQ ID NO:16 (salvia) or SEQ ID NO:18 (sollya) or SEQ ID NO:32 (lavender) or SEQ ID NO;27 (kennedia). A further aspect of the present invention provides a method for producing a transgenic 10 flowering plant capable of synthesizing a F3'S'H said method comprising stably transforming a cell of a suitable plant with a nucleic acid sequence which comprises a sequence of nucleotides encoding said F3'5'H under conditions permitting the eventual expression of said nucleic acid sequence, regenerating a transgenic plant from the cell and growing said transgenic plant for a time and under conditions sufficient to permit the 15 expression of the nucleic acid sequence. The expression of the nucleic acid sequence generally results in a transcription of sufficient level and length to encode a F3'5'H. This is conveniently determined by detectable levels of delphinidin or delphinidin-based molecules as measured by obromatographic techniques such as TLC or HPLC. The transgenic plant may thereby produce a non-indigenous F3'5'H at elevated levels relative to 20 the amount expressed in a comparable non-transgenic plant. This generally results in a visually detectable color change in the plant or plant part or preferably in the inflorescence or flowers of said plant. Another aspect of the present invention contemplates a method for producing a transgenic 25 plant with reduced F3'5'H activity, said method comprising stably transforming a cell of a suitable plant with a nucleic acid molecule which comprises a sequence of nucleotides encoding or complementary to a sequence encoding a F3'5H activity, regenerating a transgenic plant from the cell and where necessary growing said transgenic plant under conditions sufficient to permit the expression of the nucleic acid. 30 WO 2004/020637 PCT/AU2003/001111 - 11 Yet another aspect of the present invention contemplates a method for producing a genetically modified plant with reduced F3'5'H activity, said method comprising altering the F3'5'H gene through modification of the indigenous sequences via homologous recombination from an appropriately altered F3'5'H gene or derivative or part thereof 5 introduced into the plant cell, and regenerating the genetically modified plant from the cell. Still another aspect of the present invention contemplates a method for producing a transgenic flowering plant exhibiting altered floral or inflorescence properties, said method comprising stably transforming a cell of a suitable plant with a nucleic acid sequence of the 10 present invention, regenerating a transgenic plant from the ecll and growing said transgenic plant for a time and under conditions sufficient to permit the expression of the nucleic acid sequence. Still a further aspect of the present invention contemplates a method for producing a 15 transgenic flowering plant exhibiting altered floral or inflorescence properties, said method comprising alteration of the F3'5'H gene through modification of nucleotide sequences via homologous recombination from an appropriately altered F3'H gene or derivative or part thereof introduced into the plant cell, and regenerating the genetically modified plant from the cell. 20 Even yet another aspect of the present invention extends to a method for producing a transgenic plant capable of expressing a recombinant gene encoding a F3'5'H or part thereof or which carries a nucleic acid sequence which is substantially complementary to all or a part of a niRNA molecule encoding said F3'5'H, said method comprising stably 25 transforing a cell of a suitable plant with the isolated nucleic acid molecule comprising a sequence of nucleotides encoding, or complementary to a sequence encoding, a F3'5'H, where necessary under conditions permitting the eventual expression of said isolated nucleic acid molecule, and regenerating a transgenic plant from the cell.

WO 2004/020637 PCT/AU2003/001111 -12 Even still another aspect of the present invention extends to all transgenic plants or parts of transgenic plants or progeny of the transgenic plants containing all or part of the nucleic acid sequences of the present invention, or antisense forms thereof and/or any homologs or related forms thereof and, in particular, those transgenic plants which exhibit altered floral 5 or inflorescence properties. Even still another aspect of the present invention extends to all transgenic plants or parts of transgenic plants or progeny of the transgenic plants containing all or part of the nucleic acid sequences of the present invention, or antisense forms thereof and/or any homologs or 10 related forms thereof and, in particular, those transgenic plants which exhibit altered aerial parts of the plant such as fruit, berries, sepal, bract, petiole, peduncle, ovaries, anthers or stem properties. Another aspect of the present invention contemplates the use of the extracts from 15 transgenic plants or plant parts transgenic plants or progeny of the transgenic plants containing all or part of the nucleic acid sequences of the present invention and, in particular, the extracts from those transgenic plants when used as a flavouring or food additive or health product or beverage or juice or coloring. 20 A further aspect of the present invention is directed to recombinant forms of F3'5H Another aspect of the present invention contemplates the use of the genetic sequences described herein in the manufacture of a genetic construct capable of expressing a F3'5'U or down-regulating an indigenous F3'5 1 H enzyme in a plant. 25 Yet another aspect of the present invention is directed to a prokaryotic or eukaryotic organism carrying a genetic sequence encoding a F3'5'H extrachromasonially in plasmid form.

WO 2004/020637 PCT/AU2003/001111 -- 13~ Still another aspect of the present invention extends to a recombinant polypeptide comprising a sequence of amino acids substantially as set forth in SEQ ID NO:10 or SEQ ID NO:12 or SEQ ID NO:14 or SEQ ID NO:16 or SEQ ID NO:18 or SEQ ID NO:32 or SEQ ID NO:27 or an amino acid sequence having at least about 50% similarity to SEQ ID 5 NO:10 or SEQ ID NO:12 or SEQ ID NO:14 or SEQ ID NO:16 or SEQ ID NO:18 or SEQ ID NO:32 or SEQ ID NO:27 or a derivative of said polypeptide. The present invention further provides promoters which operate efficiently in plants such as rose and gerbera or botanically related plants. Such promoters include a rose CHS 10 promoter, chrysanthemum CHS promoter and a CaMV 35S promoter. A summary of sequence identifiers used throughout the subject specification is provided in Table 1: 15 TABLE 1 Summary of sequence iden~fiers SEQ ID NAME SPECIES TYPE DESCRIPTION NO: OFSEQ 1 pet /fl.nt Petunia hybrid nucleotide F3'5'H cDNA 2 petHf.aa - Petunia hybrida amino acid translation of F3'5'HcDNA 3 petHf2.nt Petunia hybrid nucleotide F3'5'H eDNA 4 peff,1.aa Petunia hybrida amino acid translation of F3'S'H cDNA 5 RoseCHlS promoter Rosa hybrida nucleotide promoterfragment 6 D8 oligo#2 Petunia hybrid nucleotide oligonucleolide 7 D8 oligo #4 Petunia hybrida nucleotide oligonucleotide 8 chrysanCHSATG chrysanthemum nucleotide oligonucleotide 9 BP#1&nt Viola spp. nucleotide F3'5'H eDNA 10 BP#18.aa Viola spp. amino acid translation of F3'5'HCDNA 11 BP#40.nt Viola spp. nuoleotide F3'5' cDNA 12 BP#40.aa Viola spp. amino acid translation of F3'5'H eDNA 13 Sal#2,nt Salvia spp. nucleotide F3'5 'HeDNA WO 2004/020637 PCT/AU200310011 11 -14 SEQ ID NAME SPECIES TYPE DESCRIPTION NO: OF SEQ 14 Sal42.aa Salvla spp. amino acid translation ofF3S'5'HcDNAA 15 .Sall147.nt Salvia spp. nucleotide PS'S5'I 7JDAA 16 Sal#47.aa &dvia $pp. amino acid translation ofP3 '5 'H cDNA 1V ' Soll#5.nt Sofiya spp. nucleotide PS '5'HcDNA 18 Soil#5 .aa Sollya spp. amino arid translation of FS 5'HcDNA 19 FLS-Nco Petunia hybrida nucleotide oligonucleotide 20 BpeaHF.Int Clitoria tern atea nucleotide PS'5H cDNA 21 BpeaFIFZ.aa Citorla tern area amnino acid .. translation of F3 57 ~FcDNA 22 Genx#48.nt Gentiana nuclootide F3 '5'H cDNA triora 23 Genll48aa Gentiana amino acid translation ofP3 'S'HvDNA 24 PeIDS 5' Petunia hybrid nucleotide oligonuclootide 25 Bpe primer Clitoria ternatea nucleotide oligonucleozide 26 Kenn#S1 .nt Kennedia spp. nucleotide FS '7 eDNA 27 Kenng31.aa Kennedia spp. amino acid translation Of F3'SHeD NA 28 chiysCHSxit chrysanthemum nucleotide CHS cDNA 29 chi'ysCHS.aa chrysanthemum amino acid translation of CHS cDNA 30 chrysCHSpromnior chrysanthemums nucleotide promoter fragment 31 LBG.nt Lavandula nil nucleotide F3'57! eDNA 32 LBG.aa Lavcznduta nil am-ino acid translation ofF H KSS c DNA WO 2004/020637 PCT/AU2003/001111 - 15 BRIEF DESCRIPTION OF TIE FIGURES Figures IA and 1B are schematic representations of the biosynthesis pathway for the flavonoid pigments. Figure 1A illustrates the general production of the anthocyanidin 5 3-glucosides that occur in most plants that produce anthocyanins. Figure 1B represents further modifications of anthocyanins that occur in petunia. Enzymes involved in the pathway have been indicated as follows: PAL = Phenylalanine ammonia-lyase; C4H=Cinnamate 4-hydroxylase; 4CL = 4-coumarate: CoA ligase; CHS = Chalcone synthase; CH = Chalcone flavanone isomerase; F3H = Flavanone 3-hydroxylase; 10 DFR = Dihydroflavonol-4-reductase; ANS = Anthocyanidin synthase, 3GT = UDP-glucose: flavonoid 3-O-glucosyltransferase; 3RT = UDP rhamnose: anthocyanidin 3-glucoside rhamosyltransferase, AR-AT = Anthocyanidin-rutinoside acyltransferase, SGT = Anthocyanin 5-glucosyltransferase; 3' OMT = Anthocyanin 3' Q-methyltransferase, 3'5' OMT Anthocyanin 3', 5' 0 --methyltransferase. Other 15 abbreviations include: DHK = dihydrokaempferol, DHQ = dihydroquercetin, DHM = dihydromyricetin, TABLE 2: Descriptions of the abbreviations used in Figures 2 to 52 ABBREVIATION DESCRIPTION Amp ampicillin resistance gene which confers resistance to the antibiotic ampicillin ColElori plasmid origin of replication fl ori (+) fl filamentous phage origin ofreplication GcntR gentamycin. resistance gene which confers resistance to the antibiotic gentamycin LB left border of the T-DNA npt1!I the neomycin phosphotransferase III gene which confers resistance to the antibiotic kanamycin ori pRi plasmid origin of replication ori 322 plasmid origin ofreplication pACYC ori modified replicon from pACYC184 fromE. coli WO 2004/020637 PCT/AU2003/001111 -16 pVS1 a broad host range origin of replication from a plasmid from Pseuodomonas aeruginasa rev approximate location of the M13 reverse primer site used in sequence analysis RB right border of the T-DNA TeIR tetracycline resistance gene which confers resistance to the antibiotic tetracycline -20 approximate location of the M13 -20 primer site used in sequence analysis RK2 broad host range Gram-negative plasmid RK2 origin Figure 2 is a diagrammatic representation of the plasmid pCGP602, pCGP601 and pCGP 176 containing petunia P3'S'HpetHfl cDNA clones from P. hybrida cv. OGB. The petunia F3'5'H petHf fragment was used in the preparation of constructs containing the 5 petunia F3'5'H cDNA clone. 3 2 P-labelled fragments of the 1.6 kb BspHI/FspI fragment were used to probe 'petal cDNA libraries. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. Figure 3 is a diagrammatic representation of the plasmid pCGP175 containing the petunia 10 F3'5'HpetHf2 cDNA clone from P. hybrida cv. OGB. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. Figure 4 is a diagrammatic representation of the plasmid pCGP1303 containing a subclone of the petunia F3'5'Hpetlfl cDNA clone from pCGP601. The construction of pCGP1303 15 is described in Example 4. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. Figure 5 is a diagrammatic representation of the binary plasmid pCGP1452. The AmCHS 5': petHf: petD8 3' gene from pCGP485 was cloned into the binary vector pWTT2132 20 (DNAP) in a tandem orientation with the chimacric SuRS gene. The construction of pCGP1452 is described in Example 4. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations.

WO 2004/020637 PCT/AU2003/001111 -17 Figure 6 is a diagrammatic representation of the binary plasmid pWTT2132 (DNAP) containing the 35S 5': SuRB selectable marker gene and a multi-cloning site. A description of pWTT2132 is given in Example 4. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. 5 Figure 7 is a diagrammatic representation of the plasmid pCGP725. The AmCHS 5'r petHf: petD8 3' gene from pCGP485 was cloned into p1luescript II (KS (+) vector. The construction of pCGP725 is described in Example 4. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. 10 Figure 8 is a diagrammatic representation of the binary plasmid pCGP1453. The Mac: petHfl: mas 3' gene from pCGP628 was cloned into the binary vector pWTT2132 (DNAP) in a tandem orientation with the chimaeric SuRB gene. The constmction of pCGP1453 is described in Example 4. Selected restriction endonuclease sites are marked, 15 Refer to Table 2 and Table 4 for a description of the abbreviations. Figure 9 is a diagrammatic representation of the binary plasmid pCxP1457. The petD8 5 petHfl: peID8 3' gene from pCGP1107 was cloned into the binary vector pWTT2132 (DNAP) in a tandem orientation with the chimaeric SuRB gene. The construction of 20 pCGP1457 is described in Example 4. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. Figure 10 is a diagrammatic representation of the binary plasmid pCGP1461. The shortpetFLS P'. petHf]: petFLS 3' gene from pCGP497 was cloned into the binary vector 25 pWTT2132 (DNAP) in a tandem orientation with the chimnaeric SuRB gene. The construction of pCGP1461 is described in Example 4. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. Figure 11 is a diagrammatic representation of the binary plasmid pCGP1 616. The petR T 30 5': petHf: nos 3' gene from pCGP846 was cloned into the binary vector pWTT2132 (DNAP) in a tandem orientation with the chimactic SuRB gene. The construction of WO 2004/020637 PCT/AU2003/001111 - 18 pCGP1616 is described in Example 4. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. Figure 12 is a diagrammatic representation of the binary plasmid pCGP1623. The 5 mas/355: patHf1: ocs 3' gene from pCGP1619 was cloned into the binary vector pWTT2132 (DNAP) in a tandem orientation with the chimaeric SuRB gene. The construction of pCGP1623 is described in Example 4. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. 10 Figure 13 is a diagrammatic representation of the binary plasmid pCGP 1638. The CaMV 353: petHfi: nos 3' gene from pCGP1636 was cloned into the binary vector pWTT2132 (DNAP) in a tandem orientation with the chimaeric SuRB gene. The construction of pCGP 1636 is described in Example 4. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. 15 Figure 14 is a diagrammatic representation of the binary plasmid pCGP1860. The RoseCHS 5': pet0f7: nos 3' gene from pCGP200 was cloned into the binary vector pWTT2132 (DNAP) in a tandem orientation with the chimaeric SuRB gene. The construction of pCGP1860 is described in Example 4. Selected restriction endonuclease 20 sites arc marked. Refer to Table 2 and Table 4 for a description ofthe abbreviations. Figure 15 is a diagrammatic representation of the binary plasmid pCGP2123. The CaMV35S: petuf2: ocs 3' gene from pCGP2109 was cloned into the binary vector pCGP1988 in a tandem orientation with the chimaeric SuRB gene. The construction of 25 pCGP2123 is described in Example 4. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. Figure 16 is a diagrammatic representation of the binary plasmid pCGP1988. The multi cloning site of the binary vector pWTT2132 (DNAP) was replaced with the multi-cloning 30 site from pNEB193 (New England Biolabs). The construction of pCGP-1988 is described in WO 2004/020637 PCT/AU2003/001111 -19 Example 4. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. Figure 17 is a diagrammatic representation of the plasmid pCGP2105. The 358 5': ocs 3' 5 expression cassette with multiple restriction endonuclease sites between the promoter and terminator fragments is in a pBluescript SK (+) vector backbone.. The construction of pCGP2105 is described in Example 4. Selected restriction endonuclease sites are marked, Refer to Table 2 and Table 4 for a description of the abbreviations. 10 Figure 18 is a diagrammatic representation of the binary plasmid pCGP1307. The petD8 5': GUS: petD8 3' gene from pCGP 1106 was cloned into the binary vector pCGN1548 in a tandem orientation to the chimaeric nprI selectable marker gene. The construction of pCGP1307 is described in Example 6. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. 15 Figure 19 is a diagrammatic representation of the binary plasmid pCGP1506. The longpetFLS 5': GUS: petFLS 3' gene from pCGP496 was cloned into the binary vector pBIN19 in a tandem orientation to the chimaeric nptIl selectable marker gene. The construction of pCGP1506 is described in Example 6. Refer to Table 2 and Table 4 for a 20 description of the abbreviations. Figure 20 is a diagrammatic representation of the binary plasmid pCGPl626. The ChrysCHS 5': GUS: nos 3' gene from pCGP1622 was cloned into the binary vector pWTT2132 in a tandem orientation with the chimaeric SuRB gene. The construction of 25 pCGP1626 is described in Example 6. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. Figure 21 is a diagrammatic representation of the binary plasmid pCGP1641. The petRT 5': GUS: petR T 3' gene from pCGP1628 was cloned into the binary vector pWTT2132 in 30 a tandem orientation with the chimacric SuRD gene. The construction of pCGP1641 is WO 2004/020637 PCT/AU2003/001111 -20 described in Example 6. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. Figure 22 is a diagrammatic representation of the binary plasmid pCGP1861. The 5 RoseCHS 5': GUS: nos .3' gene from pCGPI97 was cloned into the binary vector pWTT2132 in a tandem orientation with the chimaeric SuRB gene. The construction of pCGP161 is described in Example 6. Refer to Table 2 and Table 4 for a description of the abbreviations. 10 Figure 23 is a diagrammatic representation of the binary plasmid pCGP1953. The AmCHS 5'; GUS: petD8 3' gene from pCQP1952 was cloned into the binary vector pWTT2132 in a tandem orientation with the chimaeric SuRB gene. The construction of pCG?1953 is described in Example 6. Refer to Table 2 and Table 4 for a description of the abbreviations. 15 Figure 24 is a diagrammatic representation of the binary plasmid pWTT2084 (DNAP) containing a 35S 5': GUS: ocs 3', gene in a convergent orientation to the chimaeric SuRB selectable marker gene. A description of pWTT2084 is given in Example 6. Refer to Table 2 and Table 4 for a description of the abbreviations. 20 Figure 25 is a diagrammatic representation of the plasmid pCGP1959 containing the F3'5'H BP#18 cDNA clone from Viola spp. cv Black Pansy in a pBluescript SK 11 (+) backbone. A description of pCGP1959 is given in Example 7. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the 25 abbreviations. Figure 26 is a diagrammatic representation of the plasmid pCGP1961 containing the F3'3'H BP#40 cDNA clone froni Viola spp. cv Black Pansy in a pBluescript SK II (+) backbone. A description of pCpP1961 is given in Example 7.' Selected restriction 30 endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations.

WO 2004/020637 PCT/AU2003/001111 -21 Figure 27 is a diagrammatic representation of the binary plasmid pCGP1972. The AmCHS 3': BP#18: petD8 3' gene from pCGP1970 was cloned into the binary vector pWTT2132 (DNAP) in a tandem orientation with the chimaeric SuRB gene. The construction of 5 pCGP 1972 is described in Example 7. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. Figure 28 is a diagrammatic representation of the binary plasmid pC

G

P1973. The AmCIIS 5': BP#40: petD 3' gene from pCGP1971 was cloned into the binary vector pWTT2132 10 (DNAP) in a tandem orientation with the chimaeri SuRB gene. The construction of pCGP1973 is described in Example 7. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. Figure 29 is a diagrammatic representation of the binary plasmid pCGP 1967. The CaMV 15 35S: BP#18:ocs 3' gene from pCGP1965 was cloned into the binary vector pWTT2132 (DNAP) in a tandem orientation with the chimaeric SuRB gene. The construction of pCGP1967 is described in Example 7, Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. 20 Figure 30 is a diagrammatic representation of the binary plasmid pCGPl969. The CaMV 35S: BP#40:ocs 3' gene from pCGP 1966 was cloned into the binary vector pWTT2132 (DNAP) in a tandem orientation with the chimaeric SuRB gene. The construction of pCGP1969 is described in Example 7. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. 25 Figure 31 is a diagrammatic representation of the plasmid pCGP1995 containing the F3 'TH Sa#2 cDNA clone from Salvia spp, in a pBluescript SK II (+) backbone. A description of pCGPl995 is given in Example 7. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. 30 WO 2004/020637 PCT/AU2003/001111 -22 Figure 32 is a diagrammatic representation of the plasmid pCGP1999 containing the F3'5'H Sal#47 eDNA clone from Salvia spp in a pBluescript SK II (+) backbone. A description of pCGP1999 is given in Example 7. Selected restriction endonuclease sites are marked, Refer to Table 2 and Table 4 for a description of the abbreviations. 5 Figure 33 is a diagrammatic representation of the binary plasmid pCGP2121. The AmCHS 5': Sa#2: petD8 3' gene from pCGP2116 was cloned into the.binary vector pCGP1988 in a tandem orientation with the chimaeric SuRB gene. The construction of pCGP2121 is described in Example 7. Selected restriction endonuclease sites are marked. Refer to Table 10 2 and Table 4 for a description of the abbreviations. Figure 34 is a diagrammatic representation of the binary plasind pCGP2122. The AmCHS 5': Sal#47: petD8 3' gene from pCGP2117 was cloned into the binary vector pCGP1988 in a tandem orientation with the chimaeric SuRB gene. The construction of pCGP2122 is 15 described in Example 7. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. Figure 35 is a diagrammatic representation of the binary plasmid pCGP2120. The CaMV 35S:Sal#2:ocs 3' gene from pCGP2112 was cloned into the binary vector pCGP1988 in a 20 tandem orientation with the chimaeric SuRS gene. The construction of pCGP2120 is described in Example 7. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. Figure 36 is a diagrammatic representation of the binary plasmid pCGP2119. The CaMV 25 35S:SaI#47:ocs 3'gene from pCGP2111 was cloned into the binary vector pCGP 1988 in a tandem orientation with the chimaeric SuRB gene. The construction of pCGP2119 is described in Example 7. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations.

WO 2004/020637 PCT/AU2003/001111 -23 Figure 37 is a diagrammatic representation of the plasmid pCQP2110 containing tie F3'S 'H Soll#5 cDNA clone from Sollya spp. in a pBluescript SK II (+) backbone. A description of pCGP2110 is given in Example 7. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. 5 Figure 38 is a diagrammatic representation of the binary plasmid pCGP2130. The AmCHS 5': Soll#5: petDB 3' gene from pCGP2128 was cloned into the binary vector pCGP1988 in a tandem orientation with the chimaeric SuRR gene. The construction of pCGP2130 is described in Example 7. Refer to Table 2 and Table 4 for a description of the 10 abbreviations. Figure 39 is a diagrammatic representation of the binary plasmid pCGP2131. The CaMV 353: Soll#5:ocs 3' gene from pCGP2129 was cloned into the binary vector pCGP1988 in a tandem orientation with the chimaeric SuRB gene. The construction of pCGP2131 is 15 described in Example 7. Refer to Table 2 and Table 4 for a description of the abbreviations. Figure 40 is a diagrammatic representation of the plasmid pCGP2231 containing the F3'5'H Kenn#31 cDNA clone from Kennedia spp. in a pBluescript SK II (+) backbone. A 20 description of pCGP2231 is given in Example 7. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. Figure 41 is a diagrammatic representation of the binary plasmid pCGP2256. The AmCHS S ': Kenn#31: petD8 3' gene from pCGP2242 was cloned into the binary vector pCGP1988 25 in a tandem orientation with the chimaeric SuRB gene. The construction of pCGP2256 is described in Example 7. Refer to Table 2 and Table 4 for a description of the abbreviations.

WO 2004/020637 PCT/AU2003/001111 '24 Figure 42 is a diagrammatic representation of the binary plasmid pCGP2252. The CaMV 35S: Kenn#31:ocs 3' gene from pCGP2236 was cloned into the binary vector pCGP1988 in a tandem orientation with the chimaeric SuRB gene. The construction of pCGP2252 is described in Example 7. Refer to Table 2 and Table 4 for a description of the 5 abbreviations. Figure 43 is a diagrammatic representation of the plasmid pBHF2F containing the full length F3'5'H BpeaHF2 cDNA clone from Clitoria ternatea in a pBluescript SK 1H (+) backbone. A description of pBHF2F is given in Example 7. Selected restriction 10 endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. Figure 44 is a diagrammatic representation of the binary plasmid pCGP2135. The AmCHS 5'. BpeaHF2: petD8 3' gene from pCGP2133 was cloned into the binary vector 15 pCGP1988 in a tandem orientation with the chimaeric SuRB gene. The construction of pCGP2135 is described in Example 7. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. Figure 45 is a diagrammatic representation of the binary plasmid pBEBFS. The eCaMV 20 355: BpeaHF2: nos 3' gene was constructed by replacing the GUS fragment from pBE2113-GUSs with the Clitoria F3'5'H BpeaHF2 cDNA clone from pBHF2F The construction of pBEBFS is described in Example 7. Refer to Table 2 and Table 4 for a description of the abbreviations. 25 Figure 46 is a diagrammatic representation of the binary plasmid pCGP2134. The CaMV 35S: BpeaHF2: ocs 3' gene from pCGP2132 was cloned into the binary vector pCGP1988 in a tandem orientation with the chimaeric SuRB gene. The construction of pCGP2134 is described in Example 7. Refer to Table 2 and Table 4 for a description of the abbreviations. 30 WO 2004/020637 PCT/AU2003/001111 -25 Figure 47 is a diagrammatic representation of the plasmid pG48 containing the P3'5'H Gen#48 eDNA clone from Gentiana triflora in a pBluescript SK 11 (+) backbone. A description of pG48 is given in Example 7. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. 5 Figure 48 is a diagrammatic representation of the binary plasmid pCGP1498. The AmCHS 5': Gen#48: petD8 3' gene from pCGP1496 was cloned into the binary vector pCGP1988 in a tandem orientation with the chimaeric SuRB gene. The construction of pCGP1498 is described in Example 7. Selected restriction endonuclease sites are marked. Refer to Table 10 2 and Table 4 for a description of the abbreviations. Figure 49 is a diagrammatic representation of the birnary plasmid pBEGHF48. The eCaMV 353: Gen#48: nos 3' gene was constructed by replacing the GUS fragment from pBE2113 GUSs with the Gentiana F3'5'H Gen#48 cDNA clone from pG48. The construction of 15 pBEGHF48 is described in Example 7. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations. Figure 50 is a diagrammatic representation of the binary plasmid pCGP1982. The CaMV 35S:Gen#48:ocs 3' gene from pCGP1981 was cloned into the binary vector pWTT2132 20 (DNAP) in a tandem orientation with the chimaeric SuRB gene. The construction of pCGP1982 is described in Example 7. Refer to Table 2 and Table 4 for a description of the abbreviations. Figure 51 is a diagrammatic representation of the plasmid pLFH8 containing the PS' 'H 25 LBG cDNA clone from Lavandula nit in a pBluescript SK H (+) backbone. A description of pLFH8 is given in Example 7. Selected restriction endonuclease sites are marked. Refer to Table 2 and Table 4 for a description of the abbreviations.

WO 2004/020637 PCT/AU2003/001111 -26 Figure 52 is a diagrammatic representation of the binary plasmid pBELF8. The eCaMV 355: LBG: nos 3' gene was constructed by replacing the GUS fragment from pBE2113 GUSs with the Lavandula F3'5'H LBG cDNA clone from pLHES The construction of pBELF8 is described in Example 7, Refer to Table 2 and Table 4 for a description of the 5 abbreviations.

WO 2004/020637 PCT/AU2003/001111 -27 DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS In accordance with the present invention, genetic sequences encoding polypeptides having F3'5'H activity have been identified, cloned and assessed. The recombinant genetic 5 sequences of the present invention permit the modulation of expression of genes'encoding this enzyme by, for example, de novo expression, over-expression, sense suppression, antisense inhibition, ribozyme, minizyme and DNAzyme activity, RNAi-induction or raethylation-induction or other transcriptional or post-transcriptional silencing activities. RNAi-induction includes genetic molecules such as hairpin, short double stranded DNA or 10 RNA, and partially double stranded DNAs or RNAs with one or two single stranded nucleotide over hangs. The ability to control F3'5'H synthesis in plants permits modulation of the composition of individual anthocyanins as well as alteration of relative levels of flavonols and anthocyanins, thereby enabling the manipulation of petal color. Moreover, the present invention extends to plants and reproductive or vegetative parts thereof 15 including flowers, fruits, seeds, vegetables, leaves, stems and the like. The present invention further extends to ornamental lransgenic or genetically modified plants. The term "transgenic" also includes progeny plants and plants from subsequent genetics and/or crosses thereof from the primary transgenic plants. .20 Accordingly, one aspect of the present invention provides an isolated nucleic acid molecule comprising a sequence of nucleotides encoding or complementary to a sequence encoding a F3'5'H or a polypeptide having F3'5'E activity wherein expression of said nucleic acid molecule in a rose petal tissue results in detectable levels of delphinidin or delphinidin-based molecules as measured by a chromatographic technique. 25 Another aspect of the present invention is directed to an isolated nucleic acid molecule comprising a sequence of nuoleotides encoding or complementary to a sequence encoding a F3'5'H or a polypeptide having F3'5'H activity wherein expression of said nucleic acid molecule in a rose petal tissue results in a sufficient level and length of transcript which is 30 translated to said F3'5'H as determined by detectable levels of delphinidin or delphinidin based molecules as measured by a chromatographic technique.

WO 2004/020637 PCT/AU2003/001111 -28 A further aspect of the present invention is directed to an isolated nucleic acid molecule comprising a sequence of nucleotides encoding or complementary to .a sequence encoding a F3'5'H or a polypeptide having F3'5'H activity wherein expression of said nucleic acid molecule in a rose petal tissue results a full-length transcript which is detectable by 5 Northem blot analysis of total RNA isolated from rose petals. The present invention is described and exemplified herein by reference to the identification, cloning and manipulation of genetic sequences encoding a F3'5'H which acts on DHK as well as DHQ. Preferably, the F3'5'H enzyme is a pansy, salvia , sollya 10 lavender or kennedia F3'5'H. The F3'5'H enzyme may also be considered to include a polypeptide or protein having a F3'5'H activity or F3'5'H-like activity. The latter encompasses derivatives having altered F3'5'H activities. A preferred aspect of the present invention, therefore, is directed to an isolated nucleic acid 15 molecule comprising a sequence of nucleotides encoding, or complementary to- a sequence encoding a F3'5'H or a functional mutant, derivative, part, fragment, homolog or analog thereof wherein the nucleic acid molecule is characterized by the following: (i) the F3'5'H transcript in rose petal tissue is of sufficient level and size to encode a 20 F3'5'H resulting in detectable delphinidin or delphinidin-based molecules in the rose petal tissue as measured by a chromatographic procedure (eg. TLC or HPLC); (ii) the F35'H transcript in rose petal tissue is full-length and detected by Northern blot analysis of total RNA isolated from rose petal tissue 25 (iii) the F3'5'H in rose petal tissue.results in detectable delphinidin or delphinidin-based molecules as measured by a ohromatographic procedure (eg. TLC or HOPLC); and/or (iv) the F3'5'H results in a visual color change in rose petal tissue. 30 WO 2004/020637 PCT/AU2003/001111 -29 The term delphinidin-based pigments includes the anthocyanidin, delphinidin or any derivatives thereof including but not limited to glycosylated, acylated, methylated or other modified forms. Methylated forms of delphinidin include but are not limited to the anthocyanidin petunidin (methylated at the 3'-position), malvidin (methylated at the 3' and 5 5' position), 5-0 methyl malvidin (methylated at the 5, 3' and 5' positions), 5, 7-0 dimethyl malvidin (methylated at the 5, 7, 3' and 5' positions). The methylated anthocyanidins can also be modified by glycosylation and acylation. The term anthooyanins defines glycosylated forms of the respective anthocyanidins. 10 By the term "nucleic acid molecule" is meant a genetic sequence in a non-naturally occurring condition. Generally, this means isolated away from its natural state or synthesized or derived in a non-naturally-occurring environment. More specifically, it includes nucleic acid molecules formed or maintained in vitro, including genomic DNA fragments recombinant or synthetic molecules and nneleic acids in combination with 15 heterologous nucleic acids. It also extends to the genomic DNA or cDNA or part thereof encoding F35'H or a part thereof in reverse orientation relative to its own or another promoter. It further extends to naturally occurring sequences following at least a partial purification relative to other nucleic acid sequences. 20 The term "genetic sequences" is used herein in its most general sense and encompasses any contiguous series of nucleotide bases specifying directly, or via a complementary series of bases, a sequence of amino acids in a F3'5'H enzyme. Such a sequence of amino acids may constitute a full-length F3'S'H such as is set forth in SEQ ID NO: 10 (pansy) or SEQ ID NO:12 (pansy) or SEQ ID NO;14 (salvia) or SEQ ID NO:16 (salvia) or SEQ ID 25 NO:18 (sollya) or SEQ ID NO:32 (lavender) or SEQ ID NO:27 (kennedia) or an active truncated form thereof or may correspond to a particular region such as an N-terminal, C tnninal or internal portion of the enzyme. A genetic sequence may also be referred to as a sequence of nuoleotides or a nucleotide sequence and includes a recombinant fusion of two or more sequences. 30 WO 2004/020637 PCT/AU2003/001111 -30 In accordance with the above aspects of the present invention there is provided a nucleic acid molecule comprising a nucleotide sequence or complementary nucleotide sequence substantially as set forth in SEQ ID NO:9 (pansy) or SEQ ID NO: 11 (pansy) or SEQ ID NO:13 (salvia) or SEQ ID NO:15 (salvia) or SEQ ID NO:17 (sollya) or SEQ ID NO:31 5 (lavender) or SEQ ID) NO:26 (cennedia) or having at least about 50% similarity thereto or capable of hybridizing to the sequence set forth in SEQ ID NO:9 or SEQ ID NO: 11 or SEQ ID NO:13 or SEQ ID NO:15 or SEQ ID NO:17 or SEQ ID NO:31 or SEQ ID NO:26 under low stringency conditions. 10 Table 1 provides a summary of the sequence identifiers. Alternative percentage similarities and identities (at the nucleotide or amino acid level). encompassed by the present invention include at least about 60% or at least about 65% or at least about 70% or at-least about 75% or at least about 80% or at least about 85% or at 15 least about 90% or above, such as about 95% or about 96% or about 97% or about 98% or about 99%, such as at least about 60%, 61%, 62%, 62%, 63%, 64%, 65%, 66%, 67%, 68%, 69%, 70%, 71%, 72%, 73%, 74%, 75%, 76%, 77%, 78%, 79%, 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, 100%. 20 In a particularly preferred embodiment, there is provided an isolated nucleic acid molecule comprising a nucleotide sequence or complementary nucleotide sequence substantially as set forth in SEQ ID NO:9_(pansy) or SEQ ID NO: II (pansy) or SEQ ID NO: 13 (salvia) or SEQ ID NO:15 (salvia) or SEQ ID NO:17 (sollya) or SEQ D NO:31 (lavender) or SEQ 25 ID NO:26 (kennedia) or having at least about 50% similarity thereto or capable of hybridizing to the sequence sot forth in SEQ ID NO:1 (petunia) or SEQ ID NO:3 (petunia) or complementary strands of either under low stringency conditions, wherein said nucleotide sequence encodes a polypeptide having a F3'5'H activity. 30 For the purposes of determining the level of stringency to define nucleic acid molecules capable of hybridizing to SEQ ID NO:1, SEQ ID NO:3, SEQ ID1 NO:9 or SEQ ID NO:1l WO 2004/020637 PCT/AU2003/001111 -31 or SEQ ID NO:13 or SEQ 1D NQ:15 or SEQ ID NO:17 or SEQ ID NO:31 or SEQ ID NO:26 reference herein to a low stringency includes and encompasses from at least about 0% to at least about 15% v/v formamide and from at least about IM to at least about 2 M salt for hybridization, and at least about 1 M to at least about 2 M salt for washing 5 conditions. Generally, low stringency is from about 25-30 0 C to about 42 0 C. The temperature may be altered and higher temperatures used to replace the inclusion of formamide and/or to give alternative stringency conditions. Alternative stringency conditions may be applied where necessary, such as medium stringency, which includes and encompasses from at least about 16% v/v to at least about 30% v/v formamide and 10 from at least about 0.5 M to at least about 0.9 M salt for hybridization, and at least about 0.5 M to at least about 0.9 M salt for washing conditions, or high stringency, which includes and encompasses from at least about 31% v/v to at least about 50% v/v formamide and from at least about 0.01 M to at least about 0.15 M salt for hybridization, and at least about 0.01 M to at least about 0.15 M salt for washing conditions. In general, 15 washing is carried out T. = 69.3 + 0.41 (G+C)% (Marmur and Doty, J. Mol. BioL. 5: 109, 1962). However, the Tm. of a duplex DNA decreases by PC with every increase of 1% in the number of mismatch base pairs (Bonner and Laskey, Eur. J. Biochem. 46: 83, 1974). Formamide is optional in these hybridization conditions. Accordingly, particularly preferred levels of stringency are defined as follows: low stringency is 6 x SSC buffer, 20 1.0% w/v SDS at 25-42"C; a moderate stringency is 2 x SSC buffer, 1.0% w/v SDS at a temperature in the range 20"C to 65*C; high stringency is 0.1 x SSC buffer, 0.1% w/v SDS at a temperature of at least 650C. Another aspect of the present invention provides a nucleic acid molecule comprising a 25 sequence of nucleotides encoding or complementary to a sequence encoding an amino acid sequence substantially as set forth in SEQ ID NO:10 or SBQ ID NO:12 or SEQ ID NO:14 or SEQ ID NO;16 or SEQ ID NO:18 or SEQ ID NO:32 or SEQ ID NO:27 or an amino acid sequence having at least about 50% similarity thereto. 30 The term similarity as used herein includes exact identity between compared sequences at the nucleotide or amino acid level. Where there is non-identity at the nucleotide level, WO 2004/020637 PCT/AU2003/001111 - 32 similarity includes differences between sequences which result in different amino acids that are nevertheless related to each other at the structural, functional, biochemical and/or conformational levels. Where there is non-identity at the amino acid level, similarity includes amino acids that are nevertheless related to each other at the structural, functional, 5 biochemical and/or conformational levels. In a particularly, preferred embodiment, nucleotide and sequence comparisons are made at the level of identity rather than similarity. Terms used to describe sequence relationships between two or more polynucleotides or 10 polypeptides include "reference sequence", "comparison window", "sequence similarity", "sequence identity", "percentage of sequence similarity", "percentage of sequence identity", "substantially similar" and "substantial identity". A "reference sequence" is at least 12 but frequently 15 to 18 and often at least 25 or above, such as 30 monomer units, inclusive of nucleotides and amino acid residues, in length. Because two polynucleotides 15 may each comprise (1) a sequence (i.e. only a portion of the complete polynucleotide sequence) that is similar between the two polynucleotides, and (2) a sequence that is divergent between the two polynucleotides, sequence comparisons between two (or more) polynucleotides are typically performed by comparing sequences of the two polynucleotides over a "comparison window" to identify and compare local regions of 20 sequence similarly. A "comparison window" refers to a conceptual segment of typically 12 contiguous residues that is compared to a reference sequence. The comparison window may comprise additions or deletions (i.e. gaps) of about 20% or less as compared to the reference sequence (which does not comprise additions or deletions) for optimal alignment of the two sequences. Optimal alignment of sequences for aligning a comparison window 25 may be conducted by computerized implementations of algorithms (GAP, BESTFIT, FASTA, and TFASTA in the Wisconsin Genetics Software Package Release 7.0, Genetics Computer Group, 575 Science Drive Madison, WI, USA) or by inspection and the best alignment (i.e. resulting in the highest percentage homology over the comparison window) generated by any of the various methods selected, Reference also may be made to the 30 BLAST family of programs as, for example, disclosed by Altschul el al. (Nucl Acids Res. 25: 3389-3402, 1997). A detailed discussion of sequence analysis can be found in Unit WO 2004/020637 PCT/AU2003/001111 -33 19.3 of Ausubel et a. ("Current Protocols in Molecular Biology" John Wiley & Sons Inc, 1994-1998, Chapter 15, 1998). The terms "sequence similarity" and "sequence identity" as used herein refers to the .extent 5 that sequences are identical or functionally or structurally similar on a nucleotide-by nucleotide basis or an amino acid-by-amino acid basis over a window of comparison. Thus, a "percentage of sequence identity", for example, is calculated by comparing two optimally aligned sequences over the window of comparison, determining the number of positions at which the identical nucleic acid base (e.g. A, T, C, G, I) or the identical amino 10 acid residue (e.g. Ala, Pro, Ser, Thr, Gly, Val, Leu, Ile, Phe, Tyr, Trp, Lys, Arg, His, Asp, Glu, Asn, Gln, Cys and Met) occurs in both sequences to yield the number of matched positions, dividing the number of matched positions by the total number of positions in the window of comparison (i.e., the window size), and multiplying the result by 100 to yield the percentage of sequence identity. For the purposes of the present invention, "sequence 15 identity" will be understood to mean the 'match percentage" calculated by the DNASIS computer program (Version 2.5 for windows; available from Hitachi Software engineering Co., Ltd., South San Francisco, California, USA) using standard defaults as used in the reference manual accompanying the software. Similar comments apply in relation to sequence similarity. 20 The nucleic acid molecules of the present invention may be further characterized by having, or previously having, prior to derivatization on overall lower AT content (or higher GC content) compared to a nucleic acid molecule which encodes a F3'5'H but which does not result in detectable intact transcript in rose petal tissue or, when expressed, does not 25 result in detectable delphinidin or delphinidin-based molecules, as measured by a chromatographic procedure such as TLC or HPLC. Furthermore, the % of A's or T's in the third position of a codon is also lower than other F4'5'H enzymes. Reference herein to a chromatographic procedure includes a related procedure. By "related" means a technically related procedure or a procedure which provides a similar result. Examples of related 30 procedures include other forms of chromatography (0g. gas chromatography), WO 2004/020637 PCT/AU2003/001111 -34 In addition, nucleotide sequences which do not express well in rose tissue may be modified such as in reducing overall % AT or at least reduce the levels of% AT in the third position of a codon. Such time expression in rose tissue is elevated. 5 The nucleic acid sequences contemplated herein also encompass oligonucleotides useful as genetic probes for amplification reactions or as antisense or Sense molecules capable of regulating expression of the corresponding gene in a plant. Sense molecules include hairpin constructs, short double stranded DNAs and RNAs and partially double stranded DNAs and RNAs which one or more single stranded nucleotide over hangs. An antisense 10 molecule as used herein may also encompass a genetic construct comprising the structural genomic or cDNA gene or part thereof in reverse orientation relative to its or another promoter. It may also encompass a homologous genetic sequence. An antisense or sense molecule may also be directed to terminal or internal portions of the gene encoding a polypeptide having a F3'5'H activity or to combinations of the above such that the 15 expression of the gene is reduced or eliminated. With respect to this aspect of the invention, there is provided an oIigonucleotide of 5-50 nucleotides such as 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 20 49, 50, 51, 52, 53, 54, 55 having substantial similarity to a part or region of a molecule with a nucleotide sequence set forth in SEQ ID NO:9 or SEQ ID NO: 11 or SEQ ID NO;13 or SEQ ID NO;1S or SEQ ID NO:17 or SEQ ID NO:31 or SEQ ID NO:26. By substantial similarity or complementarity in this context is meant a hybridizable similarity under low, alternatively and preferably medium and alternatively and most preferably high stringency 25 conditions specific for oligonucleotide hybridization (Sambrook et aL., Molecular Cloning: A Laboratory Manual, 2 1d edition, Cold Spring Harbor Laboratories, Cold Spring Harbor, NY, USA, 1989). Such an oligonucleotide is useful, for example, in screening for F3'5'H genetic sequences from various sources or for monitoring an introduced genetic sequence in a transgenic plant. The preferred oligonucleotide is directed to a conserved F3'5'H 30 genetic sequence or a sequence conserved within a plant genus, plant species and/or plant variety.

WO 2004/020637 PCT/AU2003/001111 -35 In one aspect of the present invention, the oligonucleotide corresponds to the 5' or the 3' end of the F3'5'H genetic sequences. For convenience, the'5* end is considered herein to define a region substantially between the start codon of the structural gene to a centre 5 portion of the gene, and the 3' end is considered herein to define a region substantially between the centre portion of the gene and the terminating codon of the structural gene. It is clear, therefore, that oligonucleotides or probes may hybridize to the 5' end or the 3' end or to a region common to both the 5' and the 3' ends. the present invention extends to all such probes. In one 'embodiment, the nucleic acid sequence encoding a F3'5'H or various functional derivatives thereof is used to reduce the level of an endogenous a F3'5-I (e.g. via co suppression or antisense-mediated suppression) or other post-transcriptional gene silencing (PTGS) processes including RNAi or alternatively the nucleic acid sequence encoding this 15 enzyme or various derivatives or parts thereof is used in the sense or antisense orientation to reduce the level of a F3'5'H. The use of sense strands, double or partially single stranded such as constructs with hairpin loops is particularly useful in inducing a PTGS response. In a further alternative, ribozymes, minizymes or DNAzymes could be used to inactivate target nucleic acid sequences. 20 Still a further embodiment encompasses post-transcriptional inhibition to reduce translation into polypeptide material. Still yet another embodiment involves specifically inducing or removing methylation. 25 Reference herein to the altering of a F3'5'H activity relates to an elevation or reduction in activity of up to 30% or more preferably of 30-50%, or even more preferably 50-75% or still more preferably 75% or greater above or below the normal endogenous or existing levels of activity. Such elevation or reduction may be referred to as modulation of a F3'5'H enzyme activity. Generally, modulation is at the level of transcription or translation of 30 F3'5'H genetic sequences.

WO 2004/020637 PCT/AU2003/001111 -36 The nucleic acids of the present invention may be a ribonucleic acid or deoxyribonucleic acids, single or double stranded and linear or covalently closed circular molecules. Preferably, the nucleic acid molecule is eDNA. The present invention also extends to other nucleic acid molecules which hybridize under low, preferably under medium and most 5 preferably under high stringency conditions with the nucleic acid molecules of the present invention and in particular to the sequence of nucleotides set forth in SEQ ID NO:9 or SEQ ID NO:l1 or SEQ ID NO:13 or SEQ ID NO:15 or SEQ ID NO!l7 or SEQ ID NO:31 or SEQ ID NO:26 or a part or region thereof In its most preferred embodiment, the present invention extends to a nucleic acid molecule having a nucleotide sequence set forth in SEQ 10 ID NO:9 or SEQ ID NO:11 or SEQ ID NO:13 or SEQ ID NO:15 or SEQ ID NO:17 or SEQ ID NO:31 or SEQ ID NO:26 or to a molecule having at least 40%, more preferably at least 45%, even more preferably at least 55%, still more preferably at least 65%-70%, and yet even more preferably greater than 85% similarity at the level of nucleotide or amino acid sequence to at least one or more regions of the sequence set forth in SEQ ID NO:9 or 15 SEQ ID NO:l 1 or SEQ ID NO:13 or SEQ ID NO:15 orSEQ ID N0:17 or SEQ ID NO:31 or SEQ ID NO:26 and wherein the nucleic acid encodes or is complementary to a sequence which encodes an enzyme having a F3'5'H activity. It should be noted, however, that nucleotide or amino acid sequences -may have similarities below the above given percentages and yet still encode a F3'5'H activity and such molecules may still be 20 considered in the scope of the present invention where they have regions of sequence conservation. The present invention further extends to nucleic acid molecules in the form of oligonucleotide primers or probes capable of hybridizing to a portion of the nucleic acid molecules contemplated above, and in particular those set forth in SEQ ID NO:9 or SEQ ID NO:ll or SEQ ID NO:13 or SEQ ID NO:15 or SEQ D) NO:17 or SEQ ID NO:31 or 25 SEQ ID NO:26, under low, preferably under medium and most preferably under high stringency conditions. Preferably the portion corresponds to the 5' or the 3' end of the gene. For convenience the 5' end is considered herein to define a region substantially between the start codon of the structural genetic sequence to a centre portion of the gene, and the 3' end is considered herein to define a region substantially between the centre portion of the 30 gene and the terminating codon of the structural genetic sequence. It is clear, therefore, that WO 2004/020637 PCT/AU2003/001111 -37 oligonucleotides or probes may hybridize to the 5' end or the 3' end or to a region common to both the 5' and the 3' ends. The present invention extends to all such probes. The term gene is used in its broadest sense and includes cDNA corresponding to the exons 5 of a gene. Accordingly, reference herein to a gene is to be taken to include! (i) a classical genomic gene consisting of. transcriptional and/or translational regulatory sequences and/or a coding region and/or non-translated sequences (i.e. introns, 5'- and 3'- untranslated sequences); or 10 (ii) mRNA or cDNA corresponding to the coding regions (i.e. exons) and 5'- and 3' untranslated sequences of the gene. The term gene is also used to describe synthetic or fusion molecules encoding all or part of 15 an expression product: In particular embodiments, the term nucleic acid molecule and gene may be used interchangeably. The nucleic acid or its complementary form may encode the full-length enzyme or a part or derivative thereof. By "derivative" is meant any single or multiple amino acid 20 substitutions, deletions, and/or additions relative to the naturally occurring enzyme and which retains a F3'5'H activity. In this regard, the nucleic acid.includes the naturally occurring nucleotide sequence encoding a F3'5'H or may contain single or multiple nucleotide substitutions, deletions and/or additions to said naturally occurring sequence. The nucleic acid of the present invention or its complementary form may also encode a 25 "part" of the F3'5'H, whether active or inactive, and such a nucleic acid molecule may be useful as an oligonucleotide probe, primer for polymerase chain reactions or in various mutagenic techniques, or for the generation of antisense molecules. Reference herein to a "part" of a nucleic acid molecule, nucleotide sequence or amino acid 30 sequence, preferably relates to a molecule which contains at least about 10 contiguous nucleotides or five contiguous amino acids, as appropriate.

WO 2004/020637 PCT/AU2003/001111 - 38 Amino acid insertional derivatives of the F3'5'H of the present invention include amino and/or carboxyl terminal fusions as well as intra-sequence insertions of single or multiple amino acids. Insertional amino acid sequence variants are those in which one or more 5 amino acid residues are introduced into a predetermined site in the protein although random insertion is also possible with suitable screening of the resulting product. Deletional variants are characterized by the removal of one or more amino acids from the sequence. Substitutional amino acid variants are those in which at least one residue in the sequence has been removed and a different residue inserted in its place. Typical 10 substitutions are those made in accordance with Table 3. TABLE 3 Suitable residues for amino acid substitutions Original residue Exemplary substitutions Ala Ser Arg Lys Asn Gln; His Asp Glu Cys Ser Gin Asn; Glu Glu Asp Gly Pro His Asn; Gln Ile Lea; Val Leu Ile; Val Lys Arg; Gin; Glu Met Leu; Ile; Val Phe Met; Lou; Tyr Ser Thr Thr Ser WO 2004/020637 PCT/AU2003/001111 -39 Original residue Exemplary substitutions Trp Tyr Tyr Tip; Phe Val lie; Leu; Met Where the F3'5'H is derivatized by amino acid substitution, the amino acids are generally replaced by other amino acids having like properties, such as hydrophobicity, hydrophilicity, electronegativity, bulky side chains and the like. Amino acid substitutions 5 are typically of single residues. Amino acid insertions will usually be in the order of about 1-10 amino acid residues and deletions will range from about 1-20 residues. Preferably, deletions or insertions are made in adjacent pairs, i.e. a deletion of two residues or insertion of two residues. 10 The amino acid variants referred to above may readily be made using peptide synthetic techniques well known in the art, such as solid phase peptide synthesis (Merifield, J Am. Chem. Soc. 85: 2149, 1964) and the like, or by recombinant DNA manipulations. Techniques for making substitution mutations at predetermined sites in DNA having known or partially known sequence are well known and include, for example, M13 15 mutagenesis. The manipulation of DNA sequence to produce variant proteins which manifest as substitutional, insertional or deletional variants are conveniently described, for example, in Sambrook et al. (1989, supra). Other examples of recombinant or synthetic mutants and derivatives of the F3'5'H enzyme 20 of the present invention include single or multiple substitutions, deletions and/or additions of any molecule associated with the enzyme such as carbohydrates, lipids and/or proteins or polypfptides. The terms "analogs" and "derivatives" also extend to any functional chemical equivalent 25 of a F3'5'H and also to any amino acid derivative described above. For convenience, reference to F3'5'H herein includes reference to any functional mutant, derivative, part, fragment, homolog or analog thereof.

WO 2004/020637 PCT/AU2003/001111 -40 The present invention is exemplified using nucleic acid sequences derived from pansy, salvia, sollya or lavender or kennedia since this represents the most convenient and preferred source. of material to date. However, one skilled in the art will immediately appreciate that similar sequences can be isolated from any number of sources such as other 5 plants or certain microorganisms. All such nucleic acid sequences encoding directly or indirectly a F3'5'H are encompassed by the present invention regardless of their source. Examples of other suitable sources of genes encoding F3'5'H include, but are not limited to Vitis spp., Babiana stricta, Pinus spp., Picea spp., Larix spp., Phaseolus spp., Vaccinium spp., Cyclamen app., Iris app., Pelargonium spp. Liparieae, Geranium spp., Pisum spp., 10 Lathyrus spp., Clitoria spp., Catharanthus spp., Malva spp., Mucuna spp., Vicia spp., Saintpaulia spp., Lagerstroemia spp., bouchina app., Plumbago app., Hypocalyptus spp., Rhododendron spp., Linum spp., Macroptilium spp., Hibiscus spp., Hydrangea spp., Cymbidium spp., Millettia spp,, Hedysarum spp., Lespedeza spp., Asparagus app. Antigonon spp., Freesia spp., Brunella spp., Clarkia spp., etc. 15 In accordance with the present invention, a nucleic acid sequence encoding a F3'5'H may be introduced into and expressed in a transgenic plant in either orientation thereby providing a means either to convert suitable substrates, if synthesized in the plant cell, ultimately into DHM, or alternatively to inhibit such conversion of metabolites by reducing 20 or eliminating endogenous or existing F3'5'H activity. The production of these 3', 5'-hydroxylated substrates will subsequently be converted to delphinidin-based pigments that will modify petal color and may contribute to the production of a bluer color. Expression of the nucleic acid sequence in the plant may be constitutive, inducible or developmental and may also be tissue-specific. The word "expression" is used in its 25 broadest sense to include production of RNA 6r of both RNA and protein. It also extends to partial expression of a nucleic acid molecule. According to this aspect of the present invention, there is provided a method for producing a transgenic flowering plant capable of synthesizing a F3'5'H, said method comprising 30 stably transfonning a cell of a suitable plant with a nucleic acid sequence which comprises a sequence of nucleotides encoding said F3'5'H under conditions permitting the eventual WO 2004/020637 PCT/AU2003/001111 -41 expression of said nucleic acid sequence, regenerating a transgenic plant from the cell and growing said transgenic plant for a time and under conditions sufficient to permit the expression of the nucleic acid sequence. The transgenio plant may thereby produce non indigenous F3'5'IH at elevated levels relative to the amount expressed in a comparable non 5 transgenic plant. Another aspect of the present invention contemplates a method for producing a transgenic plant with reduced indigenous or existing flavonoid 3', 5'-hydroxylase activity, said method comprising stably transforming a cell of a suitable plant with a nucleic acid 10 molecule which comprises a sequence of nueleotides encoding or complementary to a sequence encoding a F3'5'H activity, regenerating a transgenic plant from the cell and where necessary growing said transgenic plant under conditions sufficient to permit the expression of the nucleic acid. 15 Yet another aspect of the present invention contemplates a method for producing a genetically modified plant with reduced indigenous or existing F3'5H activity, said method comprising altering the F3'S'H gene through modification of the indigenous sequences via homologous recombination from an appropriately altered F3'S'H gene or derivative or part thereof introduced into the plant cell, and regenerating the genetically modified plant from 20 the cell. As used herein an "indigenous" enzyme is one, which is native to or naturally expressed in a particular cell. A "non-indigenous" enzyme is an enzyme not native to the cell but expressed through the introduction of genetic material into a plant cell, for example, 25 through a transgene. An "endogenous" enzyme is an enzyme produced by a cell but which may or may not be indigenous to that cell. In a preferred embodiment, the present invention contemplates a method for producing a fransgenic flowering plant exhibiting altered floral or inflorescence properties, said method 30 comprising stably transforming a cell of a suitable plant with a nucleic acid sequence of the present invention, regenerating a transgenic plant from the cell and growing said transgenic WO 2004/020637 PCT/AU2003/001111 -42 plant for a time and under conditions sufficient to pnnit the expression of the nucleic acid sequence. Alternatively, said method may comprise stably transforming a cell of a suitable plant with 5 a nucleic acid sequence of the present invention or its complementary sequence, regenerating a transgenic plant from the cell and growing said transgenic plant for a time and under conditions sufficient to alter the level of activity of the indigenous or existing F3'5'H. Preferably the altered level would be less than the indigenous or existing level of F3'5'H activity in a comparable non-transgenic plant. Without wishing to limit the present 10 invention, one theory of mode of action is that reduction of the indigenous F3'5'H activity requires the expression of the introduced nucleic acid sequence or its complementary sequence. However, expression of the introduced genetic sequence or its complement may not be required to achieve the desired effect: namely, a flowering plant exhibiting altered floral or inflorescence properties. 15 In a related embodiment, the present invention contemplates a method for producing a flowering plant exhibiting altered floral or inflorescence properties, said method comprising alteration of the flavonoid 3', 5'-hydroxylase gene through modification of the indigenous sequences via homologous recombination from an appropriately altered F3'5'H 20 gene or derivative or part thereof introduced into the plant cell, and regenerating the genetically modified plant from the cell. Preferably, the altered floral or inflorescence includes the production of different shades of blue or purple or red flowers or other colors, depending on the genotype and physiological 25 conditions of the recipient plant. Accordingly, the present invention extends to a method for producing a transgenic plant capable of expressing a recombinant gene encoding a F3'5'H or part thereof or which carries a nucleic acid sequence which is substantially complementary to all or a part of a 30 mRNA molecule encoding the F3'5'H, said method comprising stably transforming a cell of a suitable plant with the isolated nucleic acid molecule comprising a sequence of WO 2004/020637 PCT/AU2003/001111 -43 nucleotides encoding, or complementary to a sequence encoding, a F3'5'H, where necessary under conditions permitting the eventual expression of said isolated nucleic acid molecule, and regenerating a transgenic plant from the cell. By suitable plant is meant a plant capable of producing DKK and possessing the appropriate physiological properties 5 required for the development of the color desired. One skilled in the art will immediately recognise the variations applicable to the methods of the present invention, such as increasing or decreasing the expression of the enzyme naturally present in a target plant leading to differing shades of colors sudh as different 10 shades of blue, purple- or red. The present invention, therefore, extends to all transgenic plants or parts or cells therefrom of transgenic plants or progeny of the transgenic plants containing all or part of the nucleic acid sequences of the present invention, or antisense forms thereof and/or any homologs or 15 related forms thereof and, in particular, those transgenic plants which exhibit altered floral or inflorescence properties. The transgenic plants may contain an introduced nucleic acid molecule comprising a nucleotide sequence encoding or complementary to a sequence encoding a F3'5'H. Generally, the nucleic acid would be stably introduced into the plant genome, although the present invention also extends to the introduction of a F3'5'H 20 nucleotide sequence within an autonomously-replicating nucleic acid sequence such As a DNA or RNA virus capable of replicating within the plant cell. The invention also extends to seeds from such transgenic plants. Such seeds, especially if colored, are useful as proprietary tags for plants. Any and all methods for introducing genetic material into plant cells including but not limited to Agrobacterium-mediated transformation, biolistic particle 25 bombardment etc. are encompassed by the present invention. Another aspect of the present invention contemplates the use of the extracts from transgenic plants or plant parts or cells therefrom of transgenic plants or progeny of the transgenic plants containing all or part of the nucleic acid sequences of the present 30 invention and, in particular, the extracts from those transgenic plants when used as a flavouring or food additive or health product or beverage or juice or coloring.

WO 2004/020637 PCT/AU2003/001111 .. 44- Plant parts contemplated by the present invention includes, but is not limited to flowers, fruits, vegetables, nuts, roots, stems, leaves or seeds. 5 The extracts of the present invention may be derived from the plants or plant part or cells therefrom in a number of different ways including but not limited to chemical extraction or heat extraction or filtration or squeezing or pulverization. The plant, plant part or cells therefrom or extract can be utilized in any number of different 10 ways such as for the production of a flavouring (e.g. a food essence), a food additive (e.g. a stabilizer, a colorant) a health product (e.g. an antioxidant, a tablet) a beverage (e.g. wine, spirit, tea) or a juice (e.g. fruit juice) or coloring (e.g. food coloring, fabric coloring, dye, paint, tint). 15 A further aspect of the present invention is directed to recombinant forms of F3'5'H. The recombinant forms of the enzyme will .provide a source of material for research, for example, more active enzymes and may be useful in developing in vitro systems for production of colored compounds. 20 Still a further aspect of the present invention contemplates the use of the genetic sequences described herein in the manufacture of a genetic construct capable of expressing a F3'5'H or down-regulating an indigenous F3'5'H enzyme in a plant. The term genetic construct has been used intorchangably throughout the specification and 25 alims with the terms "fusion molecule", "recombinant molecule", "recombinant nucleotide sequence". A genetic construct ifnay include a single nucleic acid molecule comprising a nucleotide sequence encoding a singal protein or may contain multiple open reading frames encoding 2 or more proteins. It may also contain a promoter operably linked to I or more of the open reading frames. 30 WO 2004/020637 PCT/AU2003/001111 -45 Another aspect of the present invention is directed to a prokaryotic or eukaryotic organism carrying a genetic sequence encoding a F3'5'H extrachromasomally in plasmid form. The present invention further extends to a recombinant polypeptide comprising a sequence 5 of amino acids substantially as set forth in SEQ ID NO.:10 or SEQ ID NO: 12 or SEQ ID NO:14 or SEQ ID NO:16 or SEQ ID NO:18 or SEQ ID NO:32 or SEQ ID NO:27 or an amino acid sequence having at least about 50% similarity to SEQ ID NO:10 or SEQ ID NO:12 or SEQ ID NO:14 or SEQ ID NO:16 or SEQ ID NO:18 or SEQ ID NO:32 or SEQ ID NO:27 or a derivative of said polypeptide. 10 A "recombinant polypeptide" means a polypeptide encoded by a nucleotide sequence introduced into a cell directly or indirectly by buman intervention or into a parent or other relative or precursor of the cell. A recombinant polypeptide may also be made using cell free, in vitro transcription systems. The term "recombinant polypeptide" includes an 15 isolated polypeptide or when present in a cell or cell preparation. It may also be in a plant or parts of a plant regenerated from a cell which produces said polypeptide. A "polypeptide" includes a peptide or protein and isencompassed by the term "enzyme". 20 The recombinant polypeptide may also be a fusion molecule comprising two or more heterologous amino acid sequences. The present invention is further described by the following non-limiting Examples.

WO 2004/020637 PCT/AU2003/001111 -46 EXAMPLE 1 General methods In general, the methods followed were as described in Sambrook et al (1989, supra) or 5 Sambrook and Russell, Molecular Cloning: A Laboratory Manual 3"d edition, Cold Spring Harbor Laboratories, Cold Spring Harbor, NY, USA, 2001 or Plant Molecular Biology Manual ( 2 "4 edition), Gelvin and Schilperoot (eds), Kluwer Academic Pablisher, The Netherlands, 1994 or Plant Molecular Biology Labfax, Croy (ed), Bios scientific Publishers, Oxford, UK, 1993. 10 The cloning vectors pBluescript and PCR script were obtained f-om Stratagene, USA. pCR7 2.1 was obtained from Invitrogen, USA. E. can transformation 15 The Escherichia coli strains used were: DH5a supB44, A (lacZYA-ArgF)U169, (08O0acZAM 15), hsdR17(r 1 , mfl, recAl, endAl, gyrA96, thi-1, relAl, deoR. (Hanahan, J. Mol Blol. 166: 557, 1983) 20 XL1 -Blue supE44, hsdRl7(rV, mkt, recAl, endAl, gyrA96, thi-l, relAl, lac,[F'proAB, lac1, lacZAM15, Tn10(tet)] (Bullock et aL, Biotechniques 5: 376, 1987). .25 BL2I-CodonPlus-RIL strain ompT hsdS(Rb- mB-) dcm+ Tet gal endA Hto [argU ileY leuW Camr M15 E. colt is derived from E-coli K12 and has the phenotype Nal", Str', Rif, Thi-, Ara, Gai , Mtl~, F, RecA, Uvr, Lon+ 30 Transformation of the e toli strains was performed according to the method of Inoue et at, (Gene 96: 23-28, 1990).

WO 2004/020637 PCT/AU2003/001111 -47 Aezrobacterium tumefaciens strains and transformadons The disarmed Agrobacterium tumefaciens strain used was AGLO (Lazo et at. Bio/technology 9: 963-967, 1991). 5 Plasmid DNA was introduced into the Agrobacterium tumefaciens strain AGLO by adding 5 pg of plasmid DNA to 100 pL of competent AGLO cells prepared by inoculating a 50 mL LB culture (Sambrook et al., 1989, supra) and incubation for 16 hours with shaking at 28*C. The cells were then pelleted and resuspended in 0.5 mt of 85% (v/v) 100 mM 10 CaC1 2 /15% (v/v) glycerol, The DNA-Agrobacterium mixture was frozen by incubation in liquid N 2 for 2 minutes and then allowed to thaw by incubation at 37"C for 5 minutes. 'The DNA/bacterial mix was then placed on ice for a further 10 minutes. The cells were then mixed with imL of LB (Sambrook et al., 1989 supra) media and incubated with shaking for 16 hours at 28 0 C. Cells of A. tumefaciens carrying the plasmid were selected on LB 15 agar plates containing appropriate antibiotics such as 50 pg/mL tetracycline or 100 pg/mL gentamycin. The confirmation of the plasmid in A. tumefaciens was done by restriction endonuclease mapping of DNA isolated from the antibiotic-resistant transformants. DNA lirations 20 DNA ligations were carried out using the Amersham Ligation Kit or Promega Ligation Kit according to procedures recommended by the manufacturer. Isolation and purification of DNA fragments Fragments were generally isolated on a 1% (w/v) agarose gel and purified using the 25 QIAEX 11 Gel Extraction kit (Qiagen) or Bresaclean Kit (Bresatec, Australia) following procedures recommended by the manufacturer. Repair of overhangring ends after restriction endonuclease digestion Overhanging 5' ends were repaired using DNA polymerase (Klenow fragment) according 30 to standard protocols (Sambrook et al., 1989 supra). Overhanging 3' ends were repaired using T4 DNA polymerase according to standard protocols (Sambrook et a., 1989 supra).

WO 2004/020637 PCT/AU2003/001111 -48 Removal of Phosphoryl groups from nucleic acids Shrimp alkaline phosphatase (SAP) (USB) was typically used to remove phosphoryl groups from cloning vectors to prevent re-circularization according to the manufacturer's 5 recommendations. Palymierase Chain Reaction (PCR) Unless otherwise specified, PCR conditions using plasmid DNA as template included using 2 ng of plasmid DNA, 100 ng of each primer, 2 pL 10 miM dNTP mix, 5 p.L 10 x 10 Taq DNA polymerase buffer, 0.5 pL Taq DNA Polymerase in a total volume of 50 pL. Cycling conditions comprised an initial denaturation step of 5 minutes at 94 0 C, followed by 35 cycles of 940C for 20 sec, 500C for 30 sec and 72*C for 1 minute with a final treatment at 7200 for 10 minutes before storage at 4*C. 15 PCRs were performed in a Perkin Elmer GeneAmp PCR System 9600. 3 2 P-Labelling of DNA Prohes DNA fragments (50 to 100 ng) were radioactively labelled with 50 pCi of [i- 32 P]-dCTP using a Gigaprime kit (Geneworks). Unincorporated [a- 32 P]-dCTP was removed by 20 chromatography on Sephadex G-50 (Fine) columns or Microbiospin P-30 Tris chromatography columns (BioRad). Plasmid Isolation Single colonies were analyzed for inserts by inoculating LB broth (Sambrook et al., 1989, 25 supra) with appropriate antibiotic selection (e.g. 100 pg/mL ampicillin or 10 to 50 g/nL tetracycline etc.) and incubating the liquid culture at 37*C (for E. coli) or 29 0 C (for A. twnefaciens) for -16 hours with shaking. Plasmid DNA was purified using the alkali-lysis procedure (Sambrook et al., 1989, supra) or using The WizardPlus SV minipreps DNA purification system (Promega) or Qiagen Plasmid Mini Kit (Qiagen). Once the presence of 30 an insert had been determined, larger amounts of plasmid DNA were prepared from 50 mL overnight cultures using the alkali-lysis procedure (Sambrook et al., 1989, supra) or WO 2004/020637 PCT/AU2003/001111 -49 QIAfilter Plasmid Midi kit (Qiagen) and following conditions recommended by the manufacturer. DNA Sequence Analysis 5 DNA sequencing was performed using the PRISM (trademark) Ready Reaction Dye Primer Cycle Sequencing Kits from Applied Biosystems. The protocols supplied by the manufacturer were followed. The cycle sequencing reactions were performed using a Perkin Elmer PCR machine (GeneAmp PCR System 9600). Sequencing runs were generally performed by the Australian Genome Research Facility at The Walter and Eliza 10 Hall Institute of Medical Research (Melbourne, Australia) or in-house on an automated 373A DNA sequencer (Applied Biosystems). Sequences were analysed using a MacVectorm application (version 6.5.3) (Oxford Molecular Ltd., Oxford, England). 15 Homology searches against Genbank, SWISS-PROT and EMBL databases were performed using the FASTA and TFASTA programs (Pearson and Lipman, Proc. Nat. Acad. Sci. USA 85(8): 2444-2448, 1988) or BLAST programs (Altschul et al., J. Mo). BioL 215(3): 403-410, 1990). Percentage sequence similarities were obtained using LALIGN 20 program (Huang and Miller, Adv. Apple. Math. 12: 373-381, 1991) or ClustalW program (Thompson et al., Nucleic Acids Research 22: 4673-4680, 1994) within the MacVector"' application (Oxford Molecular Ltd., England) using default settings. Multiple sequence alignments were produced using ClustalW (Thompson et al., 1994, 25 supra) using default settings, WO 2004/020637 PCT/AU2003/001111 -50 EXAMPLE 2 Plant transformations Petunia hybrida transformations (Sw63 x Skr4) 5 As described in Holton et al. (1993a, supra) by any other method well known in the art. Rosa hybrida transformations As described in U.S. Patent Application No. 542,841 (PCT/US91/04412) or Robinson and Firoozabady (Scientia Horticulturae, 35: 83-99, 1993). Rout et al. (Scientia Horiculturae, 10 81: 201-238, 1999) or Marchant at aL (Molecular Breeding 4; 187-194, 1998) or by any other method well known in the art. Cuttings of Rosa hybrida were generally obtained from Van Wyk and Son Flower Supply, Victoria. 15 Dianthus caryophlius transformations International Patent Application No. PCT/US92/02612 (carnation transformation). As described in International Patent Application No. PCT/AU96/00296 (Violet carnation), Lu at al. (Bio/Technology 9: 864-868, 1991), Robinson and Firoozabady (1993, supra) or by 20 any other method well known In the art. - Cuttings of Dianthus caryophyllus ov. Kortina Chanel or Monte Lisa were obtained'from Van Wyk and Son Flower Supply, Victoria. 25 WO 2004/020637 PCT/AU2003/001111 - 51 EXAMPLE 3 Transgenic Analysis Color coding 5 The Royal Horticultural Society's Color Chart (Kew, UK) was used to provide a description of observed color. They provide an alternative means by which to describe the color phenotypes observed. The designated numbers, however, should be taken only as a guide to the perceived colors and should not be regarded as limiting the possible colors which may be obtained. 10 Chronatographic analysis Thin Layer Chromatography (TLC) and High Performance Liquid Chromatography (HPLC) analysis was performed generally as described in Brugliera et al. (Plant J 5, 81 92, 1994). 15 Extraction of anthoevanidins Prior to HPLC analysis, the anthocyanin and flavonol molecules present in petal and stamen extracts were acid hydrolysed to remove glycosyl moieties from the anthocyanidin or flavonol core. Anthocyanidin and flavonol standards were used to help identify the 20 compounds present in the floral extracts. Anthocyanidins in the reaction mixture were analysed by HPLC via gradient elution using gradient conditions of 50%B to 60%B over 10 minutes, then 60% B for 10 minutes and finally 60% B to 100% B over 5 minutes where solvent A consisted of TFA: H20 (5:995) 25 and solvent B consisted of acetonitrile: TFA: H20 (500:5:495). An Asahi Pac ODP-50 cartridge column (250 mm x 4.6 mmn ID) was used for the reversed phase chromatographic separations. The flow rate was 1 mL/min and the temperature was 40"C. The detection of the anthocyanidin compounds was carried out using a Shimadzu SPD-M6A three dimensional detector at 400-650 un. 30 WO 2004/020637 PCT/AU2003/001111 -52 The anthocyanidin peaks were identified by reference to known standard, viz delphinidin or delphinidin-based molecules, petunidin, malvidin, cyanidin and peonidin Stages of flower development 5 Petunia Petunia hybrida cv. Skr4 x Sw63 flowers were harvested at developmental stages defined as follows: 10 Stage 1: Unpigmented, closed bud. Stage 2: Pigmented, cosed bud. Stage 3: Pigmented bud with emerging corolla Stage 4: Pigmented, opened flower with anther intact (pre-dehiscence) Stage 5: Fully opened flower with all anthers dehisced. 15 For TLC or HPLC analysis, petals were collected from stage 4 flowers at the stage of maximum pigment accumulation. For Northern blot analysis, petals were collected from stages 2 to 3 flowers at the stage of 20 maximal expression of flavonoid pathway genes. Carnation Dianthus caryophyllus flowers were harvested at developmental stages defined as follows: 25 Stage 1: Closed bud, petals not visible. Stage 2: Flower buds opening: tips of petals visible. Stage 3: Tips of nearly all petals exposed, "Paint-brush stage", Stage 4: Outer petals at 45' angle to stem. Stage 5: Flower fully open. 30 WO 2004/020637 PCT/AU2003/001111 -53 For TLC or HPLC analysis, petals were collected from stage 4 flowers at the stage of maximanm pigment accumulation. For Northern blot analysis, petals were collected from stage 3 flowers at the stage of 5 maximal expression of flavonoid pathway genes. Rose Stages of Rosa hybrida flower development were defined as follows: 10 Stage 1: Unpigmented, tightly closed bud. Stage 2: Pigmented, tightly closed bud Stage 3: Pigmented, closed bud; sepals just beginning to open. Stage 4: Flower bud beginning to open; petals heavily pigmented; sepals have separated. 15 Stage 5: Sepals completely unfolded; some curling. Petals are heavily pigmented and unfolding. For TLC or HPLC analysis, petals were collected from stage 4 flowers at the stage of maximum pigment accumulation. 20 For Northern blot analysis, petals were collected from stage 3 to 4 flowers at the stage of maximal expression of flavonoid pathway genes (Tanaka et at, Plant Cell Physiol., 36(6): 1023-1031, 1995). 25 Anthacyanin/flavonol measurement by spectrophotometric measurements Approximately 200mg of fresh petal tissue was added to 2 mL of methanol/1% (v/v) HCI and incubated for -16 hours at 4'C. A I in 20 dilution (e.g. 50 pL made to 1000 pL) was then made and the absorbance at 350 urn and $30 nm was recorded. 30 The approximate flavonols and anthocyanin amounts (m-noles/gram) were then calculated according to the following formulae: WO 2004/020637 PCT/AU2003/001111 -54 AnIhocyanin content

(AS

3 , / 34,000) x volume of extraction buffer (mL) x dilution factor x 10' mass of petal tissue (grams) 5 Flavonol content (A 1 114,300) x volume of extraction buffer (mL) x dilution factor x 100 mass of petal tissue (grams) 10 Northern/RNA blot analysis Transcription of a transferred gene was monitored by isolating RNA and estimating the quantity and size of the expected transcript. Northern blot analysis was used to monitor the steady-state level of particular transcripts in petals. A transcript was determined to be intact or full-length based on the estimated size expected from the gene used. In general 15 when cDNAs were used as coding sequences the size of the transcript expected would be the size of the cDNA plus any 5' untranslated component of the fused promoter fragment plus any 3' untranslated sequence from the fused terminator fragment. In some cases where a cDNA region contained a putative polyadenylation site and the terminator region contained a putative polyadenylation site, 2 transcripts would be detected. One would be of 20 a size consistent with polyadenylation occurring just downstream from the polyadenylation site within the cDNA sequence. The second transcript would be larger and consistent with the transcript being polyadenylated after the polyadenylation site within the terminator fragment. 25 Total RNA was isolated from petals or leaves using a Plant RNAeasy kit (QIAGEN) following procedures recommended by the manufacturer. For rose samples 1% (w/v) PVP was added to the extraction buffer.

WO 2004/020637 PCT/AU2003/001111 RNA samples (5 xg) were electrophoresed through 2.2 M formaldehyde/1.2% w/v agarose gels using running buffer containing 40 mM morpholinopropanesulphonie acid (pH 7.0), 5 mM sodium acetate, 0.1 mM EDTA (pH 8.0). The RNA was stained with ethidiumn bromide and visualised under UV-light. The ribosomal RNA was generally used as a guide 5 in conflring that the RNA had 'not been degraded by intra- or extra- cellular ribonucleases. The RNA was transferred to Hybond-N membrane filters (Amersham) and treated as described by the manufacturer. Control samples were included on RNA gels as a measure of the integrity of the 10 radiolabelled probe and as guides to expected transcript sizes. Controls for perHf and petH2 genes included RNA isolated from petunia OGB petals (stages 3 to 4) or from flowers of transgenic carnations shown previously to accumulate petHfl transcripts. Controls for other F3'5'H genes generally included RNA isolated from petals of the same species from which the F3'5'H sequence had been isolated. 15 RNA blots were probed with 32 P-labelled fragments. Prehybridization (1 hour at 42 0 C) and hybridization (16 hours at 42C) of the membrane filters were carried out in 50% v/v formamide, 1 M NaCl, 1% w/v SDS, 10% w/v dextran sulphate. The membrane filters were generally washed in 2 x SSC, 1% w/v SDS at 65"C for between I to 2 hours and then 20 0.2 x SSC, 1% w/v SDS at 65*C for between 0,5 to 1 hour. Membrane filters were generally exposed to Kodak XAR film with an intensifying screen at -70'C for 16 to 72 hours. EXAMPLE 4 25 Introduedon of chimeric petunia F3'S'H genes into rose As described in the introduction, the pattern of hydroxylation of the B-ring of the anthocyanidin molecule plays a key role in determining petal color. The production of the dihydroflavonol DHM, leads to the production of the purple/blue delphinidin-based 30 pigments in plants such as petunia. The absence of the F3'5'H activity has been correlated WO 2004/020637 PCT/AU2003/001111 -56 with the absence of blue flowers in many plant species such as Rosa, Gerbera, Antirrhinum, Dianthus and Dendranthema. Based on success in producing delphinidin-based pigments in a mutant petunia line 5 (Holton et al, 1993a, supra and International Patent Application No. PCT/AU92/00334), in tobacco flowers (International Patent Application No. PCT/AU92/00334) and in carnation flowers (International Patent Application No. PCT/AU96/00296), similar chimeric petunia F3'5'B genes were also introduced into roses in order to produce novel delphinidin-based pigments and modify flower color. 10 Preparation of cidmeric petania Ff3'5'Hene constructs A sunmary of promoter, terminator and coding fragments used in the preparation of constructs and the respective abbreviations is listed in Table 4. 15 TABLE 4 Abbreviations used in construct preparations ABBREVIATION DESCRIPTION 1.2 kb promoter fragment from the Antirrhinum majus chalcone AmCHS 5' synthase (CHS) gene (Sommer and Saedler, Mol Gen. Gent., 202: 429-434, 1986) -0.2 kb incorporating BglII fragment containing the promoter CaMV3ss region from the Cauliflower Mosaic Virus 35S (CaMV 35S) gere (Franck et at, Cell 21: 285-294, 1980, Guilley et at, Cell, 30: 763-773. 1982) promoter fragment from CaMV 3538 gene (Franck et at, 1980, 35S5' supra) with an -60bp 5' untranslated leader sequence from the petunia chlorophyll a/b binding protein gene (Cab 22 gene) (Harpster et al., MGG, 212: 182-190, 1988) chrys CR8 5' promoter region from a CHS gene from chrysanthemum (SEQ 1D NO: 30) WO 2004/020637 PCT/AU2003/001111 -57 ABBREVIATION DESCRIPTION enhanced CaMY 35S promoter as described in Mitsuhara et at., sCaMVt 35S Plant Cell Physiol. 37: 49-59, 1996 GUS p-glucuronidase (GUS) coding sequence (Jefferson, et al, EMBO J. 6: 3901-3907, 1987) Hybrid promoter consisting of the promoter from the mannopine Mac synthase (mas) gene and a CaMV' 353 enhancer region (Comai et al., Plant Mol. Biol. 15: 373-381, 1990) Hybrid promoter consisting of a promoter region from CaMV 35S gene with enhancer elements from a promoter fragment of mas/35S marmopine synthase (mas) gene of Agrobaclerium tumefaciens (Janssen and Gardner, Plant Molecular Biology, 14: 61-72, 1989) mas 5' Promoter region from the mas of A. tuinefaciens mas 3' Terminator region from the mas gene of A. tumefaciens Promoter region from the nopalino synthase (nos) gene of A. nos 5' tumefaciens (Depicker et a., J Mol. and App. Genetics 1: 561 573, 1982) nos 3' Terminator region from the nos gene of A. tumefaciens (Depicker et al:, 1982, supra) Kanamycin-resistance gene (encodes neomycin np tI phosphotransferase which deactivates aminoglycoside antibiotics such as kanamycin, neomycin and G418) ocs 3' -1.6kb terminator fragment from octopine synthase gene of A. tumefaciens (described in Janssen and Gardner, 1989, supra) 3.2kb promoter region from a phospholipid transfer protein gene petD8 S (DS) of Petunia hybrida (Holton, Isolation and characterization of petal specific genes from Petunia hybrida. PhD thesis, University of Melboume, Australia, 1992) (SEQ 1D NO: 24) -0.7kb terminator region from a phospholipid transfer protein petD8 3' gene (D8) of Petunia hybrida cv. OGB (Holton, 1992, supra) WO 2004/020637 PCT/AU2003/001111 -58 ABBREVIATION DESCRIPTION -4.0kb fragment containing the promoter region from a flavonol synthase (FLS) gene of P. hybrida -2.2kb fragment containing the promoter region from FLS gene of short petMFL S ' P. hybrida -0.95kb fragment containing the terminator region from FLS gene of P. hybrida Petunia F3'S'H Hfl cDNA clone (Holton et aL, 1993a, supra) (SEQ ID NO: 1) Petunia F35'H H2 cDNA clone (Holton et a, 1993a, supra) (SEQ ID NO: 3) Promoter region of an anthocyanidin-3- glucoside peiRT 5rhamnosyltransferase (3RT) gene from P. hybrida (Brugliera, Charatterization of floral specific genes isolated from Petunia hybrida. RM1T, Australia. PhD thesis, 1994) Terminator region of a 3RT gene from P. hybrida (Brgliera, 1994, supra) RoseCHS 5' -2.8kb fragment containing the promoter region from a CHS gene of Rosa hybrida (SEQ ID: 5) Chlorsulfuron-resistance gene (encodes Acetolactate Synthase) SuRB with its own terminator from Nicotiana tabacum (Lee et at., EMBOJ. 7:1241-1248, 1988) In order to produce delphinidin or delphinidin-based molecules in rose petals, a number of binary vector constructs were prepared atilising the petunia F3 TH cDNA fragments and various promoter and terminator fragments. The chimaeric petunia F35'H genes had 5 proved successful in carnation and petunia leading to detectable intact F3'S 'H transcripts (as detected by Northern blot analysis) and to the production of delphinidin or delphinidin based molecules pigments. Table 5 summarises the list of binary vector constructs containing petunia F3'5'H cDNA fragments.

WO 2004/020637 PCT/AU2003/001111 -59 TABLE 5 Summary of chimueric petunia F3'5I gene expression cassettes contained in binary vector constructs used in the transfonnation of roses (see Table 4 for an explanation of abbreviations). SELECTABLE ARKER PLASMID F3'5'H GENE GENE GENE pCGP1452 AmCHS 5': petHfl: petD8 3' 35S 5': SuRB pCGP1453 Mac: petHfff: mas 3' 355 5': SuRB pCGP1457 petD8 5': petHf: petD8 3' 355 5': SuRB pCGP1461 short petFLS 5': petifi: pet FLS 3' 35S 5': SuRB pCGP1616 petRT S': petfl: nos 3' 35S 5': SuRB pCGP1638 CaMY 35S: petHf: ocs 3' 35S 5': SuRB pCGP1623 mas 35S: petHfl: acs 3' 35S 5': SuRB pCGP1860 RoseCHS 5': petHfl: nos 3' 35S 5': SuRB pCGP2123 CaMV 35S: petHf2: ocs 3' 35 . ': SuRB 5 Isolation of petunia F3'5'H eDNA clones (petH fl and petM 2) The isolation and characterisation of cDNA clones of petunia F3'S'H (petHfl and petHJ2 contained in pCGP602 (Figure 2) and pCGPI 75 (Figure 3) respectively) (SEQ ID NO: 1 and SEQ ID NO:3, respectively) have been described in International Patent Application 10 No. PCTIAU92/00334 and Holton et al. (1993a, upra). The plasmids pCGP601 (Figure 2), pCOP602 (Figure 2), pCGP176 (Figure 2) contain homologs of the petunia petHf F3'5'H cDNA clone. The plasinid pCGP601 contains a petunia F3'5'H petHfl homolog that includes 52bp of 5' untranslated sequence. The 15 plasmid pCGP602 contains a petunia F3'5'H petHf homolog that includes 125bp of 5' untranslated sequence (SEQ ID NO: 1). The plasmid pCGP176 (described in Holton et aL, 1993a supra) contains a petunia F3'5'H petHf homolog that includes 27bp of 5' untranslated sequence and a further -127bp of 3' untranslated sequence over the petunia F3'5'HpetHJfl cDNA clone in pCGP602. 20 WO 2004/020637 PCT/AU2003/001111 -60 Construction of pCGP1303 (petff in nUC19 backbone) The petunia F3'5'H cDNA clone contained in the plasmid pCGP601 (described above) (Figure 2) included 52 bp of 5' untranslated sequence and 141 bp of 3' untranslated sequence including 16 bp of the poly A tail. The plasmid pCGP601 (Figure 2) was firstly 5 linearized by digestion with the restriction endonuclease BspHI. The ends were repaired and the petunia F3'5 2petHf cDNA clone was released upon. digestion with the restriction endonuclease FspI. The BspHI recognition sequence encompasses the putative translation initiating codon and the PspI recognition sequence commences 2 bp downstream from the stop codon, The 1.6 kb fragment containing the petunia F3'5"H 10 petHfl cDNA clone was purified and ligated with repaired EcoRI ends of pUC19 (New England Biolabs). Correct insertion of the fragment was established by restriction endonuclease analysis of DNA isolated from ampicillin-resistant transformants. The resulting plasmid was designated as pCGP1303 (Figure 4). 15 Construction of pCGP627 (short petH(1 in pBluesript backbone) The plasmid pCGP176 (Holton et al, 1993a, supra) (Figure 2) was digested with the restriction endonuclease SpeI and EcoR. The ends were then repaired and allowed to religate. The resulting plasmid was designated as pCGP627 and contained the identical cDNA clone as in pCGP176 except that the restriction endonuclease sites Pstl, BamHl and 20 Smal were removed from the multi-cloning site of the pBluescript vector at the 5' end of the cDNA clone. The binary vector pCGPI452 (AmiCHS 5: petH(1: peD8 3,) The plasmid pCGP1452 (Figure 5) contain's a chimaeric petunia F3'5'H (petHf) gene 25 under the control of a promoter fragment from the Antirrhinum majus chatcone synthase gene (CHS) (Sommer and Saedler, 1986, supra) with a terminator fragment from the petunia phospholipid transfer protein (PLTP) gene (petDS 3) (Holton, 1992, supra). The chimeric petunia 173''H cassette is in a tandem orientation with respect to the 35S 5' SuRB gene of the binary vector, pWTT2132 (DNA Plant Technologies, USA = DNAP) 30 (Figure 6).

WO 2004/020637 PCT/AU2003/001111 -61 Intermediates in the preparation of the binary pCGP145 2 The binary vector PWTT2132 The binary vector plasmid pWTT2132 (DNAP) (Figure 6) contains a chimeric gene comprised of a 35S 5' promoter sequence (Franck et al., 1980, supra), ligated with the 5 coding region and terminator sequence for acetolactate syntbase (ALS) gene from the SuRB locus of tobacco (Lee et al., 1988, supra). An -60bp 5' untranslated leader sequence from the petunia chlorophyll a/b binding protein gene (Cab 22 gene) (Harpster et al., MGG, 212: 182-190, 1988) is included between the 3555' promoter fragment and the SuRB sequence. 10 Construction of pCGP725 (Am CHS 5': petH(1: petD8 3'in pfllueseript) A chimerio petunia F3'5'H gene under the control Antirrhinum majus CHS (AmCHS 5) promoter with a petunia PLTP terminator (petD8 3') was constructed by cloning the 1.6kb BclI/FspI petunia P3'5'H (petHfl) fragment from pCGP602 (Holton et al., 1993a, supra) 15 (Figure 2) between a 1.2 kb Antirrhinum majus CHS gene fragment 5' to the site of translation initiation (Sommer and Saedler, 1986, supra) and a 0.7 kb SmaI/AhoI PLTP fragment (petD8 3) from pCGP13ABam (Holton, 1992, supra), 3' to the deduced stop codon. The resulting plasmid in a pBluescript II KS (Stratagene, USA) backbone vector was designated pCGP725 (Figure 7). 20 Construction ofn CGP485 and oCGP1452 (Am CHS 5': petH(1: petD8 3' binary vectors) The chimeric F3'5'H gene from pCGP725 (Figure 7) was cloned into the binary vector pCGN1547 containing an nptll selectable marker gene cassette (McBride and Summerfelt Plant Molecular Biology 14: 269-276, 1990) to create pCGP485. A 3.5 kb fragment 25 containing the AmCHS 5'. petlifi: petD8 3' cassette was released upon digestion of pCGP485 with the restriction endonuclease PsAt. The overhanging ends were repaired and the purified 3.5.kb fragment was ligated with SmaI ends of the binary vector, pWTT2132 (DNAP). Correct insertion of the fragment in a tandem orientation with respect to the 35S 5': SuRB selectable marker gene cassette was established by restriction endonuclease 30 analysis of plasmid DNA isolated from tetracycline-resistant transformants. The plasmid was designated as pCGP1452 (Figure 5).

WO 2004/020637 PCT/AU2003/001111 - 62 Plant transformation with pCGP1452 The T-DNA contained in the binary vector plasmid pCGP1452 (Figure 5) was introduced into rose via Agrobacterium-mediated transformation. 5 The binary vector pCGPI453 (Mac: petHfl: mas 3) The plasmid pCGP 1453 (Figure 8) contains a chimeric petunia F3'5'H (pet/fl) gene under the control of a Mac promoter (Comai et al, 1990, supra) with a terminator fragment from the mannopine synthase gene of Agrobacterium (mas 3'. The chimeric petunia F3'S'H 10 cassette is in a tandem orientation with respect to the 355': SuRB selectable marker gene cassette of the binary vector, pWTTZ132 (DNAP) (Figure 6). A 3.9 kh fragment containing the Mac: peiHf: mas 3' gene was released from the plasmid pCGP6 2 8 (described in International Patent Application No. PCTIAU94/00265) upon 15 digestion with the restriction endonuclease PsiT. The overhanging ends were repaired and the purified fragment was ligated with Smai ends of pWTT2132 (DNAP). Correct insertion of the Mac: petHf: mas 3' gene in a tandem orientation with respect to the 358 5': SuRB selectable marker gene cassette was established by restdction endonuclease analysis of plasmid DNA isolated from tetracycline-resistant transformants. The plasmid was 20 designated as pCGP1453 (Figure 8). Plant transformation with pCGP1453 The T-DNA contained in the binary vector plasmid pCGP1453 (Figure 8) was introduced into rose via Agrobacterium-mediated transformation. 25 WO 2004/020637 PCT/AU2003/001111 - 63 The binary vector pCGPI457 (petD8 5'; petift: pet D 3) The plasmid pCGP1457 (Figure 9) contains a ehimaeric petunia F3'5'H (petHff) gene under the control of a promoter fragment from the petunia PLTP gene (petDS 5) with a terminator fragment from the petunia PLTP gene (petD8 3). The chimeric petunia F3'5'U 5 cassette is in a tandem orientation with respect to the 35S 5': SuRB gene of the binary vector, pWTT2132 (DNA?) (Figure 6). Intermediates in the preparation of the binary vector pCGP1457 Isolation of petania D8 zenomic clone 10 Preparation of . hybrida cv. OGB (Old Glory Blue) genomic library in 42001 A genoinic DNA library was constmcted from Petunia hybrida cv. OGB DNA in the vector 22001 (Kam et aL., Gene 32: 217-224, 1984) using a Sau3A partial digestion of the genomic DNA as described in Holton, 1992 (supra). Screening of the OGB genomic library for the petunia D8 gene was as described in Holton, 1992. supra. 15 Isolation ofD8 genomic clone OGB2.6 PCR was performed in order to find a non-mutant genomic clone representing D8. Oligo #2 (5' to 3' GTTCTCGAGGAAAGATAATACAAT) (SEQ ID NO:6) and Oligo #4 (5' to 3' CAAGATCGTAGGACTGCATG) (SEQ ID NO:7) were used to amplify D8 gene fragments, 20 across the intron region, using 4 pL of phage suspension from the clones isolated from the primary screening of the OGB genomic library. The reactions were carried out in a total volume of 50 1 rL containing I x Amplification buffer (Cetus), 0.2 mM dNTP mix, <1 jg of template DNA, 50 pmoles of each primer and 0.25 pL of Taq polymerase (5 units/pL Cetus). The reaction mixtures were overlaid with 30 pL of mineral oil and temperature cycled 25 using a Gene Machine (Innovonics). The reactions were cycled 30 times using the following conditions: 94'C for I minute, 55 0 C for 50 seconds, 72*C for 2 minutes. One quarter of each PCR reaction was min on an agarose gel using TAB running buffer. Three clones, XOCB-2.4, 10OB-2.5, and XOGB-2.6, gave fragments of approximately I kb 30 whereas the mutant clone, XOGB-32 (described in Holton, 1992, supra), had produced a product of 1.25 kb. The X0GB-2.6 clone was chosen for further analysis, WO 2004/020637 PCT/AU2003/001111 -64 pCGP382 The genornic clone, 2OGB-2.6, contained a single 3.9 kb XbaI fragment that hybridized with the D8 rDNA. This Xbal fragment was isolated and purified and ligated with the Xbal ends of 5 pBluescriptll KS- (Stratagene, USA). Restriction mapping of this clone revealed an internal PstI site 350 hp from the 3' end. However, the 'mutant" genomic clone in pCGP13, bad an internal PstI near the putative initiating "ATG" of the coding region (approximately 1.5 kb from its 3' end). The difference in the position of the Pstl site in both clones suggested that the LOGB-2.6 Xbal fragment did not contain the whole genomic sequence of D8. A Southen 10 blot was performed on PstI digested .0GB-2.6 DNA, and a fragment of 2.7 kb was found to hybridize with the D8 cDNA. Restriction endonuclease mapping confirmed that this fragment contained the 3' coding region and flanking sequences. In order to obtain a fragment containing the whole D8 genomic sequence, a number of 15 cloning steps were undertaken. The AOGB-2.6 PstI fragment of 2.7 kb was purified and ligated with PstI ends of pfluescriptil KS- (Stratagene, USA). The resultant clone was digested with Xba to remove the 350 bp PstI/Xbal fragment. This fragment was replaced by the 3.9 kb Xbal fragment frorn kOGB-2.6 to produce the plasmid pCGP382. 20 A 3.2 kb fragment containing the promoter region from the D8 2.6 gene in pCGP382 was released upon digestion with the restriction endonucleases HindIII and NcoL The fragment was purified and ligated with the 4.8 kb NcoIlHindfI fragment of pJB1 (Bodeau, Molecular and genetic regulation of Bronze-2 and other maize anthocyanin genes. Dissertation, Stanford University, USA, 1994) to produce pCGPi101 containing a 25 petD85': GUS: nos 3' cassette. A 1.6 kb petunia F3'5'H petHfl fragment was released from the plasnid pCOP602 (Holton et al,, 1993a, supra) (SEQ ID NO: 1) (Figure 2) upon digestion with the restriction endonucleases EspHi and BamHfL. The fragment was purified and ligated with the 6.2 kb 30 Ncol/BamHl fragment of pCGP1 101 to produce pCGP1 102 containing apetD8 5': petHfI: nos expression cassette.

WO 2004/020637 PCT/AU2003/001111 -65 A 0.75 kb BamHl petD8 3' fragment (Holton, 1992, supra) was purified from the plasmid pCGP13ABamHI and ligated with BamHI/BglIl ends of pCGP1102 to produce the plasmid pCGPI107 containing apetD8 5': petlfl: peD8 3' expression cassette. 5 The plasmid pCGP1107 was linearised upon digestion with the restriction endonuclease XbaL. The overhanging ends were repaired and then the 5.3 kb fragment containing the petD8 5. petHfi: petD8 3' expression cassette was released upon digestion with the restriction endonucleaso Pst. The fragment was purified and ligated with SmaI/Pstl ends 10 of the binary vector pWTT2132 (DNAP) (Figure 6). Correct insertion of the petD8 5': peif: petD8 3'gene in a tandem orientation with respect to the 35S 5': SuRB selectable marker gene cassette was established by restriction endonuclease analysis of plasmid DNA isolated from tetracycline-resistant transformants. The plasmid was designated as pCGP1457 (Figure 9). 15 Plant transformation with pCGPJ457 The T-DNA contained in the binary vector plasmid pCGP1457 (Figure 9) was introduced into rose via Agrobacterium-mediated transformation, 20 The binary vector PCGP1461 (short PetFLS 5': PetH.f: Pet FLS 37) The plasmid pCGP1461 (Figure 10) contains a chimeric petunia F3'5'J (petRHf) gene under the control of a promoter fragment from the petunia flavonol synthase (FLS) gene (short petFLS 5') with a terminator fragment from the petunia FLS gene (petFLS 3'). The chimerio petunia F3'5'S gene is in a tandem orientation with respect to the 35S 5': SuRB 25 gene of the binary vector, pWTT2132 (Figure 6).

WO 2004/020637 PCT/AU2003/001111 - 66 Intermediates in the preparation of the binary vector pCGP1461 Isolation of petunia FLS rene Preparation of P, hybrida cv. Th7 genoinic library A P. hyorida ov, Th7 genomic library was prepared according to Sambrook et al. (1989, 5 supra) using a Sau3A partial digestion of the genomic DNA. The partially digested DNA was cloned into EMBL-3 lambda vector (Stratagene, USA). The Th7 genomic DNA library was screened with 32 P-labelled fragments of a petunia FLS cDNA clone (Holton et al., Plant J. 4: 1003-1010, 1993b) using high stdugency 10 conditions. Two genomic clones (PLS2 and FLS3) were chosen for further analysis and found to contain sequences upstream of the putative initiating methionine of the petunia FLS coding region with PLS2 containing a longer promoter region than FLS3. 15 pCGP486 A 6 kb fragment was released upon digestion of the genomic clone FLS2 with the restriction endonuclease AoL The fragment containing the short petunia FLS gene was purified and ligated with AoI ends of pBluescript SK (Stratagene, USA). Correct insertion 20 of the fragment was established by restriction endonuclease analysis of DNA isolated from ampicillin-resistant transformants, The resulting plasmid was designated as pCGP486. pCGP487 A 9 kb fragment was released upon digestion of the genomic clone FL33 with the 25 restriction endonuclease Xhol. The fragment containing the petunia FLS gene was purified and ligated with Xhol ends of pBluesoript SK (Stratagene, USA). Correct insertion of the fragment was established by restriction endonuclease analysis of DNA isolated from amnpicillin-resistant transformants. The resulting plasmid was designated as pCGP487.

WO 2004/020637 PCT/AU2003/001111 - 67 pCGP7I7 A 2.2 kb petunia FLS promoter fragment upstream from the putative translational initiation site was released from the plasmid pCGP 4 87 upon digestion with the restriction endonucleases AoI and PstL The fragment generated was purified and ligated with 5 Xhol/PstI ends of pBluescript R KS+ (Stratagene, USA). Correct insertion of the fragment was established by restriction endonuclease analysis of DNA isolated from ampicillin resistant transformants. The resulting plasmid was designated as pCGP717. pCGP716 10 A 0.95 kb petunia PLS terminator fragment downstream from the putative translational stop site was released from the plasmid pCGP487 upon digestion with the restriction endonwcleases HindIII and Sad, The fragment generated was purified and ligated with HindJII/SacI ends of pBluescript II KS+ (Stratagene, USA). Correct insertion of the fragment was established by restriction endonuclease analysis of DNA isolated from 15 ampicillin-resistant transformants. The resulting plasmid was designated as pCGP716. Construction of pCGP493 (short petFLS 5petFL3' expression cassette) A 1.8 kb fragment containing the short petunia FLS promoter fragment was amplified by PCR using the plasmid pCGP717 as template and the T3 primer (Stratagene, USA) and an 20 FLS-Nco primer (5' AAA ATC GAT ACC ATG GTC TTT TTT TCT TTG TCT ATA C 3') (SEQ ID NO:19). The PCR product was digested with the restriction endonucleases XloI and CaI and the purified fragment was ligated with XhoIIClaI ends of pCGP716. Correct insertion of the fragment was established by restriction endonuclease analysis of DNA isolated from ampicillin-resistant transformants. The resulting plasmid was 25 designated as pCGP493. Construction of P CGP497 (short pPLS 5': petH fPetFLS3' expryson cassette The petunia FPS'5H (petHf) cDNA clone was released from the plasmid pCGP6 2 7 (described above) upon digestion with the restriction endonucleases BspHI and FspL The 30 BspHI recognition sequence encompasses the putative translation initiating codon and the FspI recognition sequence commences 2 bp downstream from the stop codon. The petunia WO 2004/020637 PCT/AU2003/001111 - 68 F3'5'Hpetlfl fragment generated was purified and ligated with Clal (repaired ends)/NcoI ends of the plasmid pCGP 4 93. Correct insertion of the fragment was established by restriction endonuclease analysis of DNA isolated from ampicillin-resistant transformants. The resulting plasmid was designated as pCGP497. 5 Constra'etion ofepCP1461 (yhort petFLS 5': pyetH fuptL3 inary vetr The plasmid pCGP497 was linearised upon digestion with the restriction endonuclease SacI. The overhanging ends were repaired and a 4.35 kb fragment containing the short petFLS 5': petHf: petFLS3' gene expression cassette was released upon digestion with the 10 restriction endonuclease KpnI. The fragment generated was purified and ligated with PstI (ends repaired)/KpnI ends of the binary vector pWTT2132 (DNAP) (Figure 6). Correct insertion of the fragment in a tandem orientation with respect to the 355 5': SuRB selectable marker gene cassette of pWTT2132 was established by restriction endonuclease analysis of plasmid DNA isolated from tetracycline-resistant transformants. The resulting 15 plasmid was designated as pCGP1461 (Figure 10). Plant transformation with pCGP1461. The T-DNA contained in the binary vector plasmid pCGP1461 (Figure 10) was introduced into rose via Agrobacteriumn-mediated transformation. 20 The binary vector P CGP1616 (petRT 5r: etHfl: nos 3') The plasmid pCGP1616 (Figure 11) contains a chimeric petunia F3'5'H (petHf) gene under the control of a prornoter fragment from the P. hybrida 3RT gene (pRT 5) (Brugliera, 1994, supra) with a terminator fragment from the nopaline synthase gene (nos 25 3') of Agrobacterium (Depicker, et al., 1982, supra). The chimeric petunia F3'5'H cassette is in a tandem orientation with respect to the 358 5': SuRB gene of the binary vector, pWTT2132 (DNAP) (Figure 6).

WO 2004/020637 PCT/AU2003/001111 - 69 Intermediates in the preparation of the binary vector pCGPX616 Isolation of petunia 3RT ene P. hybrida cv. Th7 eenomic DNA library construction in EMBL3 A Petunia hybrida cv. Th7 genomic library was prepared according to Sambrook et al. 5 1989, supra using a Sau3A partial digestion of the genomic DNA. The partially digested DNA was cloned into EMBL-3 lambda vector (Stratagene, USA). Screening of the Th7 genomic library for the petunia 3RT gene was as described in Brugliera, 1994, supra. A 3 kb fragment containing the petRT 5' petHf: nos 3' cassette was released from the 10 plasmid pCGP846 (described in Brugliera, 1994, supra) upon digestion with the restriction endonucleases PstI and BamHI. The purified fragment was ligated with PstI/BamHI ends of pWTT2132 (DNAP) (Figure 6). Correct insertion of the fragment in a tandem orientation with respect to the 35S 5': SuRB selectable marker gene cassette was established by restriction endonuclease analysis of plasmid DNA isolated from 15 tetracycline-resistant transfonnants. The plasmid was designated as pCGP1616 (Figure 11). Plant transformation with pCGPJ616 The T-DNA contained in the binary vector plasmid pCGP1616 (Figure 11) was introduced 20 into rose via Agrobacterium-mcdiated transformation. The binary vector pCGPI623 (mas/3SS: petHfJI: ocs 3 ) The plasmid pCGP1623 (Figure 12) contains a chimeric petunia FY5'H (petHff) gene under the control of the expression cassette contained in pKIWI 01 (Janssen and Gardner, 25 1989, supra) consisting of a promoter fragment from the cauliflower mosaic virus 358 gene (35S 5') with an enhancing sequence from the promoter of the mannopine synthase gene (mias) of Agrobacterium and a terminator fragment from the octopine synthase gene of Agrobacterium (ocs 3'). The chimeric petunia F3',5H cassette is in a tandem orientation with respect to the 3585': SuRB gene of the binary vector, pWTT2132 (DNAP) (Figure 6). 30 WO 2004/020637 PCT/AU2003/001111 - 70 Intermediates in the preparation of the binary vector pCGP1623 The -1.6 kb fragment of the petunia F3'5' pet Ilfi cDNA clone contained in the plasmid pCGP1303 (Figure 4) was released upon digestion with the restriction endonucleases BspHI and SmaL. The petunia F3t5'HpetHf fragment was purified and ligated with a -5.9 5 kb NcoIEcoRI (repaired ends) fragment of pKlWf0l1 (Janssen and Gardner, 1989, supra) to produce the plasmid pCGP1619. A partial digest of the plasmid pCGP1619 with the restriction endonuclease Xhol released a 4.9 kb fragment containing the mas/35S: petHf: oca 3' expression cassette. The 10 fragment was purified and ligated with Sall ends of pWTT2132 (DNAP) (Figure 6). Correct insertion of the fragment in a tandem orientation with respect to the 35S ': SuRB selectable marker gene cassette was established by restriction endonuclease analysis of plasmid DNA isolated from tetracycline-resistant transformants. The plasmid was designated as pCOP1623 (Figure 12). 15 Plant transformation with pCGP1623 The T-DNA contained in the binary vector plasmid pCGP1623 (Figure 12) was introduced into rose via Agrobacterium-mediated transformation. 20 The binary vector pCGP1638 (355 5': petHfl: ycs 3') The plasmid pCGP1638 (Figure 13) contains a chimeric petunia F3'5'H (petUfi) gene under the control of a CaMV 35S promoter (35S 5') with an octopine synthase terminator (ocs 3'). A -60 bp 5' untranslated leader sequence from the petunia chlorophyll a/b binding protein gene (Cab 22 gene) (Harpster et al., 1988, supra) is included between the CaMV 25 35S promoter fragment and the petunia FS''HpetHfl cDNA clone. The chimeric petunia F3'S'H cassette is in a tandem orientation with respect to the 35S 5': SuRB selectable marker gene cassette of the binary vector, pWTT2132 (Figure 6).

WO 2004/020637 PCT/AU2003/001111 -71 Intermediates in the preparation of the binary vector pCGP1638 Construction ofpCGP1273 The plasmid pCGP1273 was constructed by subcloning a -3kb HindIII/Hpal fragment containing 35S 5'; GUS; cvs 3' gene from the binary vector pJJ3499 (Jones et al., 5 Transgenic Research, 1: 285-297, 1992) with the HindIII/Smal ends of the plasmid pBluescript KS I (+) (Stratagene, USA). Construction afpCGP1634 A -3kb HindIII/BamHI fragment containing the 35S 5': GUS: ocs 3' gene from 10 pCGP1273 was then isolated and ligated with the HindIIBamHI ends of the cloning vector pUC 1 9 (New England Biolabs) to create the plasmid pCGP1634. Construction ofpCGP1636 The GUS fragment from the plasmid pCGPI634 was removed by digesting pCGP1634 15 with the restriction endonucleases NcoI. and XbaI and purifying the -3.7kb fragment containing the 353 5' promoter fragment, the ocs 3' terminator fragment and the pUC19 vector backbone. The petunia F3'5'HpetHfl eDNA clone was released from pCGP1303 (Figure 4) upon digestion with the restriction endonucleases BspHI and XbaI. The resulting -16kb 20 fragment was purified and ligated with the -3.7kb Ncot/XbaI fragment from pCGP 1634. Correct insertion of the petunia F3'5'H petHfl fragment was established by restriction endonuclease analysis of plasmid DNA isolated from ampicillin-resistant transfonnants. The resulting plasmid containing a 3535': petHf: ocs 3'gene was designated pCGP1636. 25 Construction ofpCGP1638 The 35S 5': petHfl; ocs 3' gene from the plasmid pCGP1636 was released upon digestion of pCOP1636 with the restriction endonucleases PsI and EcoRI. The ends were repaired and the -2.6kb fragment was purified and ligated with the Sam ends of the binary vector, pWTT2132 (DNAP). Correct insertion of the 353 S': petHf: ocs 3' gene in a tandem 30 orientation with respect to the 358 5': SuRB selectable marker gene cassette was established by restriction endonuclease analysis of plasmid DNA isolated from WO 2004/020637 PCT/AU2003/001111 -72 tetracycline-resistant transfonnants. The plasmid was designated as pCGP1638 (Figure 13). Plant transformation with pCGP1638 5 The T-DNA contained in the binary vector plasmid pCGPt638 (Figure 13) was introduced into rose via Agrobacterium-mediated transfonuation. The binary vector pCGP1860 (RoseCHS 5' petHfl: nas 3') The plasmid pCGP1860 (Figure 14) contains a chimeric petunia P3'3'H (pettfi) gene 10 under the control of a promoter fragment from the chalcone synthase gene of Rosa hybrida (RoseCHS 5') with a terminator fragment from the nopaline synthase gene of Agrobacterium (nos 3'). The chimeric petunia F3'5'H cassette is in a tandem orientation with respect to the 35S S': SuRB selectable marker gene cassette of the binary vector, pWTT2132 (DNAP) (Figure 6). 15 Intermediates in the preparation of the binary vector pCGP186O Isolation of Rose CHS promoter A rose genomic DNA library was prepared from genomic DNA isolated from young leaves of Rosa hybrid cv. Kardinal. 20 The Kardinal genornic DNA library was screened with 32 P-labelled fragment of rose CHS eDNA clone contained in the plasmid pCGP634. The rose CHS cDNA clone was isolated by screening of a petal cDNA library prepared from RNA isolated from petals of Rosa hybrida ov Kardinal (Tanaka et al., 1995, supra) using a petunia CHS cDNA fragment as 25 probe (clone iI1 contained in pCGP701, described in Brugliera et al., 1994, supra). Conditions are as described in Tanaka et al, 1995 (supra). A rose genomic clone (roseCHS20) was chosen for further analysis and found to contain -6.4 kb of sequence upstream of the putative initiating methionine of the rose CHS coding 30 region.

WO 2004/020637 PCT/AU2003/001111 -73 An -6.4 kV fragment upstream from the translational initiation site was cloned into pBluescript KS (-) (Statagene) and the plasmid was designated as pCGP1 114. The plasmid pCOP1114 was digested with the restriction endonucleases HindffI and 5 EcuRV to release a 2.7-3.0kb fragment which was purified and ligated with the HindIII/SmaI ends of pUC19 (New England Biolabs). Correct insertion of the rose CHS promoter fragment was established by restriction endonuclease analysis of DNA isolated from ampicillin-resistant transformants, The resulting plasmid was designated as pCGP1116. The DNA sequence of the rose CHS promoter fragment was determined using 10 pCGP1116 as template (SEQ ID NO;5). Constructiont of pCGP1 97 (RoseCHS 5':- GUS , n os 3' in PUC18 backbon e) An ~3.0 kb fragment containing the rose chalcone synthase promoter (RoseCHS 5') was released from the plasmid pCGP 1116 upon digestion with the restriction endonueleases 15 Hindm and Asp7lS. The fragment was purified and ligated with a HindlIAsp718 fragment from pJBI (Bodeau, 1994, supra) containing the vector backbone, P-glucoronidase (GUS) and nos 3'fragments. Correct insertion of the rose CHS promoter fragment upstream of the GUS coding sequence was established by restriction endonuclease analysis of DNA isolated from ampicillin-resistant transformants. The resulting plasmid was designated as 20 pCGP197. Construction of pCGP200 (RoseCHS 5': petIf(: qos 3' in pUCI8 backbone) A 1.8 kb fragment containing the petunia F3'5'T (pet/f)) fragment was released from the plasmid pCOP1303 (described above) (Figure 4) upon digestion with the restriction 25 endonucleases BspHI and Sacl. The petunia F3 '5'H petHf fragment was purified and ligated with NcoI/SacI ends of pCGP197. Correct insertion of the petunia F3'5H petH1 fragment between the rose CHS promoter and nos 3' fragments was established by restriction endonuclease analysis of DNA isolated from ampicillin-resistant transfornants. The resulting plasmid was designated as pCGP200. 30 WO 2004/020637 PCT/AU2003/001111 - 74 Consftrction of pCGPJS60 (RoseCHS 5': petHf(: nos 3F in a hinary vector) An -4.9 kb fragment containing the RoseCHS 5': petHf: nos 3' cassette was released from the plasmid pCGP200 upon digestion with the restriction endonuclease BglII. The fragment was purified and ligated with BamHI ends of the binary vector, pWTT2132 5 (DNAP) (Figure 6). Correct insertion of the fragment in a tandem orientation with respect to the 353 5': SuRB selectable marker gene cassette of pWTT2132 was established by restriction endonuclease analysis of plasmid DNA isolated from tetracycline-resistant transformants. The resulting plasmid was designated as pCGP1860 (Figure 13). 10 Plant transformation with pCGP1860 The T-DNA contained in the binary vector plasmid pCGP1 860 (Figure 14) was introduced into rose via Agrobacterium-mediated transformation. Th e bin ary vector P CGP2123 (CaMV 3SS: pyet HM oces 3' 15 The plasmid pCGP2123 (Figure 15) contains a chimeric petunia F3'S1H (petHf2) gene under the control of a CaMV35S promoter with a terminator fragment from the octopine synthase gene of Agrobacterium (ocs 39. The chimeric petunia F3'5'H cassette is in a tandem orientation with respect to the 35S5': SuRB selectable marker gene cassette of the binary vector, pCGP1988 (Figure 16), 20 Intermediates in the preparation of the binary vector pCGP2123 Construction of pCGP1988 (a derivative ofthe binary vector, pW'T2132) The binary vector pCGP1988 (Figure 16) is based on binary vector pWTT2132 (DNAP) (Figure 6) but contains the multi-cloning site from pNEB193 (New England Biolabs). The 25 plasmid pNEB193 was frstly linearized by digestion with the restriction endonuclease EcoRl. The overhanging ends were repaired and the multi-cloning fragment was released upon digestion with the restriction endonuclease PstL The fragment was purified and ligated with SalI (ends repaired)/Pstl ends of the binary vector pWTT2132 (DNAP). Correct insertion of the multi-cloning fragment into pWTT2132 was established by 30 restriction endonuclease analysis of plasmid DNA isolated from tetracycline-resistant transformants. The resulting plasmid was designated as pCGP1988 (Figure 16).

WO 2004/020637 PCT/AU2003/001111 - 75 Construction of pCGP2000 (CaMV 35S promoter frag ment in pBluescrip) The plasmid pCGP2000 was an intermediate plasmid containing a cauliflower mosaic virus (CaMV) 35S promoter fragment in a pBluescript SK (Stratagene, USA) backbone. 5 The CaMV 358 promoter fragment from pKIWI101 (Janssen and Gardner, 1989, supra) was released upon digestion with the restriction endonucleases AXbal and PstI. The -0,35kb fragment generated was purified and ligated with XbaJ/PstI ends of the vector pBluescript SK. Correct insertion of the fragment was established by restriction endonuclease analysis of plasmid DNA isolated from ampicillin-resistant transformants. The plasmid was 10 designated as pCGP2000. Construction oft CGP21OS (CaMV 35S 5' and ocs 3' fragments in pluescript) The plasmid pCGP2105 (Figure 17) contained a CaMY 35S promoter fragment along with a terminator fragment from the octopine synthase gene of Agrobacterium (ocs 3') both 15 from pKIWI1Ol (Janssen and Gardner, 1989, supra). The ocs 3' fragment from pKIWI101 (Janssen and Gardner, 1989, supra) was isolated by firstly digesting the plasmid pKIWI101 with the restriction endonuclease EcoRI, followed by repair of the overhanging ends, and finally by digestion with the restriction 20 endonuclease AoI to release a 1.6 kh fragment. This fragment was then ligated with HincIXhoI ends of pCGP2000. Correct insertion of the fragment was established by restriction endonuclease analysis of plasmid DNA isolated from ampicillin-resistant transformants. The plasmid was designated pCGP2105 (Figure 17). 25 Construction of GP21 09 (CaMV35S: pEt: ocs 3' ene in p)lueseriIg The plasmid pCGP2109 contained the CaMV 35: petHJ2: ocs 3' expression gene cassette in a pBluescript backbone. The 1.8 kb petunia F3'5'Hpetf2 cDNA clone was released from pCGP17 5 (Holton et at, 30 1993a, supra) upon digestion with the restriction endonucleases XbaI and SspI. The overhanging ends were repaired and the purified f-agment was ligated with PstI (ends WO 2004/020637 PCT/AU2003/001111 -76 repaired)/EcoRV ends of pCGP2105 (described above) (Figure 17). Correct insertion of the fragment was established by restriction endonuclease analysis of plasmid DNA isolated from ampicillin-resistant transformants. The plasmid was designated pCGP2109, 5 ConsructionofpCGP2123 (CaMV 353: petHf2: nes 3' binary vector) The CaMV 35S: petfl2: ocs 3'cassette was released from pCCGP2109 upon digestion with the restriction endonucleases Asp7l8 and AbaI. The overhanging ends were repaired and the resultant -3.7 kb fragment containing the CaMV 353: perH2: ocs 3'gene was purified and ligated with repaired ends of Asp71 of the binary vector, pCGP1988 (Figure 16). 10 Correct insertion of the CaMV 35S: petHf2: ocs 3' gene in a tandem orientation with respect to the 358 5': SuAB selectable marker gene cassette was established by restriction endonuclease analysis of plasmid DNA isolated from tetracycline-resistant transformants. The plasmid was designated as pCGP2123 (Figure 15). 15 Plant transformation wit pCGP223 The T-DNA contained in the binary vector plasmid pCGP2123 (Figure 15) was introduced into rose via Agrobacterium-mediated transfonnation. EXAMPLE 5 20 Analysis of transgenic roses The transgenic roses produced in the experiments described in Example 4 were grown to flowering. Flowers were collected and the colors of the petals were coded using the Royal Horticultural Society Colour Charts (RHSCC). The anthocyanins were extracted and the anthocyanidins (specifically the presence of delphinidin or delphinidin-based molecules) 25 analysed by TLC and/or HPLC analysis. Total RNA was also isolated from petal tissue and Northern blot analysis was used to detect transcripts of petunia F3'3'H transgenes, endogenous rose CBS gene and SuRB transgene. The results of the transgenic analysis are summarised in Table 6.

WO 2004/020637 PCT/AU2003/001111 -77 Although over 250 transgenic Kardinal roses were produced (Table 6) none produced flowers with a change in color. TLC and/or HPLC analysis failed to detect accumulation of delphinidin or delphinidin-based molecules pigments confirming the absence of efficient F3'5'H activity. Subsequent Northern analysis on total RNA isolated from petal tissue of 5 these transgenic roses revealed either no detectable intact petunia F3'3'H (petHfl or petH2) transcripts, or in some cases (see footnotes), degraded transcripts. Hybridization of the same membranes with the selectable marker gene (SuRB) or with an endogenous rose CIS cDNA probe revealed discrete hybridizing transcripts indicating that the total RNA isolated was not degraded. The detection of the SuRB transgene transcripts confirmed that 10 the roses were transgenic. TABLE 6 Results of transgenic analysis of rose petals transformed with the T-DNA from various petunia F3'5H (petHf or petHf) gene expression cassettes. PLASMID F3 'S'H GENE EVENTS DEL RNA pCOP1452 AmCHS 5': petHfl: petD8 3' 34 0128 0/34 pCGPI453 Mac: pezHfl: mas 3' 16 0/14 pCGP1457 petD85':petHf:petDS3' 11 0/11 0/11 pCGP1461 short petFfLS 5' petHf: petFLS 3' 11 0/11 0/11 pCGP1616 petRT ':petHfl: nos 3' 4 0/4 014 pCGP1623 mas/35S: petHfl: oas 3' 27 0/20 0112 pCGP1638 CaMV 35S: petHf: ocs 3' 22 0/14 0114 pCGP1860 RoseCHS 5: petHfl: nos 3' 15 0/13 0/13 pCOP2123 CaMV 355: petHj2: ocs 3' 40 0/26 0/10 15 EVENTS = number of independent transgenic events produced DEL = number of transgenic events in which delphinidin or delphinidin based molecules was detected (by TLC or HPLC) in petals over the total number of events analyzed WO 2004/020637 PCT/AU2003/001111 -78 RNA = number of transgenic events in which intact F3'5'H (petHfl or petHf2) transcripts were detected by Northern blot analysis in total RNA isolated from rose petals over the total number of events analyzed Degraded transcripts were detected in 5 of the 34 analyzed 5 = Degraded transcripts were detected in 8 of the 13 analyzed 3 . = Degraded transcripts were detected in 8 ofthe 12 analyzed The fact that no intact petunia F3'5 'H petHf orpetf2) transcripts were ever detected in traisgenic rose petals transformed with the T-DNAs described (Table 6) suggested a 10 number of possibilities: 1. that the RNA isolated was degraded. This was not the case as the RNA had been stained by ethidium bromide and visualised under UV-light. The intact visible ribosomal RNA bands were used as an indicator of the quality of the RNA isolated. Furthermore the 15 detection of full-length transcripts of the endogenous rose CHS and SuRD transgenes confined that the RNA preparation was not degraded. 2. that there was no initiation of transcription of the chimeric P3'S'H genes evaluated. This was a possibility with some of the expression cassettes analysed, as no F3'5'H transcripts 20 were detected by Northern analysis. However all of the petunia F3'5'H expression cassettes had proven to be functional (ie. result in an intact transcript and result in the production of delphinidin-based pigments) in other plants such as carnation and petunia. a, that the petunia F3'5'H petHfl and petl-42 mRNAs were unstable in roses. This was 25 also a possibility as degraded petunia F3S''H transcripts were detected by Northern analysis in total RNA isolated from petals of some events. However the petunia petHf! and petHJ2 mRNAs had been proven to be stable in other plants such as carnation and petunia. Such instability could be due to aberrant translation leading to mRNA turnover, some feature of the sequence inherently unstable in rose cells, some other factor or factors. 30 WO 2004/020637 PCT/AU2003/001111 -79 There was a need therefore to find suitable promoter fragments that would efficiently drive expression of genes in rose petals and find suitable F3'5'H sequences that would result in intact transcripts accumulating in rose petals leading to functional F3'5'H activity and to the production of delphinidin-based pigments. EXAMPLE 6 Evaluation ofpromoters in roses Development of GUS gene expression cassettes. The evaluation of the promoter and terminator fragments was performed using the GUS 10 reporter gene. Therefore, a number of promoters were linked to the -glucuronidasc reporter gene (GUS) (Jefferson et aL, 1987, supra) and introduced into roses in an attempt to identify expression cassettes that lead to effective initiation of transcription in rose flowers. 15 A summary of the promoters and terminator fragments evaluated is given in Table 7. TABLE 7 List of chimaeric GUS gene expression cassettes evaluated in roses PLASMID GUS EXPRESSION SELECTABLE BACKBONE CASSETTE MARKER GENE VECTOR pCGP1307 peD8 5': GUS: peDS 3' mas 5: nptff : mas 3' pCGN1548 pCGP1506 long petLS 5': GUS: petFLS 3' nos 5': nptIL- nos 3' pBIN19 pCGP1626 chrysCHS 5': GUS petRT 3' 35 5':- SuRB pWTT212 pCGP1641 petRT S' GUS: petRT3' 355': SuRB pWTT2132 pCGP1861 RoseCHS 5': GUS: nos 3' 353 S': SuRB pWTT2132 pCGP1953 AmCHS 5': GUS: petD 3' 35S S: SuRB pWTT2132 pWTT2084 358 5': GUS: ocs 3' 3535' SukB pWTT2 132 WO 2004/020637 PCT/AU2003/001111 -80 The binar, vector pCGP1307 (petD8 5': GUS; netD8 3') The plasmid pCGP1307 (Figure 18) contains a chimeric GUS gene under the control of a promoter and terminator fragment from the petunia PLTP gene (petD8 3' and petD8 3', respectively). The chimeric GUS reporter gene cassette is in a tandem orientation with 5 respect to the mas 5': nprff: mas 3' selectable marker gene cassette of the binary vector pCGN1548 (McBride and Summerfelt, 1990, supra). Intermediates in the preparation of the binary vector pCGPI307 The nos 3' fragment from pCGP1101 (see Example 4) was replaced with the 0.75 kb petD8 10 3' fragment (Holton, 1992, supra) to produce the plasmid pCGP1106 containing a petD8 5': GUS: petD8 3'expression cassette. The 5.3 kb fragment containing the petD8 5': GUS: petD8 3' expression cassette was released from the plasmid pCGP1 106 upon digestion with the restriction endonucleases 15 HindIII and Pstl. The fragment was purified and ligated with HindIlIPstI ends of the binary vector, pCGN1548 (McBride and Summerfelt, 1990, supra). Correct insertion of the fragment was established by restriction endonuclease analysis of DNA isolated from gentamycin-resistant transformants. The resulting plasmid was designated as pCGP1307 (Figure 18). 20 Plant transformation with pCGP1307 The T-DNA contained in the binary vector plasmid pCGP1307 (Figure 18) was introduced into rose via Agrobacterium-mediated transformation. 25 The binary vector pCGP1506 (ong petFLS 5': GUS: petPLS 3') The plasmid pCGP1506 (Figure 19) contains a chimeric GUS gene under the control of promoter and terminator fragments from the petunia flavonol synthase gene (petPLS 5'and petFLS 3', respectively). The chimeric GUS reporter gene cassette is in a tandem orientation with respect to the nos 5': nptI: nos 3 selectable marker gene cassette of the 30 binary vector pBIN19 (Bevan, NucleicAcids Rs 12: 8711-8721, 1984).

WO 2004/020637 PCT/AU2003/001111 - 81 Intermediates in the preparation of the binary vector pCGPl506 A 4 kb long petunia FLS promoter fragment upstream from the putative translational initiation site was released from the plasmid pCGP486 (described in Example 4) upon digestion with the restriction endonucleases XhoI and PstI. The fragment generated was 5 purified and ligated wit XhoIlPstI ends of pBluescript II KS+ (Stratagene, USA). Correct insertion of the fragment was established by restriction endonuclease analysis of DNA isolated from ampicillin-resistant transformants. The resulting plasmid was designated as pCGP715. 10 Construction ofpCGP494 (iong petFLS S':petFLS3' expression cassette) A 4.0 kb fragment containing the long petunia FLS promoter fragment was amplified by PCR using the plasmid pCGP715 as template and the T3 primer (Stratagene, USA) and an FLS-Noo primer (5' AAA ATC GAT ACC ATG GTC TTT TTT TCT TTG TCT ATA C 3') (SEQ ID NO:19). The PCR product was digested with the restriction endonucleases 15 Xhol and ClaI and the purified fragment was ligated with Xho/ClaI ends of pCGP716 (described in Example 4). Correct insertion of the fragment was established by restriction endonuclease analysis of DNA isolated from ampicillin-resistant transformants. The resulting plasinid was designated as pCGP494. 20 Construction of pCGP496 (iong petFLS 5': GUS: petFLS3' expression cassette) The GUS coding sequence from the plasmid piBi (Bodeau, 1994, supra) was released upon digestion with the restriction endonucleases NeoI and SmaL The GUS fragment generated was purified and ligated with Clal (repaired ends)/NcoI ends of the plasmid pCGP494. Correct insertion of the fragment was established by restriction endonuclease 25 analysis of DNA isolated from ampicillin-resistant transfonnats. The resulting plasmid was designated as pCGP496. Construction of pCGP506 long petFLS 5': GUS: petFLS3' hinar' vecto) The plasmid pCGP496 was firstly linearised upon digestion with the restriction 30 endonuclease Xhol, The overhanging ends were partially repaired (using only dTTP and dCTP in the reparation reaction) and a 6.7 kb fragment containing the long petFLS 5'; WO 2004/020637 PCT/AU2003/001111 -82 GUS: petFLS3' gene expression cassette was released upon digestion with the restriction endonnelease SacL The fragment generated was purified and ligated with BamHI(partially repaired ends using dGTP and dATP in the reparation reaction)/SacI ends of the binary vector pBIN19. Correct insertion of the fragment in a tandem orientation with respect to 5 the nos 5': npl/I: nos 3' selectable marker gene cassette was established by restriction endonuclease analysis of plasmid DNA isolated from kanamycin-resistant transformants. The resulting plasmid was designated as pCGP1506 (Figure 19). Plant transformation with pCGP1506 10 The T-DNA contained in the binary vector plasmid pCGP1506 (Figure 19) was introduced into rose via Agrobacterium-mediated transfonuation. The binary vector pCGPI626 (chrvsCHS 5': GUS: petR T 3') The plasmid pCGP1626 (Figure 20) contains a chimeric GUS gene under the control of 15 promoter fragment from the chalcone synthase gene of chrysanthemum (chrysCHS 5) and a terminator fragment from the 3RT gene of petunia (petRT 3) (Brugliera, 1994, supra). The chimeric GUS reporter gene cassette is in a tandem orientation with respect to the 35S 5': SuRB selectable marker gene cassette of the binary vector pWTT2132 (DNAP) (Figure 6). 20 Intermediates in the preparation of the binary vector pCGPI626 Isolation of ehrysanthemum CHS Promoter A chrysanthemum genomic DNA library was prepared from genonric DNA isolated from young leaf material of the chrysanthemum cv Hero. 25 The chrysanthemum genornic DNA library was screened with 32 P-labelled fragments of a chrysanthemun CHS rDNA clone (SEQ 1D NO:28) (contained in the plasmid pCGP856) using high stringency conditions. The plasmid pCGP856 contains a 1.5 kb cDNA clone of CHS isolated from a petal cDNA library prepared from RNA isolated from the 30 chrysanthemum cv. Dark Pink Pom Pom.

WO 2004/020637 PCT/AU2003/001111 - 83 A genomic clone (CHSS) was chosen for father analysis and found to contain -3 kb of sequence upstream of the putative initiating methionine of the chrysanthemum CHS coding region. 5 A 4 kb fragment was released upon digestion of the genomic clone CHS5 with the restriction endonuclease HindL. The fragment containing the chrysanthemum CHS promoter was purified and ligated with HindIll ends of pBluescript SK (Stratagene, USA). Correct insertion of the fragment was established by restriction endonuclease analysis of DNA isolated from ampicillin-resistant transfornants. The resulting plasmid was 10 designated as pCGP1316 A 2.6 kb chrysanthemum CHS promoter fragment upstream from the putative translational initiation site was amplified by PCR using pCGP1316 as template and primers "chrysanCHSATG" (5'-GTTAAGGAAGCCATGGGTGT-3') (SEQ ID NOY3) and the 15 M13 reverse primer (Stratagene, USA). Primer "cbrysanCHSATG" incorporated an Neol restriction endonuclease recognition sequence at the putative translation initiation point for case of cloning. The PCR fragment was purified and ligated with EcoRV (dT-tailed) ends of pBluescript KS (Holton and Graham, Nuc. Acids Res. 19: 1156, 1990). Correct insertion of the fragment was established by restriction endonuclease analysis of DNA isolated from 20 ampicillin-resistant transformants. The resulting plasmid was designated as pCGP620. The nucleotide sequence of the chrysanthemum CHS promoter fragment contained in pCGP1620 is represented as SEQ ID NO:30. Construction of pCGP1622 (chrysCHS 5': GUS; nos 3' in UC backbone) 25 A -2.5 kb fragment containing the chrysanthemum CHS promoter was released from the plasmid pCGP1620 upon digestion with the restriction endonucleases Ncol and PstI. The fragment was purified and ligated with a 4.8 kb NcoIPstl fragment of p.TB1 (Bodeau, 1994, supra) containing the backbone vector with the GUS and nos 3' fragments. Correct insertion of the fragment was established by restriction endonuclease analysis of DNA 30 isolated from ampicillin-resistant transfonmants. The resulting plasmid was designated as pCGP1622.

WO 2004/020637 PCT/AU2003/001111 -84 Construction o pCGPI626 (chrysCHS 5': GUS: nos 3'in binarV vector) A -4.6 kb fragment containing the chrysCHS 5'. GUS: nos 3' cassette was released from the plasmid pCGP1622 upon digestion with the restriction endonucleases PstI and Bglf, 5 The fragment was purified and ligated with PstBamIHI ends of the binary vector pWTT2132 (DNAP) (Figure 6). Correct insertion of the cassette in a tandem orientation with respect to the 35S 5': SuRB selectable marker gene cassette was established by restriction endonuclease analysis of DNA isolated from tetracycline-resistant transformants. The resulting plasmid was designated as pCGP1626 (Figure 20). 10 Plant transformation with pCGP1626 The T-DNA contained in the binary vector plasmid pCGP1626 (Figure 20) was introduced into rose via Agrobacterium-mediated transformation. 15 The binary vector pCGP1I 641 (petRT 5': GUS: petRT 3,) The plasmid pCGP1641 (Figure 21) contains a chimeric GUS gene under the control of a petunia 3RT promoter (petRT 5'). covering 1,1kb upstream from the putative 3RT translation initiation codon with a petunia 3RT terminator (petRT 3') covering 2.5 kb downstream from the 3RT stop codon. The chimeric GUS cassette is in a tandem 20 orientation with respect to the 35S 5": SuRB selectable marker gene cassette of the binary vector, pWTT2132 (DNAP) (Figure 6). Intermediates in the preparation of the binary vector pCGP1641 Isolation of petunia 3RT gene 25 The isolation of the petunia 3RT gene corresponding to the Rt locus of P. hybrida has been described in Brugliera, 1994, supra.

WO 2004/020637 PCT/AU2003/001111 - 85 - Construction of pCGP1625 (CaMV 353: GUS: petRT 3' cassette) The intermediate plasmid pCGP1625 contains a CaMV 35S: GUS: petRT 3' cassette in a pUC backbone. The 2.5 kb fragment containing a petRT terminator sequences was released from the plasmid pCGP1610 (described in Brugliera, 1994, supra) upon digestion 5 with the restriction endonucleases BamHI and SaM. The fragment was purified and ligated with the BglII/SacI 4.9kb fragment of pJBI (Bodeau, 1994, supra) containing the vector backbone and the CaMV 35S promoter and GUS fragments. Correct insertion of the petunia 3RT terminator fragment downstream of the GUS fragment was established by restriction endonuclease analysis of DNA isolated from ampicillin-resistant transformants. 10 The resulting plasmid was designated as pCGP1 625. Construction ofpCGP)628 (petRT 5': GUS: petRT3' cassette) A 1.1 kbpetRTpromoter fragment was released from the plasmid pCGP1611 (described in Brugliera, 1994, supra) upon digestion with the restriction endonucleases NeoI and Pstl. 15 The purified fragment was ligated with NeoIPstI ends of the 7kb fragment of pCOP1 625 containing the vector backbone and the GUS and petRT 3' fragments. Correct insertion of the petR promoter fragment upstream of the GUS fragment was established by restriction endonuclease analysis of DNA isolated from ampicillin-resistant transfornimants. The resulting plasmid was designated as pCGP1628. 20 Construction of pCGPI 641 (petRT 5!: GUS: peR T 3 ' binary vector) A 5.4 kb fragment containing the petRT 5': GUS: petRT 3' cassette was released from pCGP1628 upon digestion with the restriction endonuclease PstL The fragment was purified and ligated with PstI ends of the binary vector pWTT2132 (DNAP) (Figure 6). 25 Correct insertion of the fragment in a tandem orientation with respect to the 355':S' SuRS selectable marker gene cassette was established by restriction endonuclease analysis of plasmid DNA isolated from tetracycline-resistant traosformants. The resulting plasmid was designated as pCGPI641 (igure 21).

WO 2004/020637 PCT/AU2003/001111 -86 Plant transformation with pCGPJ641 The T-DNA contained in the binary vector plasmid pCGP1641 (Figure 21) was introduced into rose vta Agrobacterium-mediated transformation. 5 The binary vector pCGP1861 (RoseCHS 5': GUS: nos3' The plasmid pCGP1861 (Figure 22) contains a chimeric GUS gene under the control of a promoter fragment from the CHS gene of R. hybrida (RoseCHS 5') with a terminator fragment from the nos gene of Agrobacterium (nos 3'). The chimeric GUS reporter gene cassette is in a tandem orientation with respect to the 35Y5': SuRB selectable marker gene 10 cassette of the binary vector, pWTT2132 (Figure 6). An -5 kb fragment containing the RoseCHS 5': GUS: nos 3' cassette was released from pCGP 197 describedd in Example 4) upon digestion with the restriction endonuclease Bgil. The fragment was purified and ligated with BamHI ends of the binary vector, pWTT2132 15 (DNAP). Correct insertion of the fragment in a tandem orientation with respect to the 35S 5': SuRB selectable marker gene cassette was established by restriction endonuclease analysis of plasnid DNA isolated from tetracycline-resistant-transformants. The resulting plasmid was designated as pCGPI 861 (Figure 22). 20 Plant transformation with pCGPI861 The T-DNA contained in the binary vector plasmid pCGP 1861 (Figure 22) was introduced into rose via Agrobacterium-mediated transformation. The hinary vector pCGP1953 (AmCHS 5': GUS: petD8 3') 25 The plasmid pCGP1953 (Figure 23) contains a chimeric GUS gene under the control of a proinoter fragment from the CHS gene of Antirrhinumn majus (AmCtIS 5') with a petunia PLTP terminator (petD8 3'. The chimeric GUS reporter gene cassette is in a tandem orientation with respect to the 35S 5': SuRB selectable marker gene cassette of the binary vector, pWTT2132 (DNAP) (Figure 6). 30 WO 2004/020637 PCT/AU2003/001111 -87 Intermediates in the preparation of the binary vector pCGP953 The plasmid pJBI (Bodeau, 1994, supra) was linearised with the restriction endonuclease NeoI. The overhanging ends were repaired and the 1.8 kb GUS fragment was released upon digestion with BamHI. The GUS fragment was purified and was ligated with the 5 kb 5 XbaI(ends repaired)/BamHI fragment of pCG?726 containing the pBluescript backbone vector and the AmCHS S' and petD8 3' fragments (described in Example 4), Correct insertion of the GUS fragment between the AmCHS 5' and petD8 3' fragments was established by restriction endonuclease analysis of plasmid DNA isolated from armpicillin resistant transformants. The plasmid was designated as pCGP1952. 10 A 3.8 kb fragment containing the AmCHS 5': GUS: petD8 3' expression cassette was released from the plasmid pCGP1952 upon digestion with the restriction endonucleases EagI and PstI. The overhanging ends were repaired and the purified fragment was ligated with the repaired ends of an Asp7l8 digested pWTT2132 binary vector (Figure 6). Correct 15 insertion of the fragment in a tandem orientation with respect to the 358 5': SuRB selectable marker gene cassette was established by restriction endonuclease analysis of plasmid DNA isolated from tetracycline-resistant transformants. The plasmid was designated as pCGP1953 (Figure 23). 20 Plant transformation with pCGP1953 The T-DNA contained in the binary vector plasmid pCGP1953 (Figure 23) was introduced into rose via Agrobacterium-mediated transformation. The binary vector oWTT2084 (35S 5': GUS: ocs 3') 25 The plasuid pWTT2084 (DNAP) (Figure 24) contains a chimeric GUS gene under the control of a CaMT 35S promoter (35S 5) with an octopine synthase teninator (Oacs 3). An -60 bp 5'untranslated leader sequence from the petunia chlorophyll a/b binding protein gene (Cab 22 gene) (Harpster et al, 1988, supra) is included between the CaMV 35S promoter fragment and the GUS clone. The chimeric GUS cassette is in a tandem 30 orientation with respect to the 35S 5': SuRB selectable marker gene cassette of the binary vector, pWTT2084.

WO 2004/020637 PCT/AU2003/001111 - 88 Plant transformation with pWTT2084 The T-DNA contained in the binary vector plasnid pWTT2084 (Figure 24) was introduced into rose via Agrobacterium-mediated transformation, 5 Transgenio analysis ofroses transformed with GUS expression cassettes Northern blot analysis was performed on total RNA isolated from petals of developmental stages 3 to 4 of transgenic Kardinal roses transformed with the T-DNA of various GUS expression cassettes. There was either no accumulating transcript or an intact full-length 10 transcript of the expected size of -1.8kb as detected by Northern blot hybridisation. The relative levels of GUS transcripts accumulating in the rose petals were recorded (se Table 8). TABLE S Summary of Northern analysis on transgenic Kardinal rose flowers (open 15 bud stage) containing GUS constructs. GUS SELECTABLE, PLASMID GUS REPORTERGENE TRANSCRIPT MARKER GENE LEVELS pCGP1307 petD8S' GUS: petD83' mas 5': nptll:mas3' pCGPJ506 petFLS 5':. GUS: petFLS 3' nos 5': nptL: nos 3' pCGP1626 chrysCHS5': GUS: petRT3' 358 S': SuRB -i-t +t + pCGPI641 petRT 5': GUS: petRT3' 3585' SuRB pCGPtS6L RoseCHS 5': GUS: nos 3' 3585': SuRB ++++ pCGP1953 AmCHS 5': GUS: petD8 3' 35S 5': SuRB pWTT2084 35S 5': GUS: ocs 3' 35S 5: SuB +++++ - = no transcripts detected +to+M++ relative levels (low to high) of fullength GUS transcript detected 20 by Northern blot analysis WO 2004/020637 PCT/AU2003/001111 - 89 Based on the above results (Table 8), the CaMV 35S (35S 5) and rose Cgs (RoseCtlS 5') promoters appear to drive relatively high levels of transcription in rose petals. The chrysanthemum CHS promoter (chrysCHS 5') appears to also lead to high transcript levels but not as high as those obtained using CaMV 35S or rose CHS promoters. Surprisingly, 5 antirrhinum (snapdragon) CHS (AmCHS 5'), petunia 3RT (petRT 5), petunia FLS (petFLS 5') and petunia PLTP -(petD8 5) promoters did not appear to function in rose petals as no GUS transcripts were detected with expression cassettes incorporating these promoters. However, these same promoters fused to petHf and/or -GUS genes had previously been proven to function well in carnation and petunia leading to relatively high full-length 10 transcript levels and for petHf genes, the production of delphinidin or delphinidin-based molecules pigments. .The result obtained with the antirrhinum CHS promoter (AmCHS 5') fUsed with the GUS gene was more surprising as promoter regions from homologous genes from two other species (rose and chrysanthemum) appeared to function relatively well in roses. The antirrhinum CHS promoter had also been successfully used in 15 conjunction with petunia F3'5'H (petl'fl) to produce the novel violet-colored carnations Florigene Moondust (see International Patent Application No. PCT/AJ96/00296). The evaluation of promoter and terminator fragments fused with the GUS gene also provided further evidence to suggest that the petunia F3'5'HpetHfl and petHf2 sequences 20 were unstable in roses as constructs containing the petunia F3'5'! sequences ligated to the CaMV 35S, -rose CHS and chrysanthemum CHS promoters (which do function in rose) did not result in intact petunia F'5'HpetHfl orpetHFJ2 transcripts in roses (see Table 6). EXAMPLE 7 25 Isolation of F3'5'H sequences from species other than petunia Since the petunia F3'5'H sequences had already been proven to function in various plants such as carnation, petunia and tobacco and ultimately resulted in the production of delphinidin-based pigments, it was reasonable to assume that these sequences would .also prove functional in roses. There was an assumption that the enzyme activity may vary 30 depending on the background of the species, indeed between cultivars of a given species, that the petunia F3'5'H was introduced into. However, there was no expectation that full- WO 2004/020637 PCT/AU2003/001111 -90 length recombinant petunia F3'5'H mRNA would not accumulate. Analysis of the petunia F3'5'H nucleotide sequences (petffi and petHJ2) did not reveal any sequences which might lead to instability and subsequent degradation (Johnson et al., In A look beyond tran&cription, ASPP, USA, Bailey-Serres and Gallie, eds, 1998), intron: exon splice 5 junctions (Brendel et al., In A look beyond transciption, ASPP, USA, Bailey-Serres and Gallie, eds, 1998), or any autocatalytic or degradation trigger sequences reported in the scientific literature to date (In A look beyond transcription, ASPP, USA, Bailey-Serres and Gallie, eds, 1998). The surprising result suggested that there were factors specific to rose that resulted in petunia F3'5'H sequences being unstable. 10 Since it was not obvious why the petunia F3'5'H sequences were unstable in roses but stable in carnation, petunia or tobacco, a number of F3'5W sequences were isolated across a range of families in an attempt to determine whether any F3'5'H sequence would be stable in rose and then identify any F3'5'H sequences that would lead to the synthesis of 15 stable F3 '5'H transcripts and F3'5'H activity and ultimately the production of delphinidin based pigments in roses leading to a change in flower color. Construction of petal cDNA libraries Petal cDNA libraries were prepared from RNA isolated from petals from bud to opened 20 flower stages from various species of plants described in Table 9. Rosa hybrida is classified in the family Rosaciae, Order Rosales, Subclass Rosidae and so species that produced delphinidin-based pigments and so contained a functional F3'5'H and belonged to the Subclass Rosidae were selected. Petunia hybrid is classified in the Family Solanaceae, Order Solanales, Subclass Asteridae and so species from the Subclass 25 Asteridac that produced delphinidin-based pigments were also selected.

WO 2004/020637 PCT/AU2003/001111 -91 TABLE 9 List of flowers from which total RNA was isolated for the preparation of petal cDNA libraries. Information obtained from National Center for Biotechnology Information (NCBI) website under Taxonomy browser (TaxBrowser) as of August 2003. FLOWER SPECIES FAMILY ORDER SUBCLASS gentian Gentiana spp. Gentianaceae Gentianales Asteridae lavender Lavandula spp. Larniaceac Lamiales Asteridae salvia Salvia spp. Lamiaceae Larmiales Asteridae sollya Sollya spp. Pittosporaoeae Apiales Asterdae kennedia Kennedia spp. Fabaceae Fabales Rosidae butterfly pea Clitoria ternatea Fabaceae Fabales Rosidae pansy Viola spp. Violaceae Malpighiales Rosidae 5' Unless otherwise described, total RNA was isolated from the petal tissue of purple/blue flowers using the method of Turpen and Griffith (BioTechniques 4: 11-15, 1986). Poly(A)+ RNA was selected from the total RNA, using oligotex-dT T h (Qiagen) or by three cycles of oligo-dT cellulose chromatography (Aviv and Leder, Proc. Nati. Acad. Sci. USA 10 69:1408, 1972). In general a %ZAPII/ Gigapack 11 Cloning kit (Stratagene, USA) (Short et al., Nucl. Acids Res. 16: 7583-7600, 1988) was used to construct directional petal cDNA libraries in XZAPn using around 5 ig of poly(A)' RNA isolated from petal as template. The total 15 number of recombinants obtained was generally in the order of 1 x 105 to 1 x 106. After transfecting XL1-Blue MRF' cells, the packaged cDNA mixtures were plated at around 50,000 pfu per 15 cm diameter plate. The plates were incubated at 37"C for 8 hours, and the phage were eluted in 100 mM NaCl, 8 mM MgSO 4 , 50 maM Tris-HCl pH 20 8.0, 0.01% (w/v) gelatin (Phage Storage Buffer (PSB)) (Sambrook at al., 1989, supra). Chloroform was added and the phages stored at 4*C as amplified libraries.

WO 2004/020637 PCT/AU2003/001111 -92 In general around 100,000 pfu of the amplified libraries were plated onto NZY plates (Sambrook et al., 1989, supra) at a density of around 10,000 pfu per 15 cm plate after transfecting XLl -Blue MRF cells, and incubated at 37"C for 8 hours. After incubation at 4"C overnight, duplicate lifts wore taken onto Colony/Plaque Screen' filters (DuPont) and 5 treated as recommended by the manufacturer. Plasmid Isolation Helper phage R408 (Stratagene, USA) was used to excise pBluescript phagemids containing cDNA inserts from amplified XZAPII or LZAP cDNA libraries using methods 10 described by the manufacturer. Screening of petal cDNA Libraries Prior to hybridization, duplicate plaque lifts were washed in prewashing solution (50 mM Tris-HCI pH7.5, 1 M NaC, 1 mM EDTA, 0.1% (w/v) sarcosine) at 65*C for 30 minutes; 15 followed by washing in 0.4 M sodium hydroxide at 65*C for 30 minutes; then washed in a solution of 0.2 M Tris-HCI pH 8.0, 0.1 x SSC, 0.1% (w/v) SDS at 65*C for 30 minutes and finally rinsed in 2 x SSC, 1.0% (w/v) SDS. The membrane lifts from the petal cDNA libraries were hybridized with "P-labelled 20 fragments of a 1.6 kb BspHIFspI fragment from pCGP602 (Figure 2) (SEQ ID NO: 1) containing the petunia F3'5'HpetHfl cDNA clone (Holton et al, 1993a, supra). Hybridization conditions included a prehybridization step in 10% v/v formamide, 1 M NaCl, 10% w/v dextran sulphate, 1% w/v SDS at 42*C for at least 1 hour. The 32 P-labelled 25 fragments (each at x 106 epm/mL) were then added to the hybridization solution and hybridization was continued at 42"C for a further 16 hours. The filters were then washed in 2 x SSC, 1% w/v SDS at 42*t for 2 x 1 hour and exposed to Kodak XAR film with an intensifying screen at -70*C for 16 hours.

WO 2004/020637 PCT/AU2003/001111 -93 Strongly hybridizing plaques were picked into'PSB (Sambrook at al., 1989, supra) and rescreened to isolate purified plaques, using the plating and hybridization conditions as described for the initial screening of the cDNA library. The plasmids contained in the aZAPII or XZAP bacteriophage vectors were rescued and sequence data was generated 5 from the 3' and 5' ends of the cDNA inserts. New F3'5'H cDNA clones were identified based on sequence similarity to the petunia F3'5'HpetHfl eDNA clone. The oDNA clones isolated were given plasmid designation numbers as described in Table 10. 10 TABLE 10 Plasmid numbers and SEQ ID NO. of F3'5'H cDNA clones isolated from various species PLASMID . FGUJRE SPECIES CLONE SEQ ID NO. NUMBER NUMBER Viola spp. BP#18 pCGP1959 25 9 Viola spp. BP#40 pCGP1961 26 11 Salviaspp. Sal#2 pCGP1995 31 13 Salviaspp. Sal#47 pCGP1999 32 15 Sollya spp. Soll#5 pCGP211O 37 17 Kennedia spp. Kenn#31 pCGP2231 40 26 Clitoria tornatea BpeaHF2 pBHF2F4 43 20 Gentiana triflora Gen#48 pG48 47 22 Lavandula nit LBG pLHF8 51 31 15 -Viola (pansy) F3'5'R constructs Isolation of F3'5'HeDNA clones from petals of Viola spp. (pansy) Total RNA and poly (A)* RNA was isolated from petals of young buds of Viola spp. cultivar black pansy as described above. A petal cDNA library was constructed using AZAPW Gigapack II Cloning kit (Stratagene, USA) and screened as described above. Two 20 full-length pansy F3'5'H cDNA clones (BP#18 (SEQ ID NO:9) in pCGP1959 (Figure 25) WO 2004/020637 PCT/AU2003/001111 -94 and BP#40 (SEQ 1D NO:11) in pCGP1961 (Figure 26)) were identified by sequence similarity to the petunia F3'5'H petHff cDNA clone (SEQ ID NO:1). The BP#18 and BP#40 shared 82% identity at the nucleotide leveL Comparison of the nucleotide sequence of pansy F3'5'H clones (BP#18 and BP#40) with that of the petunia F3'5'H revealed 5 around 60% identity to the petunia F3'5'HpetHfl clone and 62% identity to the petunia F3'5'HpetHf2 clone. The binar' vectors, PCGPI972 and pCGP1973 (AmCHS 5': BP#18 or 2P#40: petD8 3') The plasmids pCGP1972 (Figure 27) and pCGP1973 (Figure 28) contain the pansy F3'S'H 10 cDNA clone (BP#18 and BP#40, respectively) between an A. majus (snapdragon) CHS promoter fragment (AmCHS 5') and a petunia PLTP tenninator fragment (petD3 3'). The chimeric F3'5'H genes are in tandem with respect to the 35S 5': SuRB selectable marker gene cassette of the binary vector, pWTT2132 (DNAP) (Figure 6). 15 The petunia F3'5'W (petHf) cDNA clone in pCGP725 (described in Example 4) (Figure 7) was removed by initially digesting pCGP725 with the restriction endonuclease BamH. The ends were repaired and the linearised plasmid was further digested with the restriction endonuclease Xbal The -4.9kb fragment containing the vector with the AmCHS 5' and perD8 3' fragments was purified and ligated with the -1.6kb KpnI (ends repaired)/XbaI 20 fragment containing the pansy F3'5'HBP#18 or BP#40 cDNA clone from pCGP1959 or pCGP1961, respectively to produce pCGP1970 and pCGP1971, respectively. The AMCHS 5': pansy F3'S'H: petD8 3' cassette was then isolated from pCGP1970 or pCGP1971 by firstly digesting with the restriction endonuclease Nod. The ends of the linearised plasmid were repaired and then the chimeric F3'S'H! genes were released upon digestion with the 25 restriction endonuclease EcoRV. The purified fragments were then ligated with Asp718 (repaired ends) of the binary vector pWTT2132 (DNAP). Correct insertion of the fragment was established by restriction endonolease analysis of plasmid DNA isolated from tetracycline-resistant transformants. The resulting plasmids were designated pCGP1972 (Figure 27) and pCGP 1973 (Figure 28), respectively. 30 WO 2004/020637 PCT/AU2003/001111 -95 Carnation and petunia transformation with pCGP1972 and 1973 The T-DNAs contained in the binary vector plasmids pCGP1972 (Figure 27) and pOOP1973 (Figure 28) were introduced separately into Dianthus caryophyllus cultivars Kortina Chanel and Monte Lisa and Petunia hybrida cv. Skr4 x Sw63 via Agrobacterhum 5 mediated transformation. The binary vectors, pCGP1967 and pCGPI969 (CaMV 35: pansy F3'5'H: ocs 3') The binary vectors pCGP1967 (Figure 29) and pCGP1969 (Figure 30) contain chimeric CaMV 35S: pansy F3'5'H: ocs 3' genes in tandem with respect to the 3SS S': SuRB 10 selectable marker gene cassette of the binary vector, pWTT2132 (DNAP) (Figure 6). Intermediates in the preparation of the binary vectors pCGP1967 and pCGP1969 The plasmids pCGP1959 (Figure 25) and pCGP1961 (Figure 26) were firstly linearized upon digestion with the restriction endonuclease KpnI. The overhanging KpnI ends were 15 repaired and the pansy F3'5'H cDNA clones, BP#18 and BP#40, were released upon digestion with the restriction endonuclease PstI. The -1.6 kb fragments generated were ligated with an -5.9 kb EcoRI (repaired ends)/PstI fragment of pKIWI10] (Janssen and Gardner, 1989, supra). Correot insertion of each fragment was established by restriction endonuclease analysis of plasmid DNA isolated from ampicillin-resistant transfonnants. 20 The resulting plasmids were designated pCGP1965 and pCGP1966, respectively. The plasmids pCGP1965 and pCGP19 6 6 were firstly partially digested with the restriction endonuclease XhoI. The resulting fragments were further digested with the restriction endonuclease Abal. The overhanging ends were repaired atid the 3.6kb fragments 25 containing the CaMV 358: pansy F3'5'h ocs 3' chimeric genes were isolated and ligated with Asp7lS repaired ends of pWTT2132 (Figure 6). Correct insertion of each fragment was established by restriction endonuclease analysis of plasmid DNA isolated from tetracycline-resistant transformants. The resulting plasmids were designated pCGP1967 (Figure 29) and pCGP1969 (Figure 30), respectively. 30 WO 2004/020637 PCT/AU2003/001111 - 96 Rose transformation with pCGP1967 and PCGPJ969 The T-DNAs contained in the binary vector plasmids pCGP1967 (Figure 29) and pCGP 1969 (Figure 31) were introduced separately into Rosa hybrida cv. Kardinal and Soft Promise via Agrobacterium-mediated transformation. The T-DNA contained in the binary 5 vector plasmids pCGP 1969 (Figure 31) was also introduced into Rosa hybrida cv. Pamela and Medeo via Agrobacterium-mediated transformation. Salvia F3*5'H constructs Isolation of a F3'5'H cDNA clone from petals of Salvia spp. 10 Total RNA and poly (A)~ RNA was isolated from young petal buds of Salvia spp. (bought from a nursery) as described above. A petal eDNA library was constructed using XZAPIL/ Gigapack II Cloning kit (Stratagene, USA). Two full-length salvia F3'5'H cDNA clones (Sal#2 (SEQ ID NO:13) in pCGP1995 (Figure 31) and Sat#47 (SEQ ID NO:15) in pCGP1999 (Figure 32)) were identified by sequence similarity with the petunia F3'5'H 15 pedtlfl cDNA clone. The Sal#2 and Sal#47 shared 95% identity at the nucleotide level. Comparison of the nucleotide sequence of salvia F3'5'H clones (Sal#2 and Sal#47) with that of the petunia F3'5'IH revealed around 57% identity to the petunia F3 '5'H petHfJ clone (SEQ ID NO:1) and 58% identity to the petunia F-3'5'TpetHf2 clone (SEQ ID NO:3). 20 The binary vectors, pCGP2121 and pCGP2122 (AmCHS 5': Salvia F3S '5H #2 or #47: petD8 3') The plasmids pCGP2121 (Figure 33) and pCGP2122 (Figure 34) contain the salvia F3'5'H cDNA clones (Sal#2 and Sal#47, respectively) between a snapdragon CHS promoter 25 fragment (AmCHS 5) and a petunia PUP terminator fragment (petD8 3') in tandem with the 355 5': SuRB selectable marker gene cassette of the binary vector pWTT2132 (DNAP) (Figure 6).

WO 2004/020637 PCT/AU2003/001111 -97 The petunia P3' 'H (petHf) cDNA clone in pCGP725 (described in Example 4) (Figure 7) was removed by initially digesting pCGP725 with the restriction endonuclease BamHI. The ends were repaired and the linearised plasmid was further digested with the restriction endonuclease XbaI. The -4.9kb fragment containing the vector with the AmCHS 5' and 5 petD8 3' fragments was purified and ligated with the -1.6kb Xhol/BamHI (ends repaired) fragment from pCGP1995 (Figure 31) containing the salvia F3'S'H#2 orXoL/EcoRI (ends repaired) fragment from pCGP1999 (Figure 32) containing the salvia F3'5'H #47, respectively to produce pCGP2116 and pCGP2117, respectively. 10 The AmCHS 5': salvia F3'5'H: petD8 3' cassette was isolated from pCGP2116 or pCGP2117 by firstly digesting with the restriction endonuclease NotL The ends of the linearized plasmid were repaired and then the chimeric F3'S'H gone cassettes were released upon digestion with the restriction endonuclease EcoRV. The -3.6kb purified fragments were then ligated with Asp718 repaired ends of the binary vector pCGP1988 (Figure 16) 15 (described in Example 4). Corect insertion of each fragment was established by restriction endonuclease analysis of plasmid DNA isolated from tetracycline-resistant transfonnants. The resulting plasmids were designated pCGP2121 (Figure 33) and pCGP2122 (Figure 34), respectively. 20 Carnation and petunia transformation with pCGP2121 and pCGP2122 The T-DNAs contained in the binary vector plasmids pCGP2121 (Figure 33) and pCCP2122 (Figure 34) were introduced separately into Dianthus caryophyllus cultivars Kortina Chanel and Monte Lisa and Petunia hybrida cv. Skr4 x Sw63 via Agrobacterium mediated transformation. 25 The hinary vectors, pCGP2120 and PCGP2119 (CaMV 35S: salvia F3'S'H: ocs 3' The binary vectors pCGP2120 (Figure 35) and pCGP2119 (Figure 36) contain chimeric CaMV 3S: salvia F3'5'H: ocs 3' gene cassettes in tandem with the 35S 5': SuRB selectable marker gene cassette of the binary vector pCGP1988 (Figure 16). 30 WO 2004/020637 PCT/AU2003/001111 -98 Intermediates in the preparation of te binary vectors pCGP2120 and pCGP2119 The plasmids pCGP1995 (Figure 31) and pCGP1999 (Figure 32) were firstly linearized upon digestion with the restriction endonuclease Xhol. The overhanging Aol ends were repaired and then the salvia F3'5'I cDNA clones Sal#2 or Sa1#47 were released upon 5 digestion with the restriction endonuclease EcoRI. In the case of pCGP1995 a partial digest with EcoRI was undertaken. The ~1L7 kb fragments were ligated with the Clal (repaired ends)/EcoRI ends of pCGP2105 (Figure 17). Correct insertion of each fragment was established by restriction endonuclease analysis of plasmid DNA isolated from ampicillin-resistant transfornants. The resulting plasmids were designated pCGP2112 and 10 pCGP21 11, respectively. The plasmids pCGP2112 and pCGP2111 were digested with the restriction endonucleases Xhol and XbaL. The resulting overhanging ends were repaired and -3.6 kb fragments containing the CaMV 35S: salvia F3'5'H: ocs 3' chimeric genes were isolated and ligated 15 with Asp718 repaired ends of the binary vector, pCGP1988 (described in Example 4). Correct insertion of each fragment was established by restriction endonuclease analysis of plasmid DNA isolated from tetracycline-resistant transformants. The resulting plasmids were designated pCGP2120 (Figure 35) and pCGP2119 (Figure 36), respectively. 20 Rose transformation with pCGP2120 and pCGP2Il9 The T-DNAs contained in the binary vector plasmids pCGP2120 (Figure 35) and pCGP2119 (Figure 36) were introduced separately into Rosa hybrid cv. Kardinal via Agrobacterium-mediatcd transformation. 25 Sollya F3'5'H constructs Isolation of a F3'5rH cDNA clone from Petals of Sollya sp. Total RNA and poly (A) t RNA was isolated from young petal buds of Sollya spp. (bought from a nursery) as described above. A petal cDNA library was constructed using XZAPI Gigapack II Cloning kit (Stratagene, USA). One full-length Sollya F3'5'H cDNA clone 30 (Sol1#5 (SEQ ID NO:17) in pCGP21 10 (Figure 37)) was identified by sequence similarity to the petunia F3'5'HpetHfl cDNA clone. Comparison of the nucleotide sequence of the WO 2004/020637 PCT/AU2003/001111 - 99 sollya A3''H clone with that of the petunia F3'5H revealed around 48% identity to the petunia F3'5'H petHfl clone (SEQ ID NO:l) and 52% identity to the petunia F3'5'H petlf2 clone (SEQ ID 140:3). 5 The binary vector pCGP2130 (Am CHS 5': SolLa F3'5'H: petD8 3') The plasmid pCGP2130 (Figure 38) contains the sollya F3'5'H Soll#5 CDNA clone between a snapdragon CHS promoter fragment (AmCHS 5') and a petunia PLTP terminator fragment (petD8 3') in a tandem orientation with respect to the 35S 5': SuRB selectable marker gene cassette of the binary vector pCGP1988 (Figure 16). 10 The petunia F3'5'H(petHf) cDNA clone in pCGP725 (described in Example 4) (Figure 7) was removed by initially digesting pCGP725 with the restriction endonucleases Xbal and BamH. The ends were repaired the -4.9kb fragment containing the vector with the AmCHS 5' andpetD8 3' fragments was purified and ligated with the repaired ends of the 15 -1.6kb Asp71 8/Psl fragment from pCGP21 10 containing the sollya F3 '1'H ODNA clone to produce pCGP2128. Correct insertion of the sollya F3'5'H fragment in tandem with the AmCHS 5' andpetD8 3' fragments was confirmed by restriction endonuclease mapping. The AmCHS 5': sollya F3'5'H: petD8 3' gene cassette was then isolated from pCGP2128 20 by firstly digesting with the restriction endonuclease Nod. The ends of the linearized plasmid were repaired and then the chimeric F3'5'H gene was released upon digestion with the restriction endonuclease EcoRV. The -3.5kb purified fragment was then ligated with Asp718 (repaired ends) of the binary vector pCGP1988 (described in Example 4) (Figure 16). Correct insertion of the fragment was established by restriction endonuclease analysis 25 of plasmid DNA isolated from tetracycline-resistant transformants. The resulting plasmid was designated pCGP2130 (Figure 38). Carnation and petunia transformation with PCGP2130 The T-DNA contained in the binary vector plasmid pCGP2130 (Figure 38) was introduced 30 into Dianthus caryoplhyllus cultivars Kortina Chanel and Monte Lisa and Petunia hybrida cv. Skr4 x Sw63 via Agrobacterium-mediated transformation.

WO 2004/020637 PCT/AU2003/001111 -100 The binary vector pCOP2131 (CaMV 35S: soilva F3'5'H: ocs 3') The binary vector pCGP2131 (Figure 39) contains a chimeric CaMV 35S: sollya F3'S'.: ocs 3' gene in tandem with the 35S5': SuRB selectable marker gene cassette of the binary 5 vector pCGP1988 (Figure 16). Intermediates in the preparation of the binary vectoipCGP2131 The plasmid pCGP2110 was firstly linearized upon digestion with the- restriction endonuclease Asp718. The overhanging ends were repaired and then the sollya F3'3'H 10 cDNA clone was released upon digestion with the restriction endonuclease PstL The -1.7 kb fragment was ligated with the EcoRV/Pstl ends of pCGP2105 (Figure 17). Correct insertion of the fragment was established by restriction endonuclease analysis of plasmid DNA isolated from anpicillin-resistant transformants. The resulting plasmid was designated pCGP2129. 15 A 3.6 kb fragment containing the CaMV 35S: sollya F3'5'H: ocs 3' chimeric gene was released upon digestion with the restriction endonuoleases Asp718 and Xbal The overhanging ends were repaired and the purified fragment was ligated with of Asp7l8 repaired ends of the binary vector, pCGP1988 (Figure 16). Correct insertion of the 20 fragment was established by restriction endonuclease analysis of plasmid DNA isolated from tetracycline-resistant transformants. The resulting plasmid was designated pCGP2131 (Figure 39). Rose transformation with pCGP213I 25 The T-DNA contained in the binary vector plasmid pCCP2131 (Figure 39) was introduced into Rosa hybrida cv. Kardinal via Agrobacterium-mediated transformation.

WO 2004/020637 PCT/AU2003/001111 - 101 Kennedia F3'5'H constructs Isolation of a F3' 5'H cDNA clone from petals of Kennedia SP. Total RNA and poly (A) 4 RNA was isolated from young petal buds of Kennedia spp. (bought from a nursery) as described above. A petal cDNA library was constructed using 5 AZAPII/ Gigapack II Cloning kit (Stratagene, USA). One full-length kennodia F3'3'!H cDNA clone (Kenn#31 in pCGP2231 (Figure 40)) (SEQ ID NO:26) was identified by sequence similarity to the petunia P3'5'H petHf cDNA clone. Comparison of the nucleotide sequence of the kennedia F3'S'H clone with that of the petunia F3'5'H revealed around 64% identity to the petunia F3'5'HpetHf clone (SEQ ID NO:1) and 60% identity 10 to the petunia F3'5'SpetHf2 clone (SEQ ID NO:3) The iMary vector PCGP2256 (Am CHS 5': kennedia F3'5'I: petO8 3 The plasmid pCGP2256 (Figure 41) contains the kennedia F3'S'H (Kenn#31) cDNA clone between a snapdragon CHS promoter fragment (Ant CHS 5') and a petunia PLTP terminator 15 fragment (peD8 3') in tandem with the 35S 5': SuRB selectable marker gene cassette of the binary vector pCOP1988 (Figure 16). The petunia F3'5'H (petHf) cDNA clone in pCGP725 (described in Example 4) (Figure 7) was removed by initially digesting pCGP725 with the restriction endonuclcases XbaI and 20 BamHL. The ends were repaired the -4.9kb fragment containing the vector with the AmCHS 5' and petD8 3' fragments was purified and ligated with the repaired ends of the -1.Skb AlhoIBamHI fragment from pCGP2231 containing the kennedia. F3'S'H cDNA clone to produce pCGP2242. Correct insertion of the kennedia F3'S'H fragment in tandem with the AmCHS 5' and peD8 3' fragments was confuimed by restriction endonuclease 25 mapping. The AmCHS 5' kennedia F35'H: petD8 3' cassette was then isolated from pCGP2242 by digesting with the restriction endonucleases Noil and EcoRLt The ends were repaired and the -3.7kb purified fragment was then ligated with Asp71 repaired ends of the binary 30 vector pCGPl988 (described in Example 4) (Figure 16). Correct insertion of the fragment was established by restriction endonuclease analysis of plasmid DNA isolated from WO 2004/020637 PCT/AU2003/001111 -102 .tetracycline-resistant transfonnants. The resulting plasmid was designated pCOP2256 (Figure 41). Petunia trans formation with pCGP2256 5 The T-DNA contained in the binary vector plasmid pCGP2256 (Figure 41) was introduced into Petunia hybrid cv. Skr4 x Sw63 via Agrobacterium-mediated transformation. The binary vector pCGP2252 (CaMV 35S: kennedia F3'5'H: ocs 3') The binary vector pCGP2252 (Figure 42) contains a chimeric CaMV 355: kennedia 10 F3'5'W: ocs 3' gene in tandem with the 35S t SuRB selectable marker cassette of the binary vector pCGP1988 (Figure 16). Intermediates in the preparation of the binary vector pCGP2252 The plasmid pCGP 2 231 was firstly linearized upon digestion with the restriction 15 endonuclease XhoL The overhanging ends were repaired and then the kennedia F3'5'H cDNA clone was released upon digestion with the restriction endonuclease Psi, The -1.7 kb fragment was ligated with the Clal (repaired ends)/PsetI ends of pCGP2105 (Figure 17). Correct insertion of the fragment was established by restriction endonuclease analysis of plasmid DNA isolated from ampicillin-resistant transformants. The resulting plasnid was 20 designated pCGP2236, A 3.6 kb fragment containing the CoMV 35S: kennedia F3'5'H vos 3' chimeric gene cassette was released from the plasmid pCGP2236 upon digestion with the restriction endonucleases AoI and NoL The overhanging ends were repaired and the purified 25 fragment was ligated with Asp718 repaired ends of the binary vector, pCGP1988 (Figure 16). Correct insertion of the fragment was established by restriction endonuclease analysis of plasmid DNA isolated from tetracycline-resistant transformnants. The resulting plasmid was designated pCGP2252 (Figure 42). 30 Ros transformation with pCGP2252 WO 2004/020637 PCT/AU2003/001111 - 103 The T-DNA contained in the binary vector plasmid pCGP2252 (Figure 42) was introduced into Rosa hybrida cv. Kardinal via Agrobacterium-mediated transformation. Butterfly pen F3'5'H constructs 5 Isolation of a F3'5'H cDNA clne m metals of Clitoria ternatea ber en Construction of butterfly pea petal eDNA library A blue variety of Clitoria ternatea (butterfly pea, the seeds were kindly provided by Osaka Botanical Garden) was grown in a field in Osaka. Total RNA was isolated from fresh and pigmented petals at a pre-anthesis stage as described above. PolyA t RNA was prepared 10 using Oligotex (Takara) according to the manufacturer's recommendation. A petal cDNA library of butterfly pea was constructed from the polyA RNA using a directional XZAP cDNA synthesis kit (Stratagene, USA) following the manufacturer's protocols. Screeninh of butterfly Pea cDNA library for a F3S'H DNA clone 15 The butterfly pea petal cDNA library was screened with DIG-labelled petunia F3'5r petHfl cDNA clone as described previously (Tanaka et al., Plant Cell Physiol 37: 711 716, 1996). Two cDNA clones that showed high sequence similarity to the petunia F3,'5'H petHf were identified. The plasmid containing the longest cDNA clone was designated pBHF2 and the cDNA clone was sequenced. Alignment between the deduced amino acid 20 sequences of the butterfly pea F3'5'H clone and the petunia F3'J'HpetHfl clone (SEQ ID NO:2) revealed that the butterfly pea F3'5'H cDNA (contained in pBHF2) did not represent a full-length eDNA and lacked first 2 bases of the putative initiation codon. These two bases along with a BamHI restriction endonuclease recognition site were added to the cDNA clone using PCR and a synthetic primer, 5' 25 GGGATCCAACAATGTTCCTTCTAAGAGAAAT-3' [SEQ ID NO:25] as described previously (Yonekura-Sakakibara et al., Plant Cell Physiol. 41: 495-502, 2000). The resultant fragment was digested with the restriction endonucleases BamiHI and PstI and the subsequent DNA fragment of about 200 bp was recovered. The DNA fragment was ligated with a 3.3 kb fragment of BamHIEcoRI digested pBHF2 to yield pBH{F2F (Figure 43). 30 The DNA sequence was confirmed to exclude errors made during PCR (SEQ ID NO:20).

WO 2004/020637 PCT/AU2003/001111 - 104 Comparison of the nucleotide sequence of butterfly pea F3'5'H clone (SEQ ID NO:20) with that of the petuniaF3'5'H revealed around 59% identity to the petunia M3'5'HpetHfJ clone (SEQ ID NO:1) and 62% identity to the petunia F3'5'H petH2 clone (SEQ ED NO:3). 5 The binary vector pCGP2135 (AmCHS 5': butterfly pea F3'5'H: peiD8 3') The plasmid pCGP2135 (Figure 44) contains the butterfly pea F3'5H cDNA clone between a snapdragon CHS promoter fragment (AmCHS 5) and a petunia PLTP terminator fragment (peD8 39 in tandem with the 3S 5': SuRB selectable marker gene cassette of 10 the binary vector pCGPl 988 (Figure 16). The petunia F3'5'H (petIfl) cDNA clone in pCGP725 (described in Example 4) (Figure 7) was removed by initially digesting pCGP725 with the restriction endonucleases XbaI and BamHI. The ends were repaired the -4,9kb fragment containing the vector with the 15 AmCHS 5' and petDB 3' fragments was purified and ligated with the repaired ends of the -1.6kb XhoI/BamHI fragment frorn pHF2F (Figure 43) containing the butterfly pea F3'5'H cDNA clone to produce pCGP2133. Correct insertion of the butterfly pea F3'5'H fragment in tandern with the AmCHS 5' and petD8 3' fragments. was confirmed by restriction endonuolease mapping. 20 The AmCHS 5': butterfly pea F3'3'H: petD8 3' cassette was then isolated from pCGP2133 by firstly digesting with the restriction endonuclease NotI. The ends of the linearised plasmid were repaired and then the chimeric F' 'H gene was released upon digestion with the restriction endonuclease EcoRV. The -3.6kb purified fragment was then ligated with 25 Asp718 repaired ends of the binary vector pCGP1988 (described in Example 4) (Figure 16). Correct insertion of the fragment was established by restriction endonuolease analysis of plasmid DNA isolated from tetracycline-resistant transformants, The resulting plasmid was designated pCGP2135 (Figure 44).

WO 2004/020637 PCT/AU2003/001111 -105 Carnation and Petunia transformation with pCGP2J3S The T-DNA contained in the binary vector plasmid pCGP2135 (Figure 44) was introduced into Dianthus caryopihyllus cultivars Kortina Chanel and Monte Lisa and Petunia hybrida cv. Skr4 x Sw63 via Agrobacterium-mediated transformation. 5 The binary vector pREBFS (eCaMV35: Butterfly pea F3'5'H: nos 3) The binary vector, pBE2113-GUS contains a GUS coding region between an enhanced CaMV35$ promoter and ns terminator in a pBI121 binary vector (Mitsuhara et al, 1996, supra), The plasmid pBE2113-GUS was digested with the restriction endonuclease SacI. 10 The overhanging ends were repaired and then ligated with a Sall linker to yield pBE2113 GUSs. The 1.8 kb BamHI-XhoI fragment from pBHF2F was ligated with BamI-SalI digested pBE2113-GUSs to create pBBBFS (Figure 45). Rose transformation with pBEJFS 15 The T-DNA contained in the binary vector plasmid pBEBF5 (Figure 45) was introduced into Rosa hybrida cultivar Lavande via Agrobacterium-mediated transformation. The binary vector pCGP2134 (CaM J35S: butterfly pea F3 15'H: ocs 3 9 The binary vector pCGP2134 (Figure 46) contains a chimeric CaIV 35S: butterfly pea 20 F3'3'H: ocs 3' gene cassette in a tandem orientation with the 355 S': SuRB selectable marker gene cassette of the binary vector pCGP1988 (Figure 16). Intermediates in the preparation of the binary vector pCGP2134 The butterfly pea P3'5'W cDNA clone was released upon digestion of the plasmid pBHF2F 25 (Figure 43) with the restriction endonucleases Ahol and BamHL, The overhanging ends were repaired and the ~L7 kb fragment was lighted with the PstI (repaired ends)/EcoRV ends of pCGP2105 (described in Example 4) (Figure 17). Correct insertion of the fragment was established by restriction endonuclease analysis of plasmid DNA isolated from ampicillin-resistant transformants. The resulting plasmid was designated pCGP2132. 30 WO 2004/020637 PCT/AU2003/001111 - 106 An -3.6 kb fragment containing the CaMV 355: butterfly pea F3'5'h: oe 3' chimeric gene cassette was released upon digestion with the restriction endonucleascs Xhol and Xbal. The overhanging ends were repaired and the purified fragment was ligated with Asp718 repaired ends of the binary vector, pCGP19S8 (described in Example 4) (Figure 16). 5 Correct insertion of the fragment was established by restriction endonuclease analysis of plasmid DNA isolated from tetracycline-resistant transformants. The resulting plasmid was designated pCGP2134 (Figure 46). Rose transformation with pCGP2134 10 The T-DNA contained in the binary vector plasmid pCGP2134 (Figure 46) was introduced into Rosa hybrida ov. Kardinal via Agrobacterium-mediated transformation. Gentia F3'5'H constructs Isolation of a F3'S'HcDNA clone from petals of Gentiana triflora (gentian). 15 Construction and screening of a gentian petal cDNA library The isolation of a gentian cDNA encoding F3'3'H has been described previously (Tanaka et al., 1996, supra) and is contained within the plasmid pG48 (Figure 47). Comparison of the nucleotide sequence of the gentia F3'5'H clone (Gen#48) (SEQ ID NO:22) contained in the plasmid pG48 (Figure 47) with that of the petunia P3'5'H revealed around 61% identity 20 to the petunia F3'5'H petHfl clone (SEQ ID NO:I) and 64% identity to the petunia F3'S'HpeIH2 clone (SEQ ID NO:3). The binary vector pCGPI498 (AmCHS 5': entia F3'S'H: petD8 3 ) The plasmid pCGP1498 (Figure 48) contains the gentia F3'5'H (Gen#48) eDNA clone 25 between a snapdragon C/S promoter fragment (AmCHS 5' and a petuniaPLTP terminator fragment (petD8 3') in tandem with the 355 5': SuRB selectable marker gene cassette of the binary vector pWTT2132 (Figure 6). The petunia F3'5'H(petHfl) cDNA clone in pCGP725 (described in Example 4) (Figure 7) 30 was removed by initially digesting pCGP725 with the restriction endonucleases Xbal and BamHL The ends were repaired the -4.9kb fragment containing the vector with the WO 2004/020637 PCT/AU2003/001111 -107 AmCHS 5' and petD8 3' fragments was purified and ligated with the repaired ends of the -1.7 kb XhoIIBamHI fragment from pG48 (Figure 47) containing the gentia F3'5'H cDNA clone to produce pCGP1496. Correct insertion of the gentia P3'5'H fragment in tandem with the AmCHS 5' and petD8 3' fragments was confirmed by restriction endonuclease 5 mapping. The AmCHS 5': gentia F3'5'H: petD8 3' cassette was then isolated from pCGP1496 by firstly digesting with the restdction endonuclease NotI. The overhanging ends of the linearised plasmid were repaired and then the chimeric F3'5'H gene was released upon 10 digestion with the restriction endonuclease EcoRV. The -3.6kb purified fragment was then ligated with Asp718 repaired ends of the binary vector pWTT2132 (Figure 6). Correct insertion of the fragment was established by restriction endonuclease analysis of plasmid DNA isolated from tetracycline-resistant transfonnants. The resulting plasmid was designated pCGP1498 (Figure 48). 15 Carnation and ehunia transformation with pCGPI498 The T-DNA contained in the binary vector plasmid pCGP1498 (Figure 48) was introduced into Dianthus caryophyllus cultivars Kortina Chanel and Monte Lisa and Petunia hybrida cv. Skr4 x Sw63 via Agrobacterium-mediated transformation. 20 The binary vector pnEGHF48 (eCaMV 35S: entia F3 'SH: nos 3'9 The gentia F3'5'H cDNA clone was released by digestion of the plasmid pG48 with the restriction endonucleases BamHI and XhoL The resulting -1.7 kb DNA fragment was isolated and ligated with Bamll/SalI digested pBE2113-GUSs (Mitsuhara et al., 1996, 25 supra) to create pBEGHF48 (Figure 49), Rose transformation with pBE GHF48 The T-DNA contained in the binary vector plasmid pBEGHF48 (Figure 49) was introduced into Rosa hybrida cv. Lavande via Agrobacterium-mediated transformation. 30 WO 2004/020637 PCT/AU2003/001111 - 108 The binary vector pCGPI982 (CaMV 35S: Rentia F3'S'H: ocs 3') The binary vector pCGP1 9 8 2 (Figure 50) contains a chimeric CaMV 35S: gentia F3'5'H: ocs 3 'gene cassette in tandem with the 353 5': SuRB selectable marker gene cassette of the binary vector pWTT2132 (Figure 6). 5 Intermediates in the preparation of the binary vector pCGP1 982 The plasmid pG 4 8 (Figure 47) was linearised upon digestion with the restriction endonuclease Asp718. The overhanging ends were repaired and then the gentia F3'5'H cDNA clone (Gen#48) was released upon digestion with the restriction endonuclease 10 BamHl. The -1.7 kb fragment was ligated with the 5.95kb EcoRI (repaired ends)/BamHI fragment of pKIWI101 (Janssen and Gardner, 1989, supra). Correct insertion of the fragment was established by restriction endonuclease analysis of plasmid DNA isolated from ampicillin-resistant transformants. The resulting plasmid was designated pCGPl981. 15 An -3.6 kb fragment containing the CaMV 35S: gentia F3'5'H: ocs 3' chinmeric gene cassette was released upon digestion of the plasmid pCGP1981 with the restriction endonucleases Xh6I and Xbal The overhanging ends were repaired and the purified fragment was ligated with repaired ends of Asp7l 8 digested binary vector, pWTT2132. Correct insertion of the fragment was established by restriction endonuclease analysis of 20 plasmid DNA isolated from tetracycline-resistant transformants. The resulting plasmid was designated pCGP1982 (Figure 50). Rose transformation with nCGPJ982 The T-DNA contained in the binary vector plasmid pCGP1982 (Figure 50) was introduced 25 into Rosa hybrida av. Kardinal via Agrobacterium-mediated transformation.

WO 2004/020637 PCT/AU2003/001111 -109 Lavender F3'5'H constructs Isolation of a P3S '5H cDNA clone rm petals of Lavandula nil lavender ) Construction of lavender petal cDNA library Cut flowers of a violet variety of Lavandula nil were purchased from a florist. Total RNA 5 was isolated from fresh and pigmented petals as described above. PolyA+ RNA was prepared using Oligotex (Takara) according to the manufacturer's recommendations. A petal cDNA library of lavender was constructed from the polyA* RNA using a directional XZAP-cDNA synthesis kit (Stratagene, USA) following the manufacturer's protocols. 10 Screening of lavender eDNA library for a F3'5'H eDNA clone The lavender petal cDNA library was screened with DIG labelled petunia F3'5'HpetHfl cDNA clone as described previously (Tanaka et al. 1996, supra). One cDNA clone (LBG) that showed high similarity to petunia F3'5'HpetHfl was identified and the plasmid was designated pLHFS (Figure 51). The nucleotide sequence of the lavender F3'5'H (LEG) 15 cDNA clone was designated as SEQ ID NO: 31. Comparison of the nucleotide sequence of lavender JY3'5'H clone with that of the petunia F3'5'H eDNA clones revealed around 59% identity to the petunia F3'5'H petHf clone (SEQ TD NO: 1) and 60% identity to the petuniapetHf2 clone (SEQ ID NO:3). 20 The binary vector pBELFS (eCaMV35S: lavender F3'5'H:- nos 3') The plasmid of pLHFS (Figure 51) was digested with the restriction endonucleases BamHI and AoI to release a DNA fragment of approximately 1.8 kb. The -1.8kb purified fragment from pLHF8 was then ligated with the BamHI-SalI digested ends of the plasmid 25 pBE2113-GUSs (described above) to create pBELFS (Figure 52). Rose Iransformation with PBELF8 The T-DNA contained in the binary vector plasmid pBELFS (Figure 52) was introduced into Rosa hybrida cultivar Lavande via Agobacterium-mediated transformation. 30 WO 2004/020637 PCT/AU2003/001111 - 110 EXAMPLE 8 Analysis oflransgenic carnation, petunia and rose The transgenic plants produced in the experiments described in Example 7 were grown to flowering. Flowers were collected and the colors of the petals were coded using the Royal 5 Horticultural Society Colour Charts (RHSCC). The anthocyanins were extracted and the anthocyanidins analysed by spectrophotometric, TLC and/or HPLC analysis. Total RNA was also isolated from petal tissue of the appropriate stages of flower development and Northern blot analysis was used to detect transcripts of P3'5'H transgenes. The results of the transgenic analysis are summarised in Tables 11, 12 and 13. 10 Carnation The F3''H genes described in Example 7 were evaluated for their ability to lead to the production of delphinidin-based pigments in carnation petals. Two carnation cultivars, Kortina Chanel (K) and Monte Lisa (ML), were used in the transformation experiments. 15 The carnation cultivar Kortina Chanel produces pink colored flowers that normally accumulate cyanidin-based anthocyanins. This cultivar therefore contains a carnation F3'H and DFR activity that an introduced F3'5'H would need to compete with for substrate. The carnation cultivar Monte Lisa produces brick red colored flowers that normally accumulate pelargonidin-based anthocyanins. This cultivar is thought to lack fully functional F3'H 20 activity and contain a DFR that is capable of acting on DHK and thus an introduced F3'5'H would only be required to compete with the endogenous DFR for substrate.

WO 2004/020637 PCT/AU2003/001111 - 111 TABLE 11 Results of transgenic analysis of petals from carnations trasformed with T DNAs containing F3'5'H gene expression cassettes (AmCHS 5' F3'5'H: petD8 3'). Highest Av. F3' 5'H pCGP 6v, #tg TLC+ HPLC+ Northern+ % del % del KC 22 2/16 3/4 12.5% 7% nd Salvia#2 2121 ML 21 17/18 9/9 76% 57% 14115 Salvia#47 2122 KC 23 6/12 8/8 29% 12% nd ML 25 21/22 17/17 88% 56% 12/14 KC 30 22/27 17/17 35% 11% nd Sollya 2130 - --- ML 23 14/15 14/14 76% 49% 13/14 KC 22 0/16 0/1 nd nd nd Butterfly pea 2135 ML 24 19/20 13/13 23% 10% 14/14 KC 22 0/14 nd nd nd 7/8 Gentian 1498 ~ i c / ML 2 2/2 1/1 nd nd 1/2 KC 26 18/20 12/12 14% 9% 19/19 pansy BP#18 1972 _ _ ML 21 15/16 8/8 80% 66% - 14/16 KC 26 11/15 7/8 18% 8% 13/17 pansy BP#40 1973 ML 33 19/22 20/20 72% 52% 12/15 petunia 1452 KC 104 41164 nd 3.5% 1.3% 15/17 petHfl ML 48 39/41 26/26 75% 30% 12/13 1524 ML 27 18/19 17/17 81% 41% 12/14 petH2 5 F3'5'H F3'5'H sequence contained on the T-DNA pCGP =plasmid pCGP number of the binary vector used in the transformation experiment cv. cultivar KC carnation cultivar Kortina Chanel (cyanidin line) 10 ML carnation cultivar Monte Lisa (pelargonidin line) #tg total number of transgenics produced WO 2004/020637 PCT/AU2003/001111 -112 TLC+ = number of individual events in which delphinidin or delphinidin based molecules was detected in petals (as determined by TLC) over the total number of. individual events analyzed HPLC+ = number of individual events in which delphinidin or delphinidin 5 based molecules was detected in petals (as determined by HPLC) over the total number of individual events analyzed Highest % del = Highest % delphinidin or delphinidin-based molecules detected in the petals for the population of transgenic events Av % del t average % delphinidin or delphinidin-based molecules detected in 10 the petals for the population of transgenic events Northern - number of individual events in which the specific intact 3'5H transcripts were detected by Northern blot analysis in total RNA isolated from petals over the total number of events analyzed nd not done 15 The results suggest that all of the F3'3'H sequences evaluated (petunia petHfl, petunia petH2, Salvia SaW2, Salvia SaI#47, Sollya Sol#5, Butterfly pea BpeaHP2, pansy BP#18, pansy BP#40 and Gentian Gen#48) were stable in carnation and resulted in the production of novel delphinidin-based pigments in carnation flowers. Intact transcripts of each F3 '5'H 20 were detected by Northern blot analysis in total RNA isolated from petals of the transgenic carnations. Petunia The F3'S'H genes described in Example 7 were evaluated for their ability to lead to the 25 production of delphinidin-based pigments in petunia petals. The P. hybrida F1 hybrid Skr4 x SW63 which is homozygous recessive for Hfi and Hf2, was used in the transformation experiments. Although Skr4 x SW63 is homozygous recessive for Hf and Hf2, these mutations do not completely block production of the endogenous F3'5'H (see US Patent Number 5,349,125) and low levels of malvidin are produced to give the petal limb a pale 30 lilac color. Malvidin is the methylated derivative of the 3'5T-hydroxylated pigment, delphinidin or delphinidin-based molecules (Figures 1A and 1B). Spectrophotometric WO 2004/020637 PCT/AU2003/001111 -1123 analysis was used as a measure of total anthocyanins accumulating in petals from the transgenic petumia flowers. The increased level of anthocyanins and/or the color change detected was used as a guide to the efficacy of the F3'5'H gene under evaluation, 5 TABLE 12 Results of transgenic analysis of petals from P. hybrida cv Skr4 x SW63 plants transformed with T-DNAs containing F3'S'f gene expression cassettes (AmCHS 5' F3'5'H: petD8 3'. F35'IH pCGP # tg TLC+ Col tA/c Best Av. Northern+Bs color 144 control na na na na na 0 75C 250 Gentian#48 1498 22 3/5 18/20 nd 6/6 72B/78A Butterfly pea 2135 24 18/20 22/24 23/24 4427 2397 nd 74A/78A Kennedia 2256 24 22/24 22/24 22/24 4212 2592 nd 74A/78A Salvia#2 2121 24 21/24 21/24 21/24 2471 1730 nd 78A Salvia#47 2122 19 17/19 16/19 16/19 2634 1755 nd 78A/80A Sollya#5 2130 22 14/16 13/16 13/16 3446 1565 nd 78A pansy BP#1 8 1972 22 nd 20/22 nd nd nd 9/9 74A/B pansy BP#40 1973 19 8/8 18/19 18/20 2583 1556 nd 74/78A petunia petHf/ 484 16 nd 9116 8/IS 2683 1250 nd 74A/B petunia 1524. 20 nd 18/20 8/8 4578 2357 8/8 74A/B 10 F3'S'H. F3'5'H sequence contained on the T-DNA pCGP plasmid pCGP number of the binary vector used in the transformation experiment #tg = total number of transgenics produced WO 2004/020637 PCT/AU2003/001111 -114 TLC+ = number of individual events in which malvidin was detected in the flowers (at a level above the Skr4 x Sw63 background) (as determined by TLC) over the total number of individual events analyzed Col = number of individual events that produced flowers with an altered flower 5 color compared to the control over the total number examined t A/c = number of individual events that had an increased level of anthocyanius in petals as measured by spectrophotometric analysis of crude extracts over the number of individual events analyzed (in pmoles/g) Best highest anthocyanin amount as measured by spectrophotometric analysis 10 of crude extracts from a flower of an individual event (in ptmoles/g) Av = the average amount - of anthocyanin detected as measured by spectrophotometric analysis of crude extracts from a flower in the population of transgenic flowers analysed (in prmoles/g) Northern - number of individual events in which the specific intact F3'5H transcripts 15 were detected by Northern blot analysis in total RNA isolated from petals over the total number of events analyzed Best color = most dramatic color change recorded for the transgenic population nd = not done na - not applicable 20 Introduction of the F3'5'N gene expression cassettes into Skr4 x SW63 led to a dramatic flower color change from pale lilac to purple with a dramatic increase in the production of malvidin in the petals.. 25 The results suggest that all of the F3'3'H sequences tested (petuniapeiftfl, petunia petff2, Salvia Sal#2, Salvia Sal#47, Sollya So)#5, Butterfly pea BpeaHF2, pansy BP#18, pansy BP#40, Gentian Gen#48, Kennedia Kenn#31) were stable in petunia petals and resulted in the complementation of the Hf or Hft mutation in the Skr4 x SW63 petunia line leading to dramatically increased levels of malvidin accumulation with a concomitant color change 30 WO 2004/020637 PCT/AU2003/001111 -115 Rose The F3'5'H genes desexibed in Example 7 were evaluated for their ability to lead to the production of delphinidin-based pigments in rose petals. A selection of three rose cultivars, Kardinal (Kard), Soft Promise (SP) or Lavande (Lav) were used in transformation 5 experiments. The rose cultivar Kardinal produces red colored flowers that normally accumulate cyanidin-based anthocyanins. This cultivar therefore contains rose F3'H and DFR activities that the introduced F3'5'H would need to compete with for substrate. The rose cultivar Lavande produces light pink colored flowers that normally accumulate cyanidin-based anthocyanins. This cultivar therefore contains functional rose F3'H and 10 DFR activities that the introduced F3'5'H would need to compete with for substrate, The rose cultivar Soft Promise produces apricot colored flowers that normally accumulate pelargonidin. This cultivar is thought to lack a fully functional rose F3'H activity and contain a DFR that is capable of acting on DHK and thus the introduced F3'5'H would only be required to compete with the eidogenous rose DFR for substrate. 15 TABLE 13 Results of transgenic analysis of petals from roses transfonned with T DNAs containing F3'5'H gene expression cassettes (CaMV 353: F3'5'H: ocs 3'. F3'S'H plasmid Cult #tg TLC + HiPLC+ Highest AV. % North % del del Salvia2 pCGP2120 .Card 30 18/20 21/21 12% 5% 18/18 Salvia47 pCGP2119 i Kard 22 11/16 9/9 7.1% 2% 12/15 SOllya pCGP2131 Kard 27 0/23 2/2 1% 0.5% 6/6 Butterfly pCGP2134 Kard 29 0/15 nd na na 0/9 pea pBEBF5 Lav 25 nd 0/25 0% 0% nd Gentian pCGP1482 Kard 27 0/23 nd na na 0/23 pBEGHF48 Lav 23 nd 0/23 0% 0% 0/23 pansy Kard 56 30/33 33/34 58% 12% 21/21 pCGP1967 - 1 - - BP18 SP 36 21/24 18/18 65% 35% 16/21 pansy Kard 22 .15/15 15/15 24% 9% 16/16 pCGP1969 BP40 SP 37 17/17 16/17 80% 54% 11/13 WO 2004/020637 PCT/AU2003/001111 -116 Aihs Av. % F3'5'E plasmid Cult #tg TLC + HPLC+ '. . Northern+ % del del Petunia pCGP1638 Kard 22 0/21 nd na na 0/16 petHfl pCGPI392 Lav. 34 nd 0134 0% 0% nd Petunia petlf2 pCGP2123 Kard 41 0/26 nd na na 0/10 Lavender pBELFS Lay 28 nd 4/28 4% 3.5% nd F3'5'H the F3'5'H sequence contained on the T-DNA plasmid = the plasmid number of the binary vector used in the transformation experiment 5 Cult Rosa hybrida cultivar Kard = Kardinal SP = Soft Promise Lav = Lavande #tg = # of independent transgenic events produced 10 TLC+ = number of individual events that accumulated detectable delphinidin or delphinidin-based molecules (as determined by TLC) in the petals over the number of individual events analyzed HPLC+ = number of individual events that accumulated detectable delphinidin or delphinidin-based molecules (as determined by HPLC) in the petals over the number of 15 individual events analyzed Northern - number of individual events in which the specific intact F3'5fI transcripts were detected by Northern blot analysis in total RNA isolated from petals over the total number of events analyzed 20 nd not donc WO 2004/020637 PCT/AU2003/001111 - 117 The results suggest surprisingly that not all of the F3'5'H sequences assessed (petunia petHII, petunia petHJ2, Salvia Sal#2, Salvia Sal#47, Sollya Sol#5, Butterfly pea BpeaHF2, pansy BP#18, pausy BP#40, Gentian GSen#48, Kennedia Kenn#31 and Lavender LBG) were functional in rose. In fact transcripts of the introduced P3'5'H sequences isolated 5 from Clitoria ternarea (butterfly pea), Gentiana triflora, (gentian) and Petunia hybrida (petunia) faled to accumulate in rose petals. Only full-length F3'5'H transcripts from pansy, salvia, kennedia, sollya and lavender accumulated in rose petals. However although Kennedia F3'5'H transcripts did accumulate in rose petals, there was either no accumulation of the enzyme or the enzyme produced was either not functional or was 10 unable to compete with the endogenous rose F3'H and DFR enzymes to allow for the production of delphinidin or delpbinidin-based molecules pigments. Of the F3'5'H sequences evaluated, only the F3'S'H sequences derived from cDNA clones from Salvia spp. (Sal#2 and Sal#47), Viola spp. (BP#18 and BP#40), Sollya qpp. (Soll5) and Lavandula nil (LBG) resulted in the production of delphinidin or delphinidin-based 15 molecules based pigments in rose petals. Based on the relative percentages of delphinidin or delphinidin-based molecules produced in rose petals, the F3'5'H sequences from pansy (BP#18 and BP#40) were revealed to be the most effective of those assessed at producing delphinidin or delphinidin-based molecules in rose petals. 20 Introduction of Viola spp. F3'5'H sequence into Rosa hybrida cv- Medeo and Pamela As described in the introduction, copigmentation with other flavonoids, further modification of the anthocyanidin molecule and the pH of the vacuole impact on the color produced by anthocyanins. Therefore, selection of rose cultivars with relatively high levels of flavonols and relatively high vacuolar pH would result in bluer flower colors upon 25 production of delphinidin or delphinidin-based molecules pigments. The rose cultivar Medeo generally produces cream-colored to pale apricot flowers (RHSCC 158C to 159A). HPLC analysis of the anthocyanidins and flavonols accumulating in Mcdco rose petals revealed that the petals accumulate high levels of flavonols (2.32 30 mg/g kaempferol, 0.03 mg/g quercetin) and very low levels of anthocyanins (0.004 mg/g WO 2004/020637 PCT/AU2003/001111 -118 cyanidin, 0.004 mg/g pelargonidin). The estimated vacuolar pH of Medeo petals is around 4.6. The rose cultivar Pamela produces white to very pale pink colored flowers. It similarly 5 accumulates low levels of anthocyanin and relatively high levels of flavonols. The T-DNA contained in the construct pCGP1969 (Figure 30) incorporating the pansy P3'5'H clone, BP#40, was also introduced into the rose cultivars Medeo and Pamela resulting in the production of over 90% delphinidin or delphinidin-based molecules in 10 these roses and leading to a dramatic color change and novel colored flowers. The most dramatic color change in transgenic Medeo flowers was to a purple/violet color of RHSCC 70b, 70c, 80c, 186b. The most dramatic color change in transgenic Pamela flowers was to a purple/violet color of RHSCC 71c, 60c, 71a, 80b, 15 In conclusion, two unexpected findings were revealed when gene sequences that had been proven to lead to functionality in petunia and carnation were introduced into roses. First, the petunia F3'5'H petHfl (and petH2) sequences that had resulted in novel color production in carnation and also proven to lead to synthesis of a functional enzyme in 20 petunia did not lead to full-length (or intact) transcript accumulation (as detectable by Northern blot analysis) in rose petals. In fact, there was either no accumulation of full length or intact transcript or the transcripts that were detected were degraded and were seen as low MW (or fast migrating) smears on RNA blots indicating the presence of low MW heterologous hybridizing RNA. Therefore in order to find a F3'5'H sequence that 25 would accumulate in rose and lead to a functional enzyme, a number of F3'5'H sequences were isolated. Again it was not obvious which sequence would lead to an active enzyme in rose petals. All of the F35

T

'H sequences isolated were tested for functionality in canation and/or petunia and all led to. accumulation of intact transcripts and production of a functional F3'5'H activity. However only F3'5'H sequences from pansy (BP#18 and 30 BP#40), salvia (Sal#2 and Sal#47), sollya (Soll#5), kennedia (Kenn#31) and lavender (LBG) resulted in accumulation of intact full-length transcripts and only those from pansy WO 2004/020637 PCT/AU2003/001111 - 119 (BP#18 and BP#40), salvia (Sal#2 and Sal#47), sollya (Soll5) and lavender (LBG) resulted in production of a functional enzyme in rose as measured by the synthesis of delphinidin or delphinidin-based molecules. 5 Secondly that it was not obvious which promoters would be effective in rose. Promoter cassettes that had been tested and proven to be functional in carnation and petunia flowers did not lead to accumulation of detectable transcripts in rose petals. Of the promoters tested in rose, only CaMV 35S, RoseCiS 5', ChrysCHS 5', mas S'and nos 5' promoters led to intact and detectable GUS or nptll or SuRB transcript accumulation in rose petals. 10 Table 14 shows a summary of the results obtained when assessing F3'5'H sequences from various species in petunia, carnation and rose. TABLE 14 Summary of effectiveness of the F3'5'H sequences in petunia, carnation and 15 rose F3'5'N . Petunia Carnation Rose Mal RNA Del RNA Del RNA Kennedia (Kenn#31) + nd nd nd - + Gentian (Gen#48) + + + + - Salvia (Sal#2) + nd + + + + Salvia (Sai#47) + nd + + + + $ollya (Sol#5) + nd + + + + Butterfly pea + nd + + - (BpeaHF2) Pansy (BP#18) + + + + + + Pansy (BP#40) + nd + + + + Petunia (petHfl) + + + + - ~ Petunia (petHf2) + + + + . Lavender (LBG) nd nd nd nd nd + WO 2004/020637 PCT/AU2003/001111 -120 nd = not done Mal = malvidin detected in petals as analysed by TLC Del = delphinidin or delphinidin-based molecules detected in petals as analysed by TLC orHPLC 5 + = yes - = no EXAMPLE 9 Use ofpansy F3'5'H sequences in species other than rose 10 Gerbera From the examples above, it was clear that the pansy F3'S'H sequences, BP#18 and BP#40, resulted in functional F3'5THactivity and lead to the production of high levels of delphinidin or delphinidin-based molecules in roses and carnations. 15 The T-DNA from binary construct pCGPl969 (described in Example 8) (Figure 30) containing the chimeric CaMV 35S: pansy BP#40 F3'5'H: ocs 3' gene expression cassette was introduced into the gerbera cultivar Boogie via Agrobacterium-mediated transformation, to test the functionality of the pansy F3'5'H sequence in gerbera. 20 Of six events produced to date, one (#23407) has produced flowers with a dramatic color change (RHSCC 70e) compared to the control flower color (RHSCC 38a, 38c). The color change of the petals of the transgenic gerbera has been correlated with the presence of delphinidin or delphinidin-based molecules as detected by TLC. 25 Other species In order to produce delphinidin or delphinidin-based molecules pigments in plants that do not normally produce dclphinidin-based pigments and does not contain a flavonoid 3'5' hydroxylase constructs containing a F3'S 'H gene (such as but not limited to a chimaeric 30 Viola App. and/or Salvia spp. and/or Sotlya spp. and/or Lavandula spp. and/or Kennedia spp. F3'5'H gene) are introduced into a species that does not normally produce WO 2004/020637 PCT/AU2003/001111 - 121 delphinidin-based pigments. Such plants may include but are not limited to carnation, chrysanthemum, gerbera, orchids, Euphorbia, Begonia and apple. EXAMPLE 10 5 Characteristics of F3'5'H sequences evaluated in petunia, carnation and rose Gene regulation in eukaryotes is, in simple terms, facilitated by a number of factors which interact with a range of sequences proximal and distal to a nucleotide sequence coding for 10 a given polypeptide. Engineering expression cassettes for introduction into plants for the generation of one or more traits is based on an understanding of gene regulation in eukaryotes in general and, in selected cases, plants in particular. The essential elements include a series of transcriptional regulation sequences typically, but not exclusively, located upstream or 5' to the point of transcription initiation. Such elements are typically 15 described as enhancers and promoters, the latter being proximal to the point of transcription initiation. Immediately downstream from, or 3' to, the initiation of transcription point is a variable region of transcribed DNA which is denoted as the 5' untranslated region (5'utr) which plays a role in transcript stability and translational efficiency. Such sequences, when engineered into expression cassettes, are frequently 20 chimeric and may be derived from sequences naturally occurring adjacent to the coding sequence and/or adjacent to a given promoter sequence. The coding sequence (sometimes disrupted by introns) lies 3' to the 5utr followed by a 3'utr important to transcript (mRNA) stability and translational efficiency. Sequences 3' to the end of the coding region and 3' to the 3'utr itself are denoted as terminator sequences. All these elements make up an 25 expression cassette. In making direct comparisons between promoters or other elements it is important to maintain uniformity in the remaining elements of an expression cassette. Hence, when comparing the efficacy of various F3'5'H sequences it was possible to confine the sequences leading to instability and the subsequent autodegradation of engineered mRNA and resultant absence of tri-hydroxylated products (delphinidin or delphinidin 30 based molecules derivatives) to the region coding for the F3'5'H and not to other elements in the expression cassette such as 5'utr and/or 3'utr sequences for example.

WO 2004/020637 PCT/AU2003/001111 -122 In an attempt to identify motifs or similarities between the F3'5'H sequences that resulted. in full-length transcripts being detected in total RNA isolated from rose flowers, and ultimately delphinidin or delphinidin-based molecules production, comparisons across a 5 range of parameters were performed. These included sequence identities at nucleic acid and amino acid levels, sequence alignments, taxonomic classifications, % of A or T nucleotides present in the sequence, % of codons with an A or T in the third position etc. Taxonomic classfication 10 The taxonomy of each species from which the P3'5'H sequences were isolated was examined (Table 15). There appeared to be no obvious link' between the subclass classification and whether the F3'S'H sequence resulted in an intact transcript and 'subsequent delphinidin or delphinidin-based molecules production in roses. 15 Table 15 Taxonomic classifications of the species that F3'S'H sequences were isolated from and whether the use of the sequences resulted in intact transcript in rose petals that were detectable by RNA blot analysis. Intact Delphinidin Flower Species Family Order Subclass transcript in rose petals Gendana NO NO gentian Gentianaceae Gentianales Asteridae triflora lavender Lavandula nil Lamiaceac. Lanilcs Asteridae YES YES salvia Salvia spp. Lamiaceae Lamiales Asteridae YES YES sollya Solla sopp, Pittospoaceae Apiales Asteridae YES YES petunia Petunia hybrida Solanaceae Solanales Asteridae NO NO kennedia Kennedia spp. Fabaccac Fabales Rosidac YES NO butterfly CItriabaca Fabae Rosidac NO NO pea ternatea pansy Viola pfp; Violaceae Malpighiales Rosidae YES YES rose Rosa hybrida Rosaciae Rosales Rosidae na na WO 2004/020637 PCT/AU2003/001111 - 123 Intact transcript fall-length F3'5'H mRNA detected by NoTtbern blot analysis in total RNA isolated from petals from transgenic roses Comparison of F3'5'H nucleotide -sequences 5 The nucleotide sequence identities between each of the F3'JH sequences evaluated were determined using the ClustalW program (Thompson et al., 1994, supra) within the MacVectorTM version 6.5.3 application program (Oxford Molecular Ltd., England) (Table 16). There were no obvious differences between the F3'5'1N sequences that resulted in the detection of intact full-length transcripts in RNA isolated from rose petals and those that 10 didn't. Table 16: Percentage of nucleic acid sequence identity between the nucleotide sequences of the F3'5'H isolated from various species. F3'S'H sequences that resulted in intact transcripts being detected in RNA isolated from rose petals are underlined and in italics. 15 IWLB &P40 Lat' &La47 Sg2 Sii K909 Bpea Gent PetIMI PetHf2 BP1S 100 82 60 61 62 51 60 62 62 59 62 OP40 100 60 57 58 50 59 62 58 60 62 Lam 100 68 68 48 57 57 58 59 60 SaI47 100 95 48 56 57 59 57 58 SL 100 49 57 58 60 57 59 sl 100 48 50 50 48 51 Kentn 100 70 56 64 60 Bpea 100 59 59 62 Gent 100 61 64 PetHfl 84 Pet~t2 100 WO 2004/020637 PCT/AU2003/001111 -124 Comparison of F3 '5'H translated nucleotide sequences The translated nucleotide sequence identities and similarities between each of the F3 'W' sequences evaluated were also determined using the ClustalW program (Thompson et al.. 1994, supra) within the MacVectorTm version 6.5.3 application program (Oxford 5 Molecular Ltd., England) (Table 17), There were no obvious differences between the F3'5'1 sequences that resulted in the detection of intact ful-length transcripts in RNA isolated from rose petals and those that didn't. Table 17: Percentage of the amino acid sequence identity and similarity (in brackets) 10 between F3'5'H sequences isolated from various species. F3'5'H sequences that resulted in intact transcripts being detected in RNA isolated from rose petals are underlined and in italics, PF18 F4_0 Lay A-W4 5a Soft Kenn .Bpea Gent I PetHil| PtUE2 SPIR 100 91 (94) 65 (77) 65(76) 65 (76) 44(63) 69(83) 64(75) 69 (80) 74(85) 74(85) SF40 100 67(9) 66 (77) 66 (77) 46(64) 69(82) 64(75) 68(79) 74(85) 75(86) Lwa 100 75 (86) 73 (86) 45 (63) 63 (79) 59(74) 66(80) 68 (82) 69(83) 5a147 100 98 45(65) 64(78) 60(72) 64(76) 68 (79) 69 (81) Sag 100 45 (65) 64(78) 60(72) 63(75) 68 (79) 69(81) g 100 46(66) 41 (61) 44(62) 46(67) 46(66) Kgn. 100 72(80) 65(73) 71(83) 72 (83) ea 100 69(81) 65(75) 65(74) Gent 100 73 (82) 73 (82) 100 98(95) 1 2etH 100 2 15 WO 2004/020637 PCT/AU2003/001111 - 125 Percentage of nucleotides A or T in the F3'S'H DNA sequences There is some evidence to suggest that the choice of codons influences the rate of translation and mRNA degradation. Certain codons are used less frequently than others are and this may be related to the abundance of isoaccepting tRNAs. Transfer RNAs 5 corresponding to rare codons are less abundant in E.coli and yeast than tRNAs corresponding tb preferred codons (van Hoof and Green, Plant Molecular Biology, 35: 383-387, 1997). Examples of altering codon usage and making a gene more "plant-like" are the bacterial B.t. toxin gene (reviewed in Diehn et al., Genet Engin, 18: 83-99, 1996) and the jellyfish gfp gene (Haseloff et al., Proc. Nat. Acad. Set USA, 94: 2122-2127, 10 1997). However as commented in van Hoof and Green, (1997) (supra), the effect of eliminating the rare codons in the At. genes increased the GC content thereby eliminating AU-rich sequences that may be responsible for improper recognition of introns and polyadenylation sites as well as removing instability determinants. Alteration of codon usage in the jellyfish gf gene also resulted in removal of a cryptic 'intron (Haseloff et al, 15 1996, supra). Studies examining the effect of codon usage and instability elements have generally been limited to differences between genes isolated from species in different kingdoms ie. bacterial versus yeast versus animal versus plant. Within the plant kingdom, differences have been observed between the dicotyledons and the monocotyledons. Studies on transgenic plants have suggested that promoter fragments used to drive gene expression 20 in dicotyledonous plants are not as effective when used in monocotyledonous plants (see Galun and Breiman, Transgenic Plants, Imperial College Press, London, England, 1997). Differences in the methylation and ultimate expression of a DFR transgene in Petunia hybrida (dicot) were detected when a maize (monocot) DFR cDNA was compared with a gerbera (dicot) DFR cDNA (Elomaa et al, Molecular and General Genetics, 248: 649 25 656, 1995). The conclusion was that the gerbera DFR cDNA had a higher AT content (lower GC content) and was more "compatible" with the genomic organization of petunia preventing it being recognised as a foreign gene and hence silenced by methylation. (Rose along with carnation and petunia are dicotyleduu and the F3'5'H genes tested were all isolated from dicotyledonous plants.) These points serve to illustrate that degradation and 30 stability mechanisms are not understood in detail and differences appear between plants and other kingdoms and within the plant kingdom.

WO 2004/020637 PCT/AU2003/001111 -126 The content of A and T was examined in the F3'5'H cDNAs evaluated along with that of four flavonoid pathway genes (F3'L DFP, CHS, FLS) that had been isolated from rose (Table 18). The third position of each codon (within the open reading frame) was also 5 examined and the percentage of codons with an A or a T in the third position was calculated (Table 18). Table 18: Summary of the percentage amount of A or T dinucleotides in the F3'5'H sequences isolated and whether the F3'5'H resulted in full-length transcripts being 10 detected in rose petals by Northern blot analysis. % A or T in F3'5'H seq %AT RNA Delphinidin 3rd Viola BP#18 50 40 YES YES Viola BP#40 51 35 YES YES Salvia#2 48 .33 YES YES Salvia#47 48 34 YES YES Sollya#5 54 54 YES YES LavenderLBG 50 37 YES* YES Kennedia#31 54 47 YES NO petuniapetfl 61 66 NO NO petuniapethrj2 59 65 NO NO Gentian#48 57 57 NO NO Butterfly 57 53 NO NO pea#HF2 rose F3'H 47 34 na rose CHS 52 42 ** na WO 2004/020637 PCT/AU2003/001111 - 127 F'3'5'H seq %AT A or T in A Delphinidin 3rd 10SeDFR 53 46 na rose FLS 56 43 na %AT % of nucleotides that are A or T in the nucleic acid sequence %A or T in 3 rd= the percentage of codons that have an A or T in the third position RNA = whether a full-length mRNA transcript was detected by Northern blot 5 analysis in total RNA isolated from rose petals Del = whether any delphinidin or delphinidin-based molecules was detected by TLC or HPLC in rose petals YES* =although Northern blot analysis of transgenic roses transformed with the lavender F3'S 'H expression cassettes was not performed, it can 10 be assumed that full-length transcript was produced since delphinidin or delphinidin-based molecules was detected in the rose petals. rose F3'H (described in International Patent Application No. PCT/AU97/00124) rose DFR (Tanaka et al., 1995, supra) rose FLS (GeniBank accession niunber AB03 8247) 15 rose CHS (GenBank accession number A13038246) The AT content of the four rose sequences (above) encoding flavonoid pathway enzymes had an AT content of between 47 and 56%. In general the AT content of the F3'5H sequences that resulted in intact transcripts in rose petals was between 48 and 54%. 20 However the F3H'5 sequences that did not result in intact transcripts accumulating in rose petals generally had a higher AT content of between 57 and 61%. Hence the AT content of the introduced F3'5'H genes into rose may be a factor in whether an intact transcript accumulates in rose petals and so leads to production of F3'5'H and delphinidin or delphinidin-based molecules. 25 WO 2004/020637 PCT/AU2003/001111 - 128 The nucleotide base at the third position of each codon of the four rose sequences encoding flavonoid pathway enzymes was generally an A or a T in 34 to 46% of the codons. In general F3'5'H sequences that resulted in intact transcripts in rose petals contained an A or a T in the third position of each codon in 33 to 54% of the codons. However the P3'5'H 5 sequences that did not result in intact transcripts accumulating in rose petals generally contained an A or a T in the third position of each codon in 53 to 66% of the codons. So the percentage of codons with an A or a T in the third position of the introduced P3'5'H genes into rose may also be a factor in whether an intact transcript is accumulates in rose petals and so leads to production of F3'5'H and dclphinidin or delphinidin-based 10 molecules. It may be that by altering the overall content of the nucleotides A and/or T in any F3'5'! DNA sequence that does not result in an intact transcript in rose such as but nor limited to the Petunia hybrida petHf, Petunia hybrida petlIf2, Clitora ternatea (butterfly pea) 15 BpeaHF2 or Gentiana triflora (gentian) Gen#48, to a level more consistent with that found in rose genes, intact transcripts will accumulate and result in the efficient translation of F,3'7'H transcripts and so to delphinidin or delphinidin-based molecules accumulation in rose petals. One way of altering the AT content of the DNA sequence without altering the amino acid sequence is to target the degeneracy of the third position of each codon. 20 Those skilled in the art will appreciate that the invention described herein is susceptible to variations and modifications other than those specifically described. It is to be understood that the invention includes all such variations and modifications. The invention also includes all of the steps, features, compositions and compounds referred to or indicated in 25 this specification, individually or collectively, and any and all combinations of any two or more of said steps or features, WO 2004/020637 PCT/AU2003/001111 -129 BIBLIOGRAPHY A look beyond transcription, ASPP, USA, Bailey-Serres and Gallie, eds, 1998 Altschul el al, J. Mo. Biot 215(3): 403-410, 1990. Altschul et al., Nucl. Acids Res. 25: 33 89-3402, 1997. Ausubel et al., "Current Protocols in Molecular Biology" John Wiley & Sons Inc, 1994 1998, Chapter 15. Aviv and Leder, Proc. Natl Acad. Sci. USA 69: 1408, 1972. Bevan, Nucleic Acids Res 12: 8711-8721, 1984. Bonner and Laskey, Eur. J. Biochem, 46: 83, 1974. Bodeau, Molecular and genetic regulation of Bronze-2 and other maize anthocyanin genes. Dissertation, Stanford University, USA, 1994 Brendel et al, In A look beyond transcription, ASPP, USA, Bailey.Serres and Gallie, eds, 1998. Brouillard and Dangles, In: The Flavonoids -Advances in Research since 1986. Harborne, J.B. (ed), Chapman and Hall, London, UK, 1-22, 1993. Bmgliera et at, Plant J. 5: 81-92, 1994. Brugliera Characterization of floral specific genes isolated from Petunia hybrida. RMIT, Australia. PhD thesis, 1994.

WO 2004/020637 PCT/AU2003/001111 -130 Bullock et at, Biotechniques 5: 376, 1987. Comai et al., Plant Mo. Biot 15: 373-381, 1990.. Depioker et al, J Mol and Appl. Genetics 1: 561-573, 1982. Diehn et al, Genet Engin, 18: 83-99, 1996 Elomaa at al., Molecular and General Genetics, 248: 649-656, 1995 Frank et al, Cell 21: 285-294, 1980. Galun and Breiman, Transgenic Plants, Imperial College Press, London, England, 1997 Guilley et al, Cell 30: 763-773. 1982. Hanahan, J. Mol Biol 166: 557, 1983. Harpster et al., MGG 212: 182-190, 1988. Haseloff et al., Proc. Natl. Acad. Sci USA, 94: 2122-2127, 1997 Holton and Cornish, Plant Cell 7:1071-1083, 1995. Holton and Graham, Nuc. Acids Res. 19: 1156, 1990. Holton et at., Nature 366 276-279, 1993a. Holton et al., Plant J. 4:1003-1010, 1993b WO 2004/020637 PCT/AU2003/001111 -131 Holton, Isolation and characterization of petal specific genes from Petunia hybrida. PhD thesis, University of Melbourne, Australia, 1992 Huang and Miller, Adv. Apple. Math, 12: 373-381, 1991. Inoue et al., Gene 96: 23-28, 1990. Janssen and Gardner, Plant Molecular Biolog), 14: 61-72, 1989 Jefferson, et al., EMBO J 6: 3901-3907, 1987. Johnson et al., In A look beyond transcription, ASPP, USA, Bailey-Serres and Gallie, eds, 1998. Katn et al, Cene 32: 217-224, 1984, Lazo et al. Bio/Technology 9: 963-967, 1991. Leeetal.,EMBOJ 7: 1241-1248, 1988. Lu et al., Bia/Technology 9: 864-868, 1991. Marchant et al, Molecular Breeding 4:187-194, 1998. Marmur and Doty, J M1. Biol 5: 109, 1962. McBride and Summerfelt, Plant Molecular Biology 14: 269-276, 1990, Merrifield. J. Am. Chem. Soc. 85: 2149, 1964. Mitsuhara aet al., Plant Cell Physiol. 37: 49-59, 1996.

WO 2004/020637 PCT/AU2003/001111 -132 Mol et al, Trends Plant Sci. 3: 212-217, 1998. Pearson and Lipman, Proc. Natl, Acad. Sci. USA 85(8): 2444-2448, 1988, Plant Molecular Biology Labfax, Croy (ed), Bios scientific Publishers, Oxford, UK, 1993. Plant Molecular Biology Manual ( 2 "d edition), Galvin and Schilperoot (eds), Kluwer Academic Publisher, The Netherlands, 1994. Robinson and Firoozabady, Scientia Horticulturae 55: 83-99, 1993. Rout et al., Sciendta Horticulturae 81: 201-238, 1999. Sambrook et aL, Molecular Cloning: -A Laboratory Manual, Cold Spring Harbor Laboratories, Cold Spring Harbor, NY, USA, 1989. Sambrook and Russell, Molecular Cloning: A Laboratory Manual 3 rd edition, Cold Spring Harbor Laboratories, Cold Spring Harbor, NY, USA, 2001 Seitz and Hinderer, Anthocyanins. In: Cell Culture and Somatic Cell Genetics of Plants. Constabel, F. and Vasil, IK. (ads.), Academic Press, New York, USA, 5: 49-76, 1988. Short et al, Nucl. Acids Res. 16: 7583-7600, 1988. Sommer and Saedler, Mol Gen. Gent., 202: 429-434, 1986. Strack and Wray, In: The Flavonolds -Advances in Research since 1986. Harbome, J.B. (ed), Chapman and Hall, London, UK, 1-22, 1993. Tanaka et al., Plant Cell Physiol. 36: 1023-1031, 1995.

WO 2004/020637 PCT/AU2003/001111 - 133 Tanaka eta?., Plant Cell Physiol. 37: 711-716, 1996. Thompson et al, Nucleic Acids Research 22: 4673-4680, 1994. Turpen and Griffith, BioTechniques 4: 11-15, 1986. van Hoof and Green, Plant Molecular Biology, 35: 383-387, 1997 Winkel-Shirley, Plant Physiol 126: 485-493, 2001a. Winkel-Shirley, Plant Physic!. 127:1399-1404,2001b. Yonekura-Sakakibara et al, Plant Cell Physiol. 41: 495-502, 2000.

Claims (92)

1. An isolated nucleic acid molecule comprising a sequence of nucleotides encoding or complementary to a sequence encoding a flavonoid 3', 5' hydroxylase (F3'5'H) or a polypeptide having F3'5'H activity wherein expression of said nucleic acid molecule in a rose petal tissue results in detectable levels of delphinidin or delphinidin based molecules as measured by a chromatographic technique.

2. An isolated nucleic acid molecule comprising a sequence of nucleotides encoding or conplemcntary to a sequence encoding a F3'5'H or a polypeptide having F3'5'H activity wherein expression of said nucleic acid molecule in a rose petal tissue results in a sufficient level and length of transcript which is translated to said F3'5'H as determined by detectable levels of delphinidin or delphinidin-based molecules as measured by a chromatographic technique.

3. The isolated nucleic acid molecule of claim 1 or 2 wherein expression of said nucleic acid molecule in said rose petal results in a visually detectable colour change.

4. The isolated nucleic acid molecule of any one of claims lto3,.wherein the nucleic acid molecule is derived frorn a plant selected from the list comprising a Viola spp., Salvia spp., Sollya spp., Lavandula spp. ;and Kennedia spp.

5. The isolated nucleic acid molecule of claim 4, wherein the nucleic acid molecule is derived from a Viola spp. plant.

6. The isolated nucleic acid molecule of claim 5, wherein the nucleic acid molecule is derived from the Viola spp., cultivar Black Pansy. WO 2004/020637 PCT/AU2003/001111 -135

7. The isolated nucleic acid molecule of any one of claims 5 or 6, wherein the nucleotide sequence encodes a F3'5' comprising an amino acid sequence selected from SEQ ID NO:10, SEQ ID NO:12, an amino acid sequence having at least about 40% similarity to SEQ ID NO:10 and an amino acid sequence having at least about 40% similarity to SEQ ID NO: 12.

8. The isolated nucleic acid molecule of claim 7, comprising a nucleotide sequence selected from SEQ ID NO:9, SEQ ID NO: 11, a nucleotide sequence having at least about 40% identity to SEQ ID NO:9, a nucleotide sequence having at least about 40% identity to SEQ ID NO:11, a nucleotide sequence capable of hybridizing to SEQ ID NO:9 or its complement under low stringency conditions and a nucleotide sequence capable of hybridizing to SEQ ID NO:11 or its complement under low stringency conditions.

9. The isolated nucleic acid molecule of claim 8, comprising the nucleotide sequence set forth in SEQ ID NO:9.,

10. The isolated nucleic acid molecule of claim 8, comprising the nucleotide sequence set forth in SEQ ID NO:11.

11. The isolated nucleic acid molecule of claim 4, wherein the nucleic acid molecule is derived from SaVia spp..

12. The isolated nucleic acid molecule of claim 11, wherein the nucleotide sequence encodes a F3'5'H comprising an amino acid sequence selected from SEQ ID NO:14, SEQ ID NO:16, an amino acid sequence having at least about 40% similarity to SEQ ID NO:14 and an amino acid sequence having at least about 40% similarity to SEQ ID NO:16. WO 2004/020637 PCT/AU2003/001111 -136

13. The isolated nucleic acid molecule of claim 12, comprising a nucleotide sequence selected from SEQ ID NO:13, SEQ ID NO:15, a nucleotide sequence having at least about 40% identity to SEQ ID NO:13, a nucleotide sequence having at least about 40% identity to SEQ ID NO; 15, a nucleotide sequence capable of hybridizing to SEQ ID NO:13 or its complement under low stringency conditions and a nucleotide sequence capable of hybridizing to SEQ ID NO: 15 or its complement under low stringency conditions.

14. The isolated nucleic acid molecule of claim 13, comprising the nucleotide sequence set forth in SEQ ID NO:13.

15. The isolated nucleic acid molecule of claim 13, comprising the nucleotide sequence set forth in SEQ ID NO:15.

16. The isolated nucleic acid molecule of claim 4, wherein the nucleic acid molecule is derived from Sollya spp.

17. The isolated nucleic acid molecule of claim 16, wherein the nucleotide sequence encodes a F3'5'H comprising an amino acid sequence selected from SEQ ID NO: 18 and an amino acid sequence having at least about 40% similarity to SEQ ID NO:1 8.

18. The isolated nucleic acid molecule of claim 17, comprising a nucleotide sequence selected from SEQ ID NO:17, a nucleotide sequence having at least about 40% identity to SEQ ID NO:17 and a nucleotide sequence capable of hybridizing to SEQ ID NO: 17 or its complement under low stringency conditions.

19. The isolated nucleic acid molecule of claim 18, comprising the nucleotide sequence set forth in SEQ ID NO:17.

20. The isolated nucleic acid molecule of claim 4, wherein the nucleic acid molecule is derived from Kennedia spp. WO 2004/020637 PCT/AU2003/001111 -137

21. The isolated nucleic acid molecule of claim 20, wherein the nucleotide sequence encodes a F3'5'H comprising an amino acid sequence selecting from SEQ ID NO;27 and an amino acid sequence having at least about 40% similarity to SEQ ID NO:27.

22. The isolated nucleic acid molecule of claim 21, comprising a nucleotide sequence selected from SEQ ID NO:26, a nucleotide sequence having at least about 40% identity to SEQ ID NO:26 and a nucleotide sequence capable of hybridizing to SEQ ID NO:26 or its complement under low stringency conditions,

23. The isolated nucleic acid molecule of claim 22, comprising the nucleotide sequence set forth in SEQ ID NO:26.

24. The isolated nucleic acid molecule of claim 4, wherein the nucleic acid molecule- is derived from Lavandula spp.

25. The isolated nucleic acid molecule of claim 24, wherein the nucleotide sequence encodes a F3'S'U comprising an amino acid sequence selected from SEQ ID NO:32 and an amino acid sequence having at least about 40% similarity to SEQ ID NO:32.

26. The isolated nucleic acid molecule of claim 25, comprising a nucleotide sequence selected from SEQ ID NO:31, a nucleotide sequence having at least about 40% identity to SEQ ID NO:31 and a nucleotide sequence capable of hybridizing to SEQ ID NO:31 or its complement under low stringency conditions,

27. The isolated nucleic acid molecule of claim 26, comprising the nucleotide sequence set forth in SEQ ID NO:31. WO 2004/020637 PCT/AU2003/001111 - 138

28. The isolated nucleic acid molecule of any one of claims 1- to 4, wherein the nucleotide sequence comprises an overall percentage of less than or equal to 54% of the nucleotides (i) Aor (ii) T, or (iii) A and T in the third nucleotide position of each codon.

29. A construct comprising a sequence of nucleotides comprising: (i) a promoter which is operable in rose petal tissue and wherein said promoter is opetably linked to, (ii) a nucleic acid molecule encoding F3'5'H, or complementary to a sequence encoding a F3'5'H or a polypeptide having F3'5'H activity wherein expression of said nucleic acid molecule in a rose petal tissue results in detectable levels of delphinidin or delphinidin-based molecules as measured by a chromatographic technique and wherein said nucleic acid molecule is derived from a plant selected from the group consisting of a Viola spp., Salvia spp., Sollya spp., Lavandula spp. and Kennedia spp.

30. A construct comprising a sequence of nucleotides comprising: (i) a promoter which is operable in rose petal tissue and wherein said promoter is operably linked to, (ii) a nucleic acid molecule encoding F3'5'H, or complementary to a sequence encoding a F3'5'H or a polypeptide having F3'5'H activity wherein expression of said nucleic acid molecule in a rose petal tissue results in a sufficient level and length of transcript which is translated to said F3'5'H as determined by detectable levels of delphinidin or delphinidin-based molecules as measured by a chromatographic technique. WO 2004/020637 PCT/AU2003/001111 -139

31. The construct of claim 29 or 30 , wherein expression of said construct in said rose petal results in a visually detectable colour change.

32. The construct of any one of claims 29 to 31, wherein said promoter is selected from the group consisting of rose CHS, chrysanthemum CBS and CaMV 35S.

33. A construct of any one of claims 29 to 31 wherein said promoter comprises SEQ ID NO:5, or a functional equivalent thereof.

34. A construct of any one of claims 29 to 31 wherein said promoter comprises SEQ ID NO;30, or a functional equivalent thereof.

35. The construct of any one of claims 29 to 34 , wherein the nucleic acid molecule is derived from a Viola spp.

36. The isolated nucleic acid molecule of claim 35, wherein the nucleotide sequence encodes F3'5'H comprising an amino acid sequence selected from SEQ ID No: 10, SEQ ID NO:12, an amino acid sequence having at least about 40% similarity to SEQ ID NO: 10 and an amino acid sequence having at least about 40% similarity to SEQ ID NO:12.

37. The isolated nucleic acid molecule of claim 36, comprising a nucleotide sequence selected from SEQ ID NO:9, SEQ ID NO:11, a nucleotide sequence having at least about 40% identity to SEQ ID N0:9, a nucleotide sequence having at least about 40% identity to SEQ ID NO:11, a nucleotide sequence capable of hybridizing to SEQ ID NO:9 or its complement under low stringent conditions and a nucleotide sequence capable of hybridizing to SEQ ID NO iI or is complement under low stringent conditions,

38. The isolated nucleic acid molecule of claim 37, wherein the gene comprises the nucleotide sequence set forth in SEQ ID NO:9. WO 2004/020637 PCT/AU2003/001111 - 140

39. The isolated nucleic acid molecule of claim 37, wherein the gene comprises the nucleotide sequence set forth in SEQ ID NO: 11.

40. The isolated nucleic acid molecule of any one of claims 29 to 34, wherein. the nucleic acid molecule is derived from Salvia app.

41. The isolated nucleic acid molecule of claim 40, wherein the gene comprises a nucleotide sequence encoding F3'5'H comprising an amino acid sequence selected from SEQ ID NO:14, SEQ ID NO:16, an amino acid sequence having at least about 40% similarity to SEQ ID NO:14 and an amino acid sequence having at least about 40% similarity to SEQ ID NO:16.

42. The isolated nucleic acid molecule of claim 41, wherein the gene comprises a nucleotide sequence selected from SEQ ID NO:13, SEQ ID NO:15, a nucleotide sequence having at -least about 40% identity to SEQ ID NO:13, a nucleic sequence having at least about 40% identity to SEQ ID NO:15, a nucleotide sequence capable of hybridizing to SEQ ID NO:13 or its complements under low stringent conditions and a nucleotide sequence capable of hybridizing to SEQ ID NO:15 or is complement under low stringent conditions.

43. The isolated nucleic acid molecule of claim 42, wherein the gene comprises the nucleotide sequence set forth in SEQ ID NO: 13.

44. The isolated nucleic acid molecule of claim 42, wherein the gene comprises the nucleotide sequence set forth in SEQ D NO:15.

45, The isolated nucleic acid molecule of any one of claims 29 to 34, wherein the gene is derived from Sollya spp. WO 2004/020637 PCT/AU2003/001111 - 141

46. The isolated nucleic acid molecule of claim 45, wherein the gene encodes a F3'5'H comprising an amino acid sequence selected from SEQ ID NO:18 and an amino acid sequence having at least about 40% similarity to SEQ ID NO:18.

47. The isolated nucleic acid molecule of claim 46, wherein the gene comprises a nucleotide sequence selected from SEQ ID NO: 17, a nucleotide sequence having at least about 40% identity to SEQ ID NO: 17 and a nucleotide sequence capable of hybridizing to SEQ ID NO: 17 or its complements under low stringent.

48. The isolated nucleic acid-molecule of claim 47, wherein the gene comprises the nucleotide sequence set forth in SEQ ID NO:17.

49. The isolated nucleic acid molecule of any one of claims 29 to 34, wherein the gene is derived from Kennedia spp.

50. The isolated nucleic acid molecule of claim 49, wherein the gene encodes F3'5'H comprising an amino acid sequence selected from SEQ ID NO:27 and an amino acid sequence having at least about 40% similarity to SEQ ID NO:27,

51. The isolated nucleic acid molecule of claim 50, wherein the gene comprises a nucleotide sequence selected from SEQ ID NO:26, a nucleotide sequence having at least about 40% identity to SEQ ID NO:26 and a nucleotide sequence capable of hybridizing to SEQ ID NO:26 or its complements under low stringent.

52. The isolated nucleic acid molecule of claim 51, wherein the gene comprises the nucleotide sequence set forth in SEQ ID NO:26.

53. The isolated nucleic acid molecule of any one of claims 29 to 34, wherein the gene is derived from Lavandula spp. WO 2004/020637 PCT/AU2003/001111 -142

54. The isolated nucleic acid molecule of claim 53, wherein the gene encodes F3'5'H comprising an amino acid sequence selected froxn SEQ ID NO:32 and an amino acid sequence having at least about 40% similarity to SEQ ID NO:32.

55. The isolated nucleic acid molecule of claim 54, wherein the gene comprises a nucleotide sequence selected from SEQ ID NO:31 and a nucleotide sequence having at least about 40% identity to SEQ ID NO:31, a nucleotide sequence capable of hybridizing to SEQ ID NO:31 or its complements under low stringent conditions.

56. The isolated nucleic acid molecule of claim 55, wherein the gene comprises the nucleotide sequence set forth in SEQ ID NO:31.

57. A method for producing a transgenic flowering plant capable of synthesizing a F3'5'H, said method comprising stably transforming a cell of a suitable plant with a nucleic acid sequence as defined in any one of claims .1 to 28, under conditions permitting the eventual expression of said nucleic acid sequence, regenerating a transgenic plant from the cell and growing said transgenic plant fbr a time and under conditions sufficient to permit the expression of the nucleic acid sequence.

58. A method for producing a transgenic plant with reduced indigenous or existing F3'5'H activity, said method comprising stably transforming a cell of a suitable plant with a nucleic acid molecule as defined in any one of claims 1 to 28, regenerating a transgenic plant from the cell and where necessary growing said transgenic plant under conditions sufficient to permit the expression of the nucleic acid.

59. A method for producing a genetically modified plant with reduced indigenous or existing F3'5'H activity, said method comprising altering the F3 TH gene through modification of the indigenous sequences via homologous recombination from an appropriately altered FS' H gene as defined in any one of claims 1 to 28, or a derivative or part thereof introduced into the plant cell, and regenerating the genetically modified plant from the-cell. WO 2004/020637 PCT/AU2003/001111 -143

60. A method for producing a wtansgenic flowering plant exhibiting altered inflorescence properties, said method comprising stably transforming a cell of a suitable plant with a nucleic acid sequence as defted in any one of claims 1 to 28, regenerating a transgenic plant from the cell and growing said transgenic plant for a time and under conditions sufficient to permit the expression of the nucleic acid sequence.

61. A method for producing a flowering plant exhibiting altered inflorescence properties, said method comprising alteration of a F3'5'H gene as defined in any one of claims 1 to 28, through modification of the indigenous sequences via homologous recombination from an appropriately altered F3'5'H gene or derivative or part thereof introduced into the plant cell, and regenerating the genetically modified plant fRom the cell.

62. - A method for producing a transgenic plant capable of expressing a recombinant gene encoding F3'5'H as defined in any one of claims 1 to 28, or part thereof or which carries a nucleic acid sequence which is substantially complementary to all or a part of an mRNA molecule encoding said F3'5'H, said method comprising stably transforming a cell of a suitable plant with the isolated nucleic acid molecule comprising a sequence of nucleotides encoding, or complementary to a sequence encoding F3'1H, where necessary under conditions permitting the eventual expression of said isolated nucleic acid molecule, and regenerating a transgenic plant from the cell.

63. A genetically modified plant or part thereof or cells therefrom comprising an isolated nucleic acid molecule of any one of claims 1 to 28.

64. A genetically modified plant or part thereof or cells therefrom comprising an isolated nucleic acid molecule of any one of claims 1 to 28 or comprising a reduced level of expression of a nucleic acid molecule of any one of claims 1 to 28. WO 2004/020637 PCT/AU2003/001111 - 144

65. A genetically modified plant or part thereof or cells therefrom comprising an isolated nucleic acid molecule of any one of claims 1 to 28 or comprising an increased level of expression of a nucleic acid molecule of any one of claims I to 28,

66. The genetically modified plant or part thereof or cells therefrom any one of claims 63 to 65 , wherein the plant part is selected from sepal, bract, petiole, peduncle, ovaries, anthers, flowers, fruits, nuts, roots, stems, leaves, seeds.

67. The genetically modified plant or part thereof or cells therefrom of any one of claims 63 to 66, wherein the plant is a horticultural species, agricultural species or ornamental species.

68. Use of an isolated nucleic acid molecule as defined in any one of claims 1 to 28, in the manufacture of a genetic construct capable of expressing F3'5'H1 or down regulating an indigenous F3'5'H1 enzyme in a plant,

69. A. gene silencing construct comprising an isolated nucleic acid molecule as defined in any one of claims 1 to 28 or a complex thereof.

70. The genetically modified plant or part thereof or cells therefrom of 63 to 66, wherein the plant is selected from a rose, carnation, lisianthus, petunia, lily, pansy, gerbera, chrysanthemum, geranium, Torenia, Begonia, Cyclamen, Nierembergia, Catharanthus, Pelargonium, orchid, grape, apple, Euphorbia or Puchsia.

71. An extract from a genetically modified plant or part thereof or cells therefrom from any one of claims 63 to 67 and 70.

72. The extract of claim 71, wherein the extract is a flavouring or food additive or health product orbeverage or juice or colouring. WO 2004/020637 PCT/AU2003/001111 - 145

73. The method of any one of claims 57 to 62 wherein the genetically modified plant or part thereof or cells therefrom exhibit altered inflorescence.

74. An isolated recombinant F3'5'H or peptide having F3'5'H activity encoded by a nucleic acid molecule as defined in any one of claims 1 to 28.

75. The isolated recombinant F3'5'H or peptide having F3'5'H activity of claim 74, wherein the recombinant F3'5'H or peptide having F3'5'H activity is a fusion molecule comprising two or more heterologous amino acid sequences.

76. An isolated recombinant F3'5'H or peptide having F3'5'H activity nucleic acid molecule of any one of claims 1 to 28 comprising a fusion of two or more heterologous nucleotide sequences.

77. A prokaryotic organism carrying a genetic sequence encoding a F3'5'H molecule according to any one of claims 1 to 28 extrachromasomally in plasmid form.

78. A eukaryotic organism carrying a genetic sequence encoding a F3'S'H molecule according to any one of claims I to 28 extrachromasomally in plasmid form.

79. The use of a nucleic acid molecule of any one of claims 1 to 28 in the manufacture of a genetically modified plant or part thereof or cells therefrom.

80. The genetically modified plant or part thereof or cells therefrom of claim 79, wherein the genetically modified plant or part thereof or cells therefrom exhibitsaltered flowers or infloresocnce.

81. The use of a nucleic acid sequence as defined in any one of claims 1 to 28 in the manufacture of a genetic construct capable of expressing F3'5'H or down-regulating an indigenous F3'5'H enzyme in a plant. WO 2004/020637 PCT/AU2003/001111 -146

82. An isolated nucleic acid molecule comprising a sequence of nucleotides encoding or complementary to a sequence encoding a F3'5'H or a polypeptide having F3'5'H activity, wherein said nucleic acid molecule is derived from butterfly pea.

83. The isolated nucleic acid molecule of claim 81, wherein the nucleotide sequence encodes a F3'5'H comprising an amino acid sequence selecting for SEQ ID NO:21 and an amino acid sequence having at least about 40% similarity to SEQ ID NO:21,

84, The isolated nucleic acid molecule of claim 83, comprising a nucleotide sequence selected from SEQ ID NO:20, a nucleotide sequence having at least about 40% identity to SEQ ID NO:20 and a nucleotide sequence capable of hybridizing to SEQ ID NO:20 or its complement under low stringency conditions.

85. The isolated. nucleic acid molecule of claim 84, comprising the nucleotide sequence set forth in SEQ ID NO:20.

86. The isolated nucleic acid molecule of any one of claims I to 4, wherein the nucleotide sequence comprises an overall percentage of less than or equal to 55% of the nucleotides (i) A, or (ii) T, or (iii) A and T.

87. An isolated nucleic acid molecule comprising SEQ ID NO:5 or a functional equivalent thereof.

88. An isolated nucleic acid molecule comprising SEQ ID NO:30 or a functional equivalent thereof. WO 2004/020637 PCT/AU2003/001111 -147

89. An isolated nucleic acid molecule which has been modified so as to comprise a sequence of nucleotides encoding or complementary to a sequence encoding F3'5'H or a polypeptide having F3'5'1 activity wherein expression of said nucleic acid molecule in a rose petal tissue results in detectable levels of delphinidin or delphinidin based molecules as measured by a chromatographic technique wherein the nucleic acid molecule is derived from a plant selected from the list comprising Petunia spp., Gentiana spp. and Clitoria spp.

90. An isolated nucleic acid molecule which has been modified so as to comprise a sequence of nucleotides encoding or complementary to a sequence encoding F3'5'H or a polypeptide having F3'5'H activity wherein expression of said nucleic acid molecule in a rose petal tissue results in a sufficient level and length of transcript which is translated to said F3'5'H as determined by detectable levels of delphinidin or delphinidin based molecules as measured by a chromatographic technique wherein the nucleic acid molecule is derived from a plant selected from the list comprising petunia, gentiana and butterfly pea.

91. . The use of a nucleic acid sequence as defined in claim 89 or 90 in the manufacture of a genetic construct capable of expressing F3'5'H or down-regulating an indigenous F3'5'H enzyme in a plant.

92. The use of a nucleic acid sequence as defined in claim 89 or 90 in the manufacture of a genetically modified plant or part thereof or cells therefrom.