Flower Color

Flower Color Modification by Engineeringof the Flavonoid Biosynthetic Pathway: Practical Perspectives

Yoshikazu TANAKA,1;y Filippa BRUGLIERA,2 Gianna KALC,2 Mick SENIOR,2 Barry DYSON,2

Noriko NAKAMURA,1 Yukihisa KATSUMOTO,1 and Steve CHANDLER2

1Institute for Plant Science, Suntory Holdings Ltd., 1-1-1 Wakayamadai, Shimamoto, Osaka 618-8503, Japan2Florigene Pty Ltd., 1 Park Drive, Bundoora, Victoria 3083, Australia

Online Publication, September 7, 2010

[doi:10.1271/bbb.100358]

The status quo of flavonoid biosynthesis as it relatesto flower color is reviewed together with a success inmodifying flower color by genetic engineering. Flavo-noids and their colored class compounds, anthocyanins,are major contributors to flower color. Many plantspecies synthesize limited kinds of flavonoids, and thusexhibit a limited range of flower color. Since genesregulating flavonoid biosynthesis are available, it ispossible to alter flower color by overexpressing heter-ologous genes and/or down regulating endogenousgenes. Transgenic carnations and a transgenic rose thataccumulate delphinidin as a result of expressing aflavonoid 30,50-hydroxylase gene and have novel bluehued flowers have been commercialized. TransgenicNierembergia accumulating pelargonidin, with novelpink flowers, has also been developed. Although it ispossible to generate white, yellow, and pink-floweredtorenia plants from blue cultivars by genetic engineer-ing, field trial observations indicate difficulty in obtain-ing stable phenotypes.

Key words: anthocyanin; flavonoid; flower; geneticengineering; transgenic plants

Most currently cultivated floricultural crops have beendeveloped by a combination of extensive hybridizationand mutation breeding, but hybridization breedingsuffers from genetic constraints within a species.Genetic engineering liberates plant breeders from suchconstraints. Genetically modified soybean, corn, canola,and cotton varieties have gained considerable marketshare since the first commercial production in 1996, andwere cultivated on 134 million ha across the world in2009 (http://www.isaaa.org).

Successful genetic engineering of plants requiredseveral technological breakthroughs, including isolationof useful genes, the establishment of an efficienttransformation system for a target species, and regu-lation of transgene expression. We have been workingon the genetic engineering of floricultural crops focusingto produce novel flower colors. We have isolated many

genes encoding enzymes that act on compounds thatcontribute to flower color, have established transforma-tion systems for several floricultural crops, and havemodified the biosynthetic pathway of flower pigments.Since this field has been reviewed several times,1–5) wefocus here on recent progress in flower color andflavonoid biosynthesis relevant to flower color modifi-cation, describing major achievements in this field,including previously unpublished results.Flower and fruit color are important in the attraction

of pollinators and animals for seed dispersal, and thusare critical factors in plant survival. Flower and fruitcolor is derived mainly from flavonoids, carotenoids,and betalains. Flavonoids and their colored class,anthocyanins, contribute to a wide range of colors; paleyellow, orange, red, magenta, violet, blue. Carotenoidsare ubiquitously distributed in plants as essentialcomponents of photosynthesis and confer a yellow orred color on flowers when they are present. Betalainsalso result in yellow and red color, but only the familiesof Caryophyllales (except for Caryophyllaceae andMolluginaceae) produce betalains. To date, no plantsproducing both anthocyanins and betalains have beendiscovered. This review focuses on the flavonoids thathave been studied and engineered most extensivelyamong these three classes of pigments.Flavonoids are a class of phenylpropanoids that have

a C3-C6-C3 structure. They are classified into more than10 groups, mainly depending on the structure of theC-ring (Fig. 1). The flavonoids relevant to flower colorare usually localized in the vacuole, mostly in glycosy-lated and acylated forms. Anthocyanidins are theaglycone form and are chromophore of anthocyanins,and are also direct precursors of anthocyanins in theirbiosynthesis (Fig. 1). In spite of the structural diversityof anthocyanins, there are basically three major antho-cyanidins: pelargonidin, cyanidin, and delphinidin. Anincrease in the hydroxyl group number confers a colorshift to blue, and thus anthocyanins derived fromdelphinidin, cyanidin, and pelargonidin tend to yieldviolet/blue, red/magenta, and orange/intense red flower

This review was written in response to the corresponding author’s receipt of The Japan Prize of Agricultural Science in 2009.y To whom correspondence should be addressed. Tel: +81-75-962-8807; Fax: +81-75-962-3791; E-mail:Yoshikazu [email protected]: THC, 20,40,60,4-tetrahydroxychalcone; ODG, 2-oxoglutarate dependent dioxygenase; DHK, dihydrokaempferol; FNS, flavone

synthase; P450, cytochrome P450; F2H, flavanone 2-hydroxylase; F30H, flavonoid 30-hydroxylase; F3050H, flavonoid 30,50-hydroxylase; FLS, flavonolsynthase; DFR, dihydroflavonol 4-reductase; DHM, dihydromyricetin; ANS, anthocyanidin synthase; F3GT, UDP-glucose: flavonoid3-glucosyltransferae; AT, acyltransferase; GST, glutathione S-transferase; MRP, multidrug resistance-associated protein; ABC, ATP bindingcassette; MATE, multidrug and toxic extrusion; bHLH, basic helix loop helix; LAR, leucoanthocyanidin reductase; ANR, anthocyanidin reductase

Biosci. Biotechnol. Biochem., 74 (9), 1760–1769, 2010

Award Review

http://dx.doi.org/10.1271/bbb.100358

color, respectively.6) Classically bred roses, carnations,chrysanthemums, and lilies lack delphinidin-basedanthocyanins, and this is the primary reason there areno blue/violet varieties in these species.

Plants adopt many sophisticated mechanisms toexhibit and stabilize the flower color they need,especially blue colors. The mechanism of blue flowercolur has attracted the attention of chemists andbiologists for many years, and is discussed in a recentexcellent review.6) The glycosylation, acylation, andmethylation of anthocyanidins are the primary sourcesof structural diversity. The glycosylation and methyl-ation of anthocyanins causes the color to becomeslightly redder. Aromatic acylation of anthocyaninsshifts their absorption maxima toward a longer wave-length (bluer). Intermolecular stacking of anthocyaninsmodified with plural aromatic acyl groups, polyacylatedanthocyanins,7) such as in gentian, butterfly pea, andcineraria, results in a stable blue color. Stacking withco-pigments, typically flavones and flavonols, results ina bathochromic shift to a deeper and bluer color.Complexation with metal ions also results in a blue

color. Anthocyanin color also depends on pH to a largeextent. Acidic and neutral vacuolar pH gives redder andbluer color respectively. Recent progress is describedbelow.

I. Anthocyanins and Their Biosynthesis: AnUpdate of Recent Progress

1. Recent progress in understanding flower color:metal ionsAlthough tulip petals accumulate delphinidin-based

anthocyanins, there are no completely blue varieties.Some purple cultivars of tulip have blue coloration at thebottom of the perianth, and the blue and the purple partsof the flower contain the same anthocyanin (delphinidin3-rutinoside) and flavonols, and have a similar vacuolarpH, but the concentration of iron is 25 times higher inthe blue perianth parts than in the purple parts. In-vitroreconstruction with delphinidin 3-rutinoside, flavonols,and Fe3þ reproduce the blue color.8) An iron transporterlocated in the vacuolar membrane (TgVit1) has beenfound to be responsible for the blue coloration. Transient

Fig. 1. Schematic of the General Flavonoid Biosynthetic Pathway Relevant to Flower Color.Anthocyanidin is further modified with glycosyl, acyl, or methyl groups catalyzed by glycosyltransferase (GT), acyltransferase (AT), and

methyltransferase (MT). Classification of flavonoids is shown in parentheses. Abbreviations: CHS, chalcone synthase; F3H, flavanone3-hydroxylase; F30H, flavonoid 30-hydroxylase; F3050H, flavonoid 30,50-hydroxylase; DFR, dihydroflavonol 4-reductase; ANS, anthocyanidinsynthase; THC40GT, tetrahydroxychalcone 40-glucosyltransferase; AS, aurusidin synthase; FLS, flavonol synthasae; FNS, flavone synthase;LAR, leucoanthocyanidin reductase; ANR, anthocyanidin reductase.

Flower Color Modification 1761

expression of the gene in a purple tulip petal resulted ina small number of cells turning blue.9) Expression ofTgVit1 and suppression of the ferritin (an iron storageprotein) gene, TgFer1, was necessary to reproduce theblue color of the bottom of the perianth. TgVit1specifically expresses in the epidermal cells of theperianth bottom, and TgFer1 is suppressed at thebottom.10)

Hydrangea macrophylla changes in sepal colordepending on the availability of Al3þ, whose impor-tance to color has been clarified by precise chemicalanalysis. The extract of the colored protoplasts derivedfrom blue and red sepals were subjected to identify thecompounds they contained. Blue cells contain 13equivalents of 5-acylquinic acids (co-pigments) and1.2 equivalents of Al3þ to anthocyanin (delphinidin3-glucoside, 12.9mM), while red cells contain only 3.6equivalents of 5-acylquinic acids and a 0.03 equivalentof Al3þ. In vitro reconstruction using 10mM delphinidin3-glucoside with 22.5 equivalents of 5-acylquinic acidsand 1 equivalent of Al3þ at pH 4.0 reproduced the colorof the blue cells, and mixing of 10mM delphinidin3-glucoside with 21.5 equivalents of 5-acylquinic acidsand a 0.03 equivalent of Al3þ at pH 3.0 reproducedthe color of the red cells.11) Although Fe3þ and Al3þ

resulted in clear bluing effects on anthocyanin color andknowledge of metal iron transfer from the roots to theaerial parts of plants is rapidly accumulating, especiallyfor iron,12) utilization of these ions to generate blueflowers in transgenic plants remains challenging sincean excess of heavy metal ions is generally detrimentalto plants.

2. Recent progress in the biosynthesis of flavonoidsrelevant to flower color

A major pathway relevant to flower color is shown inFig. 1.3,13) Chalcone synthase is the first committedenzyme in flavonoid biosynthesis. 20,40,60,4-tetrahydroxy-chalone (THC) is stereospecifically isomerized to (2S)-naringenin (a flavanone) by the catalysis of chalconeisomerase. Flavanone 3�-hydroxylase, a 2-oxoglutaratedependent dioxygenase (ODG), catalyzes 3-hydroxyla-tion of flavanones to the corresponding dihydroflavo-nols, e.g., naringenin to dihydrokaempferol (DHK).

Flavanones are direct precursors of the correspondingflavones. Flavones act as co-pigments, and contribute toa bathochromic shift (bluing and intensifying of colors)of anthocyanins. Flavone synthesis is catalyzed byflavone synthase (FNS). There are two kinds of FNS:an ODG type (FNSI) found in parsley14) and rice,15) anda more ubiquitous cytochrome P450 (P450) type. Inlegumes, flavones are biosynthesized from flavanones intwo steps, by the reaction of flavanone 2-hydroxylase(F2H) and then dehydration. In other plants (torenia,snapdragon, gentian, perilla, gerbera, etc.) flavones aredirectly biosynthesized by the action of flavone synthaseII (FNSII) on flavanones.16) F2H and FNS belong to thesame subfamily (CYP93B) of the P450 superfamily.Recently a P450 that is more closely related to FNSIIthan F2H was identified in soybean, suggesting thatlegume plants have both FNSII and F2H.17)

Flavone C-glucoside (e.g., isovitexin) has a strongco-pigment effect, as in Japanese iris18) and a transgeniccarnation.19) Recent work on rice has revealed that

2-hydroxyflavanone is glucosylated by the catalysis aUDP-glucose dependent C-glucosyltransferase, and thendehydrated to flavone C-glucoside.20) It is yet to bedetermined whether this pathway is the same in plantspecies that accumulate a flavone C-glucoside.Flavonoid 30-hydroxylase (F30H) and flavonoid 30,50-

hydroxylase (F3050H) catalyze the hydroxylation offlavanones, dihydroflavonols, flavones, and flavonols.They determine the hydroxylation pattern of the B-ringof anthocyanins, and are thus critical enzymes in flowercolor engineering.16) They belong to the same family,CYP75, of the P450 superfamily. In the chrysanthemumfamily, the F3050H gene evolved from the F30H geneafter the speciation of the chrysanthemum21) and a fewamino acid substitutions of F30H resulted in F3050Hactivity.22)

Dihydroflavonols are also at a branching point offlavonoid biosynthesis, and are the direct precursors offlavonols. Flavonols are colorless or pale yellow tohuman eyes by themselves, and contribute to bluing asco-pigments. Flavonol biosynthesis is catalyzed byflavonol synthase (FLS), an ODG. Dihydroflavonolsare reduced to leucoanthocyanidins by the action ofdihydroflavonol 4-reductase (DFR). Competition be-tween FLS and DFR affects flower color and theamounts of flavonols and anthocyanins are oftenreciprocal in transgenic plants.23,24) Native plants canavoid such competition, since FLS expression (flavonolbiosynthesis) usually precedes DFR expression (antho-cyanin biosynthesis) in petal development.The DFRs of many plant species recognize DHK,

dihydroquercetin, and dihydromyricetin (DHM). Eventhe DFR of rose, carnation, and chrysanthemum cancatalyze DHM, although they do not normally produceDHM due to a lack to F3050H. On the other hand,petunia25) and cymbidium26) DFR cannot utilize DHK asa substrate. Petunia DFR catalyzes DHM efficiently, andis thus a suitable molecular tool to enhance the pathwayleading to delphinidin, as described below.Leucoanthocyanidins are converted to the correspond-

ing anthocyanidins by the catalysis of anthocyanidinsynthase (ANS, often called leucoanthocyanidin dioxy-genase). FLS and ANS are multifunctional enzymes(FLS has ANS activity and ANS has FLS activities) inArabidopsis, as recently reviewed.27) It is still an openquestion whether such multifunctionality is applicable toother plant species and how significant it is for flowercolor.Anthocyanidins are unstable at neutral pH and must

be stabilized by glycosylation and acylation. The firstglucosylation reaction is usually 3-glucosylation cata-lyzed by UDP-glucose:flavonoid glucosyltransferase(F3GT) resulting in the production of colored antho-cyanins. Species specific glycosylation and acylation ofanthocyanins can then occur to result in diverseanthocyanins. It was thought that anthocyanin biosyn-thesis proceeds in the cytosol, but recently someacylation reactions have been found most likely toproceed in the vacuole, catalyzed by acyl-glucosedependent serine carboxypeptidase like enzymes, asrecently reviewed.1) Another anthocyanin acyltransfer-ase (AT), one better characterized, is acyl CoA depend-ent BAHD type anthocyanin AT.28)

1762 Y. TANAKA et al.

3. Transport of anthocyanins to the vacuole andregulation of vacuolar pH

An excellent review29) is available on this topic, andwe summarize this subject only briefly here. Floralflavonoids accumulate in the vacuole, while biosynthesisof them occurs mainly in the cytosol. The transportmechanism to the vacuole is not as clear as thebiosynthesis. A better understanding of the sequestrationof anthocyanins and flavonoid glycosides to the vacuolemay be important in the further alteration of flowercolor, since blue flowers tend to accumulate largeranthocyanin molecules than red flowers, and it may bethat the transport system in red-flower species is notcompatible with the larger anthocyanin molecules.

Two routes to vacuolar storage have been described.One is membrane transporter-mediated, and the other isvesicle-mediated. These are not necessarily exclusive,and are dependent on plant species, organ type, and stageof plant development. The membrane transporter-mediated route in maize has been found to consist of aglutathione S-transferase (GST, BZ2) and a multidrugresistance-associated protein (MRP)-type ATP bindingcassette (ABC) transporter (MRP3). Various resultsindicate that GST is involved in anthocyanin transportin other plants. The involvement of MRP has beenconfirmed only for maize.30) Proton-gradient dependentmultidrug and toxic extrusion (MATE) transporters havebeen found to facilitate epicatechin 30-O-glucoside inMedicago truncatula and Arabidopsis.31) Grapevine(Vitis vinifera) MATE transporters (AM1 and AM3)carry acylated anthocyanins (3-p-coumaroylglucosylatedanthocyanins) but not malvidin 3-glucoside, the predom-inant anthocyanin in grape skin.32) Malvidin 3-glucosidemight be carried into vacuoles by another mechanism.

Anthocyanin color depends on vacuolar pH. Lower pHconfers red color and neutral pH, blue. Intense red rosepetals tend to have lower vacuolar pH (about pH 4), andhence the ability to manipulate vacuolar pH is importantin engineering flower color. Pelargonium and impatienspetals contain delphinidin, but these species lack violet/blue varieties, probably due to their lower vacuolar pH.Regulation of vacuolar pH in petals has been studied inmorning glory and petunia. They regulate vacuolar pH inthe reverse direction. The vacuolar pH in wild-typemorning glory flowers increases from pH 6.6 to 7.7before flower opening through the expression of sodiumproton antiporter homolog33,34) recently shown to be apotassium proton antiporter.35) Wild-type petunias acid-ify the petal vacuole by expressing a P3A-H

þ-ATPaselocalized on the vacuolar membrane,36) due to whichflower color changes from violet to magenta.

A recent example of vacuolar pH affecting flowercolor was reported for Veronica petals. The petalscontain delphinidin 3-di-p-coumaroylsophoroside-5-ma-lonylglucoside and 7-glucuronyl or 7-glucuronyl glucur-onyl apigenin. Purple, violet, and blue cells have beenobserved in the petals and in in vitro reconstitutionexperiments, indicating that the vacuolar pH levels ofthese cells are pH 5, 6, and 7, respectively.37)

4. Transcriptional control of anthocyanin biosynthesisin flowers and fruits

A complex consisting of two transcriptional factors,basic helix loop helix (bHLH) and R2R3-MYB, and

WD40 (MBW complex) activates the transcription offlavonoid biosynthetic genes.38,39) In the case of petunia,they are ANTHOCYANIN1 (AN1), AN2, and AN11respectively. bHLH and R2R3-MYB bind specific se-quence motifs in regulatory regions of the transcriptionof flavonoid biosynthetic genes. The WD40 proteinfacilitates protein-protein interaction. bHLH factorsand WD40 proteins are ubiquitously expressed, whilespecific MYB proteins regulate specific biosyntheticpathways. This process of regulation has been studied inmodel species such as petunia, Arabidopsis, and snap-dragon.38) The basic regulatory mechanisms of antho-cyanin biosynthesis are well conserved and are regulateddifferently in distinct species.40) The biosynthesis ofanthocyanins, flavones/flavonols, and proanthocyani-dins is regulated differently depending on the species.One plant species has more than 100 R2R3-MYB genesand forms clusters consisting of orthologs from variousplants. R2R3-MYB, regulating anthocyanins and proan-thocyanidins, forms distinct but related clusters.GtMYB3 and GtbHLH1, which belong to the same

clusters as petunia AN2 and AN1 respectively, have beenisolated from gentian petals. Their expression profilesmatched the accumulation of gentian anthocyanin. TheGtMYB3 and GtbHLH1 proteins have been found tointeract in a yeast two-hybrid system. Co-expression ofthese genes up-regulated the transcription of gentiananthocyanin biosynthetic genes. A white cultivar wasshown to have a transposon inserted in the GtMYB3gene.41) Two R2R3-Myb genes, LhMyb6 and LhMyb12(homolog of petunia AN2), were isolated from an Asiatichybrid lily. Their proteins interacted with a lily bHLH(LhbHLH2), and transient coexpression of LhMyb6 andLhbHLH2 or LhMyb12 and LhbHLH2 activated thetranscription of anthocyanin biosynthetic genes in lilybulb scales. LhMyb12 positively regulates anthocyaninbiosynthesis in the tepals, filaments, and styles, andLhMyb6 in the anthocyanin spots in the tepals and light-induced pigmentation in leaves.42)

Genes of R2R3-MYB10 belonging to the same clusteras petunia AN2 have been isolated from many membersof the rosaceous family (apple, pear, plum, cherry, peach,raspberry, rose, and strawberry). The expression of thesegenes correlates with fruit and flower anthocyanin levels.Constitutive expression of strawberry MYB10 in thestrawberry resulted in elevated anthocyanin levels in theleaves, fruits, red roots, and stigmas.43) Rearrangement inthe upstream regulatory region of appleMYB10 activatedanthocyanin production, producing red foliage and fruitflesh. This rearrangement generated an autoregulatorylocus, and this autoregulation was sufficient to increaseMYB10 transcript levels and the accumulate anthocya-nins throughout the plant ectopically.44)

Gerbera hybrida R2R3-MYB10 is highly homologousto petunia AN2. Transgenic gerbera expressing the geneconstitutively showed enhanced accumulation of antho-cyanin pigments and an altered pigmentation pattern.Microarray analysis of transgenic gerbera indicated thatit activated the transcription of a GST involved inanthocyanin transport and a serine carboxypeptidase-like gene, in addition to flavonoid biosynthetic genes.45)

Ectopic expression of Arabidopsis PAP1 (AtMYB75), anR2R3-MYB gene, activates anthocyanin biosynthesis intobacco, but not in Medicago truncatula or alfalfa. On


the other hand, the expression of Legume AnthocyaninProduction 1 (LAP1), an R2R3-MYB gene, fromM. truncatula, induced an accumulation of anthocyaninsand proanthocyanidins by elevating transcripts of twoflavonoid and anthocyanin biosynthetic genes, a MATE-like antiporter and ABC transporter.46)

Overexpression of Arabidopsis PAP2 (AtMYB90)also activates anthocyanin biosynthesis in tobacco.AtMYB90, lacking a C-terminal 78 amino acid se-quence, converted the R2R3-MYB gene that activatesplant-wide anthocyanin biosynthesis to a dominant-negative allele that interfered with normal tobaccopigment production.47) Overexpression of the bHLH orR2R3-MYB gene resulted in ectopic expression ofanthocyanins and often darker flowers, as reviewed.2)

Expression of a bHLH (snapdragon Delila) and a R2R3-MYB (snapdragon Rosea 1) under a fruit-specificpromoter in tomato yielded purple fruits accumulatinganthocyanins at a concentration comparable to theanthocyanin levels found in blackberries and blue-berries.48) Such tomatoes might be beneficial to health.

Grapevine R2R3 MYB genes have been analyzedcomprehensively.49) Grape skin color is determined bythe genotype of a Myb gene, R2R3 Myb (VvmybA1c),which regulates anthocyanin biosynthesis by control-ling the transcription of the F3GT gene. Insertion of aretrotransposon into the promoter region of VvmybA1cresults in VvmybA1a and acyanic skin color.50) Analy-sis of the grapevine genome has revealed that itcontains more than 100 R2R3 MYB genes. They areclassified in terms of their structures and similarity totheir Arabidopsis orthologs.49) Anthocyanin relatedMYBA clusters are located on chromosomes 2 and14.49) Another R2R3-MYB protein, VvMYBPA1, regu-lates proanthocyanidin biosynthesis in berry skin andseeds by controlling the transcription of leucoantho-cyanidin reductase (LAR) and anthocyanidin reductase(ANR), both of which are critical enzymes in proan-thocyanidin biosynthesis (Fig. 1). VvMYBPA1 does notactivate the F3GT gene.51) Proanthocyanidin biosyn-thesis in persimmon fruit is also regulated by an R2R3-MYB, DkMyb4, that belongs to the same cluster asVvMYBPA1. Ectopic expression of DkMyb4 in kiwifruitinduces proanthocyanidin biosynthesis but not antho-cyanin biosynthesis.52)

VvMYB5a and VvMYB5b activate the transcription offlavonoid biosynthetic enzyme genes (F3050H, ANS,LAR) but not the F3GT or the ANR genes.53,54) Theyare more closely related to petunia PH455) than AN2.PH4 encodes an R2R3MYB that interacts with AN1 (abLHL). The AN1/PH4 complex activates genes such asPH5, which encodes a P3A-H

þ-ATPase36) required forvacuolar acidification via PH3, probably a transcriptionalfactor.40,55) Overexpression of VvMYB5b in tobaccoresults in the accumulation of anthocyanin- and proan-thocyanidin-derived compounds.54) VvMYB12 mightregulate flavonol biosynthesis by controlling the expres-sion of the FLS gene,56) and its expression and flavonolbiosynthesis is suppressed by post-veraison sun expo-sure. Flavonol synthesis is regulated differently fromanthocyanin biosynthesis in Arabidopsis by AtMYB12.57)

In addition to the R2R3-Myb and bHLH transcrip-tional factors, other transcriptional factors also regulateflavonoid and anthocyanin biosynthesis. A small MYB

protein, R3-MYB (MYBL2), suppresses flavonoid bio-synthesis by interacting with a MBW complex. The lossof MYBL2 resulted in a dramatic increase in anthocya-nins in Arabidopsis seedlings.58,59) Overexpression ofMYBL2 inhibited the biosynthesis of proanthocyanidinsin seeds.58) Plant-specific NAC transcriptional factor hasbeen found to regulate anthocyanin biosynthesis. AnArabidopsis ANAC078 induced anthocyanin accumula-tion by activating anthocyanin biosynthetic genes viatranscriptional factors regulating biosynthesis understrong-light conditions.60)

5. Unsolved problems in flavonoid biosynthesis andmetabolismThe flavonoid biosynthetic pathway is probably the

most thoroughly studied plant secondary metabolismpathway. Except for the glucosyltransferases and ATsthat are necessary for polyacyl anthocyanins, most of theenzymatic genes have been characterized. Isolation ofthe genes of these enzymes should provide moremolecular tools to engineer anthocyanin structure andflower color. It is not clear whether these modificationsoccur in the cytosol or in the vacuole or on membranes.The enzymes involved in some plant metabolic

pathways, including flavonoid biosynthesis, perhapsform a macromolecular complex (metabolon) for meta-bolic channeling and efficient biosynthesis.61) Metabo-lons are thought to be anchored onto the ER membraneutilizing cytochromes P450 such as F30H.61) Moreexperimental data are necessary to confirm this. If ametabolon is necessary for efficient synthesis of antho-cyanins, an enzyme from an exogenous gene must becompatible with the endogenous metabolon.The petals of some plants fade or lose color during

development. Anthocyanin degradation and disappear-ance have been reviewed,62) and also is important interms of engineering flower color. Further understandingof these problems should contribute to engineering ofthe pathway.

II. ColorModification by Genetic Engineering

1. Color modification towards blueCarnationsBlue flowers exhibit color through a combination of

the various factors mentioned above. Although manygenes regulating these factors have been isolated, onlydelphinidin production is a proven genetic modificationtactic to alter flower color towards blue. Here we refer tocarnations and roses, species in which this tactic hasbeen used and new varieties have been commercialized.Naturally, both species accumulate pelargonidin andcyanidin-based anthocyanins, but they do not producedelphinidin-based anthocyanins due to a lack of F3050H.In order to achieve significant color change, it isnecessary to elevate the delphinidin content to morethan 90%, and ideally close to 100%. This cannot beachieved by a simple overexpression of a F3050H gene.Avoiding competition with endogenous enzymes (F30H,DFR, FLS) or downregulation of endogenous enzymesis necessary to obtain a high content of delphinidin.Since flower color depends on vacuolar pH and co-pigment, choosing a host with a proper genetic back-ground is also important.


Almost exclusive accumulation of delphinidin wasachieved in carnations, by expressing a F3050H gene anda petunia DFR gene in white carnation cultivars thatwere specifically deficient in the DFR gene (Fig. 2A).The final color hue depends on the genetic backgroundsof the host cultivars. Since the activity of transgenesvaries between transgenic lines, the color and theamount of anthocyanins also vary. It is important toestablish efficient transformation systems and to gen-erate a large number of transgenic lines in order toobtain plants with the desirable phenotype. Twelve, six,and three lines are now sold in the USA, Japan, and EU,assuming permits for commercialization. The permit isnecessary for every country or area, and regulatoryprocedures depend on the country. Commercial andregulatory issues relating to carnations have beendiscussed in the literature.1,4)

RoseThe cultivated rose (Rosa hybrida) was bred from

about eight wild species, including R. chinensis,R. galica, R. gigantia, R. moschata, R. multiflora,R. wichuraiana, and R. foetida after extensive inter-specific hybridization (http://www.gifu-u.ac.jp/�fukui/02-1-2-2.htm). Pink and red colors are derived frompelargonidin or cyanidin-based anthocyanins. The ex-pression of a pansy (Viola spp) F3050H gene resulted insignificant amounts of delphinidin-derived anthocya-nins accumulating in the petals of the transgenic plants.Expression of the pansy F3050H gene in rose cultivarsthat have higher vacuolar pH, relatively large amountsof flavonols (co-pigments), and weak or no F30Hactivity resulted in transgenic lines in which 95% ofthe anthocyanidins was delphinidin. The colors of theflowers in these lines were of a significantly bluer hue

A B

C

E

D

Fig. 2. Color Modified Transgenic Flowers.A, Transgenic carnations, Moon Series, that accumulate anthocyanins derived from delphinidin and their production farm in Colombia.

B, Transgenic roses, Suntory Blue Rose APPLAUSE. C, Schematic of the T-DNA designed to deviate the anthocyanin pathway fromdelphinidin to pelargonidin in Nierembergia. The host flower is shown on the left and a transgenic flower right. P1, Mac1 promoter;78) P2,enhanced cauliflower mosaic virus 35S promoter;79) T, mannopine synthase terminator, NF3050H contains the inverted repeat of theNierembergia F3050H sequence, and NFLS contains an inverted repeat of the Nierembergia FLS sequence. D, Outdoor field trial of colormodified transgenic torenia plants in Australia. Half-shade trial is shown above and full-sun trial below. Strong growth retardation was observedfor SWB/805-27. These photos were taken December 2007. E, Comparison of flowers of torenia plants grown indoors (first row of each photo)and outdoors (second row of each photo). Some lines showed unstable phenotypes, especially outside. These photos were taken in March 2008.See the text for details.


than any conventionally bred cultivar.63) The toreniaanthocyanin 5-AT or perilla anthocyanin 3-AT genewas co-expressed with the pansy F3050H but this didnot significantly enhance bluing. Two lines, WKS82-130-4-1 and WKS82-130-9-1, were granted commer-cial release in Japan after proof that their release wouldbe most unlikely to affect biodiversity in Japan (https://ch.biodic.go.jp/bch/OpenSearch.do) under the ‘‘LawConcerning the Conservation and Sustainable Use ofBiological Diversity through Regulations on the Useof Living Modified Organisms’’ (Cartagena ProtocolDomestic Law). One line, WKS82-130-4-1, was com-mercially propagated and launched in 2009 (Fig. 2B).This was the first and currently is the only commercialproduction of a genetically modified plant in Japan.

In order to achieve exclusive delphinidin accumula-tion irrespective of the endogenous pathway leading topelargonidin or to cyanidin, a genetic construct designedto downregulate the endogenous DFR and overexpressboth a viola F3050H and a Dutch iris DFR genes wasintroduced into rose cultivars. The petals of the resultingplants accumulated delphinidin-based anthocyanins ex-clusively.When a selected line (LA919-4-10) was crossedwith a deep red rose (a cyanidin producer), progenyharboring the transgenes produced delphinidin exclu-sively. The amount of delphinidin varied between lines.The flower color was generally magenta, presumablybecause of a low vacuolar pH.63) However, this modifi-cation of the pathway also caused some growth retarda-tion and hence is not suitable for commercialization.

As previously mentioned, blue flower color is usuallythe additive result of various factors, such as anthocya-nin modification, co-pigment concentration, vacuolarpH, and metal ion type and concentration.6) Targeting ofthese factors in addition to delphinidin production intransgenic carnation and rose is expected to yield bluerflower colors.

2. Color modification towards redTo cause an accumulation of cyanidin or pelargonidin

is a proven way to generate pink or red flowers.Downregulation of F3050H in petunia, torenia, osteo-spermum, and gentian (as reviewed by Tanaka et al.1))and also cyclamen64) has been shown to redirect thedelphinidin pathway to cyanidin.

Some species lack red or orange colored flowers dueto a deficiency of pelargonidin. Efforts to producepelargonidin started with the petunia.65) The petuniadoes not accumulate pelargonidin since its DFR does notcatalyze DHK. Expression of the maize DFR gene andthose of a few other species in a petunia mutant,deficient in FLS, F30H and F3050H, resulted in trans-genics with orange flowers, as a result of an accumu-lation of pelargonidin. It is interesting that the expres-sion of a gerbera DFR gene resulted in the production ofmore pelargonidin than in the maize DFR gene. Geneexpression appeared to be more stable.66) The choice ofgene source is important even when enzyme itself hasthe same activity. Similarly, campanula F3050H wasshown to be more efficient than petunia or lisianthusF3050H in tobacco,67) and butterfly F3050H more efficientthan verbena F3050H in verbena.68)

Red tobacco accumulating pelargonidin was producedby downregulating the F30H and FLS genes and over-

expressing a gerbera DFR gene.69) Downregulation ofthe F3050H gene and the gene encoding anthocyanin5,30-AT in gentian changed the flower color from blue topink, and was associated with an accumulation of non-acylated cyanidin glucosides.70)

Nierembergia, a popular bedding Solanaceae plant,has only violet and white varieties. Downregulation ofthe F3050H gene and additional expression of a rose DFRgene in Nierembergia resulted in white flowers only.71)

This was probably due to its weak F30H activity andstrong FLS activity. Downregulation of both the F3050Hand the FLS gene and overexpression of the rose DFRgene (Fig. 2C) in blue-flowered Nierembergia cultivarNB18 (in which 99.7% of total anthocyanins aredelphinidin-based) resulted in transgenic plants withpink flowers accumulating pelargonidin (Fig. 2C). In thepetals of the transgenic line with the strongest colorchange 74% of total anthocyanins were pelargonidin-based and 26% were cyanidin-based. The amount offlavonols decreased from 2.05mg/g fresh petal in NB18to 0.58mg/g fresh petal in the flowers of the transgenicline, but the total amount of anthocyanidin did notincrease. Enhancement of the flux to pelargonidin-basedanthocyanins and more complete suppression of FLSgene might lead to transgenic lines with redder flowers.

3. Outdoor trial of color-modified transgenic toreniaTransgenic carnations have been produced under

greenhouse conditions for more than 15 years, and thecolor-modified phenotypes of the selected transgeniclines have been stable throughout this continuous periodof vegetative propagation. In flower color-altered trans-genic petunia harboring the maize DFR gene under thecontrol of cauliflower mosaic virus 35S promoter,65)

however, the phenotype was not stable and variegatedflowers were observed due to methylation of thepromoter when the plants were grown outdoors.72) It isnot clear whether the differences in phenotypic stabilityof the carnation and petunia are due to differences in thepromoter or coding sequences of the transgene or togrowing conditions. A related question is whetherdownregulation of a target gene by the transcription ofdouble-strand RNA (RNAi technology73)) is stable ordeleterious in transgenic plants grown under outdoorconditions (which are more extreme). This prompted usto carry out field trial of color-modified transgenictorenia by RNAi.Two cultivars of torenia (Torenia hybrida) we have

used for genetic modification are Summerwave Blue(SWB) and Summerwave Violet (SWV). In terms ofassessing the commercial potential of transgenic plants,two critical factors are the possible effects on biodiver-sity of the gene flow from the transgenic plants to otherrelated species and the possibility of the transgenic linesbecoming weeds. SWB and SWV are benign in thisrespect as they are both male and female sterile.The petals of both SWB and SWV mainly accumulate

anthocyanins derived from delphinidin. Cultivar SWVhas darker-colored flowers, but both contain a largerconcentration of flavones than anthocyanins in thepetals. In the course of a program to modify flowercolor in these two varieties, we generated about 100independent transgenic plants of each cultivar. The lineswith the most stable flower-color phenotype over several


years of maintenance in a contained greenhouse wereselected for outdoor trials. The lines selected were asfollows: (i) SWB308-25, SWB 308-35, and SWB-308-36. In these transgenic lines, the petals accumulateaurone and are yellow due to expression of thesnapdragon THC 40-glucosyltransferase and aureusidinsynthase genes and suppression of the endogenous F3Hgene to downregulate anthocyanin biosynthesis,74) (ii)SWB805-27 and SWB 805-36. These transgenic lineshave white or partially white (a white cross in bluepetals) flowers respectively. This is due to the knock-down of expression of the endogenous ANS gene inSWB,75) (iii) SWB1322-45 and SWB-1322-90. Thesehave pink flowers due to an accumulation of cyanidin-based anthocyanins as a result of knockdown of F30Hand F3050H genes,76) (iv) SWV1341-17 and SWV 1341-85. The flowers of these two lines accumulate pelargo-nidin and have strongly pink-colored petals as a result ofknockdown of the F30H and F3050H genes, comple-mented by the expression of a pelargonium DFR gene.76)

All lines selected were grown in replicated outdoor trialsin Melbourne, Australia, under a license (DIR068/2006,http://www.ogtr.gov.au/ir/dir068.htm) issued by theOffice of the Gene Technology Regulator. Permissionfor this trial, containing multiple lines and constructs,was more readily granted than would have been possiblein Japan.

All nine transgenic torenia lines, and two unmodifiedcontrols, SWB and SWV, were grown outdoors inhanging baskets for two spring-summer-autumn seasons:October 2007-July 2008 (trial 1), and October 2008-May2009 (trial 2). In each trial, plants were hung in twopositions, one protected from the wind in partial sun andthe other in a fully exposed, the full-sun position.

Trial 1By November in each trial, the plants were fully

grown in the shaded area but flowered less vigorouslythan those in the full-sun area. The results are summa-rized in Table 1. Several transgenic lines did not grow aswell as the control or the other transgenic lines. Thesepoor lines were SWB308-36, SWB308-35, andSWB805-27. In full-sun, SWB805-27 stood out as beingextremely poor as the plants rarely flowered (Fig. 2D).The growth of line SWV1341-17 was as good as theuntransformed control, and line SWB1322-90 stood outas a very good line, with slightly more vigorous growththan the control. Reversion to parental flower color wasobserved in some plants in some lines. This was mostpronounced in line SWB308-25, where the parental blueflower color was observed, and in the lines SWB308-36and SWB308-35, where the petals also partly regained

the blue parental flower color (Fig. 2E). The transgenicphenotype was stable in lines SWB805-27, SWB1322-45, SWB1322-90, and SWV1341-17 (Fig. 2E). It wasfound that outdoor plants had smaller flowers than thesame lines grown in an adjacent greenhouse at the sametime (Fig. 2E).Trial 2The second trial was held from October 2008 to May

2009. In this trial the morphological characters of theflowers were measured, and it was confirmed that theoutdoor plants had smaller flowers than the same linesgrown in an adjacent greenhouse at the same time. Inthis trial the only lines SWB805-36, SWB1322-45,SWB1322-90, and SWV1341-17 had both a fully stablemodified flower color phenotype and growth character-istics as good as the host variety.In these experiments, the instability of the phenotype

was apparent only under outdoor conditions, and only insome lines. During the trials, the plants outside weresubjected to much higher temperatures and higher lightlevels than inside. The implication is that theseconditions interfered with the expression of the auronebiosynthesis genes transformed into the 308 series oftransgenic lines, or with the method used to suppress theendogenous F3H gene.As far as growth of plants outside is concerned, the

results indicated that artificial induction of RNAi doesnot necessarily have a negative effect on plant growthbut SWB308-35, SWB308-36, and SWB805-27 grewmore slowly the host and the other transgenic lines.Where the transgenic phenotype was lost in lineSWB308-25 and flowers reverted to blue during thetrial, plant growth was comparable to the host. Toreniapetals contain flavones and anthocyanins, but only theanthocyanin biosynthetic pathway was downregulated inthe transgenic plants exhibiting poor growth outdoors.These lines still produced flavonoids, mainly flavones,which are necessary for UV-protection. This implies thatanthocyanins are necessary for normal plant growthunder outdoor conditions, playing a role that flavonescannot play. Another possible explanation is that newcompounds having negative effects on plant growthaccumulate in transgenic plants in which aurone biosyn-thesis has been engineered.

III. Future Perspectives

Since the beginning of flower color alteration usinggenetic modification more than 20 years ago, consid-erable progress has been made. Though novel blue-violetflower color carnations and roses have been successfully

Table 1. Summary of Phenotypes of the Transgenic Torenia in Field Trial (trial 1)

Genetic constructs Line numbers Color stabilityGrowth compared

to the host

Over expression of snapdragon THC 40-GT and aureusidin synthase genes SWB/308-25 Unstable

and knock down of F3H gene in SWB74) SWB/308-35 Unstable Poor

SWB/308-36 Unstable Poor

Knock down of ANS gene in SWB75) SWB/805-27 Stable Poor

SWB/805-36 Unstable

Knock down of F30H and F3050H genes in SWB76) SWB/1322-45 Stable

SWB/1322-90 Stable

Over expression of pelargonium DFR gene and knock down of F30H SWV/1341-17 Stable

and F3050H genes in SWV76) SWV/1341-85 Unstable


marketed, much remains to be done to create morecolorful or even bluer flowers by genetic engineering.For example, among many useful genes for colormodification, F30H and F3050H have been the primaryfocus of modification so far. The genes relevant toflavone synthesis, polyacylation of anthocyanins, vacuo-lar pH regulation, and metal transport have not been fullyexploited. This is partly due to the difficulty of optimiz-ing the expression of multiple genes in target speciesand the provision of enough substrates for the heterol-ogous enzymes derived from transgenes. A repeatedtrial-and-error approach is inevitable to obtain transgeniclines exhibiting stable phenotypes, and it is thereforealso desirable to develop new technologies to suppressendogenous gene expression more efficiently than cur-rent RNAi. Irreversible genome modification by zinc-finger nuclease77) is a candidate for such technologies.

Although flavonoid biosynthesis is the most charac-terized plant secondary metabolism pathway, the proc-ess of flavonoid transport to the vacuoles is not yetunderstood well, and the proposed metabolon forflavonoid biosynthetic enzymes is still hypothetic inour view. Specificity for vacuolar transport or metabolonformation, if any, has to be considered in developingstrategies to produce novel flower color.

These scientific and technological difficulties may beresolved in due course, but the, costs of developmentand regulation remain high. Proper permits must beobtained in countries and areas for commercialization,and regulatory systems vary greatly depending on thecountry and area. Simple, global standards have to be setfor the application of genetic engineering to plantbreeding to flourish. This is especially true for floricul-tural crops, whose market is segmented and sales percultivar relatively small.

Acknowledgments

We wish to thank the current and former members ofSuntory and Florigene for their contributions, over manyyears. In order to focus on recent progress, only alimited number of original papers are cited here.

References

1) Tanaka Y, Brugliera F, and Chandler S, Int. J. Mol. Sci., 10,

5350–5369 (2009).

2) Tanaka Y and Ohmiya A, Curr. Opin. Biotechnol., 19, 190–197

(2008).

3) Tanaka Y, Sasaki N, and Ohmiya A, Plant J., 54, 733–749

(2008).

4) Chandler S and Tanaka Y, CRC Crit. Rev. Plant Sci., 26, 169–

197 (2007).

5) Davies KM and Schwinn KE, ‘‘Molecular Biology and

Biotechnology of Flavonoid Biosynthsis,’’ eds. Andersen OM

and Markham KR, CRC, Boca Raton, pp. 143–218 (2006).

6) Yoshida K, Mori M, and Kondo T, Natl. Prod. Rep., 26, 884–

915 (2009).

7) Honda T and Saito N, Heterocycles, 56, 633–692 (2002).

8) Shoji K, Miki N, Nakajima N, Monomoi K, Kato C, and

Yoshida K, Plant Cell Physiol., 48, 243–251 (2007).

9) Momonoi K, Yoshida K, Mano S, Takahashi H, Nakamori C,

Shoji K, Nitta A, and Nishimura M, Plant J., 59, 437–447

(2009).

10) Shoji K, Momonoi K, and Tsuji T, Plant Cell Physiol., 51, 215–

224 (2010).

11) Ito D, Shinkai Y, Kato Y, Kondo T, and Yoshida K, Biosci.

Biotechnol. Biochem., 73, 1054–1059 (2009).

12) Ishimaru Y, Masuda H, Bashir K, Inoue H, Tsukamoto T,

Takahashi M, Nakanishi H, Aoki N, Hirose T, Ohsugi R, and

Nishizawa NK, Plant J., 62, 379–390 (2010).

13) Grotewold E, Annu. Rev. Plant Biol., 57, 761–780 (2006).

14) Martens S, Forkmann G, Britsch L, Wellmann F, Matern U, and

Lukacin R, FEBS Lett., 544, 93–98 (2003).

15) Lee YJ, Kim JH, Kim BG, Lim Y, and Ahn J-H, BMB Rep., 41,

68–71 (2008).

16) Tanaka Y, Phyochem. Rev., 5, 283–291 (2006).

17) Fleigmann J, Furtwangler R, Malterer G, Cantarello C,

Schuler G, Ebel J, and Mithofer A, Phyochemistry, 71, 508–

514 (2010).

18) Yabuya T, Nakamura M, Iwashina T, Yamaguchi M, and

Takehara T, Euphytica, 98, 163–167 (1997).

19) Fukui Y, Tanaka Y, Kusumi T, Iwashita T, and Nomoto K,

Phytochemistry, 63, 15–23 (2003).

20) Brazier-Hicks M, Evans KE, Gershater MC, Puschmann H,

Steel PG, and Edwards R, J. Biol. Chem., 17926–17934 (2009).

21) Seitz C, Eder C, Deiml B, Kellner S, Martens S, and Forkmann

G, Plant Mol. Biol., 61, 365–381 (2006).

22) Seitz C, Ameres S, and Forkmann G, FEBS Lett., 581, 3429–

3434 (2007).

23) Davies KM, Schwinn KE, Deroles SC, Manson DG, Lewis DH,

Bloor SL, and Bradley JM, Euphytica, 131, 259–268 (2003).

24) Tsuda S, Fukui Y, Nakamura N, Katsumoto Y,

Yonekura-Sakakibara K, Fukuchi-Mizutani M, Ohira K,

Ueyama Y, Ohkawa H, Holton TA, Kusumi T, and Tanaka Y,

Plant Biotechnol., 21, 377–386 (2004).

25) Forkman G and Ruhnau B, Z. Naturforsch., 42c, 1146–1148

(1987).

26) Johnson ET, Yi H, Shin B, Oh BJ, Cheong H, and Choi G, Plant

J., 19, 81–85 (1999).

27) Martens S, Preuss A, and Matern U, Phyochemistry, 71, 1040–

1049 (2010).

28) Luo J, Nishiyama Y, Fuell C, Taguchi G, Elliott K, Hill L,

Tanaka Y, Kitayama M, Yamazaki M, Bailey P, Parr A,

Michael AJ, Saito K, and Martin C, Plant J., 50, 678–695

(2007).

29) Zhao J and Dixon RA, Trends Plant Sci., 15, 72–80 (2010).

30) Goodman CD, Casati P, and Walbot V, Plant Cell, 16, 1812–

1826 (2004).

31) Zhao J and Dixon RA, Plant Cell, 21, 2323–2340 (2009).

32) Gomez C, Terrier N, Torregrosa L, Vialet S, Fournier-Level A,

Verries C, Souquet JM, Mazauric JP, Klein M, Cheynier V, and

Ageorges A, Plant Physiol., 150, 402–415 (2009).

33) Fukada-Tanaka S, Inagaki Y, Yamaguchi T, Saito N, and Iida S,

Nature, 407, 581 (2000).

34) Yamaguchi T, Fukada-Tanaka S, Inagaki Y, Saito N,

Yonekura-Sakakibara K, Tanaka Y, Kusumi T, and Iida S,

Plant Cell Physiol., 42, 451–461 (2001).

35) Yoshida K, Miki N, Nomonoi K, Kawachi M, Katou K, Okazaki

Y, Uozumi N, Maeshima M, and Kondo T, Proc. Jpn. Acad.

Ser. B, 85, 187–197 (2009).

36) Verweij W, Spelt C, Di Sansebastiano G-P, Vermeer J, Reale L,

Ferranti F, Koes R, and Quattrocchio F, Nat. Cell Biol., 10,

1456–1462 (2008).

37) Mori M, Kondo T, and Yoshida K, Biosci. Biotechnol.

Biochem., 73, 2329–2331 (2009).

38) Koes R, Verweij W, and Quattrocchio F, Trends Plant Sci., 10,

236–242 (2005).

39) Ramsay NA and Glover BJ, Trends Plant Sci., 10, 63–70

(2005).

40) Tornielli G, Koes R, and Quattrocchio F, ‘‘The Genetics of

Flower Colour’’ 2nd. Edition, eds. Gerats T and Strommer J,

Springer, New York, pp. 269–299 (2009).

41) Nakatsuka T, Haruta KS, Pitaksutheepong C, Abe Y, Kakizaki

Y, Yamamoto K, Shimada N, Yamamura S, and Nishihara M,


42) Yamagishi M, Shimoyamada Y, Nakatsuka T, and Masuda K,


43) Lin-Wang K, Bolitho K, Grafton K, Kortstee A, Karunairetnam


S, McGhie TK, Espley RV, Hellens RP, and Allan AC, BMC

Plant Biol., 10, 50 (2010).

44) Espley RV, Brendolise C, Chagne’ D, Kutty-Amm S, Green S,

Volz R, Putterill J, Schouten HJ, Gardiner SE, Hellens RP, and

Allana AC, Plant Cell, 29, 168–183 (2009).

45) Laitinen RA, Ainasoja M, Broholm SK, Teeri TH, and Elomaa

P, J. Exp. Bot., 59, 3691–3703 (2008).

46) Peel GJ, Pang Y, Modolo LV, and Dixon RA, Plant J., 59, 136–

149 (2009).

47) Velten J, Cakir C, and Cazzonelli CI, PLoS One, 5, e9917 (2010).

48) Butelli E, Titta L, Giorgio M, Mock H-P, Matros A, Peterek S,

Schijlen EGWM, Hall RD, Bovy AG, Luo J, and Martin C, Nat.

Biotechnol., 26, 1301–1308 (2008).

49) Matus JT, Aquea F, and Arce-Johnson P, BMC Plant Biol., 8, 83

(2008).

50) Kobayashi S, Goto-Yamamoto N, and Hirochika H, Science,

304, 982 (2004).

51) Bogs J, Jaffe FW, Takos AM, Walker AR, and Robinson SP,

Plant Physiol., 143, 1347–1361 (2007).

52) Akagi T, Ikegami A, Tsujimoto T, Kobayashi S, Sato A, Kono

A, and Yonemori K, Plant Physiol., 151, 2028–2045 (2009).

53) Deluc L, Barrieu F, Marchive C, Lauvergeat V, Decendit A,

Richard T, Carde JP, Merillon JM, and Hamdi S, Plant Physiol.,

140, 499–511 (2006).

54) Deluc L, Bogs J, Walker AR, Ferrier T, Decendit A, Merillon

JM, Robinson SP, and Barrieu F, Plant Physiol., 147, 2041–

2053 (2008).

55) Quattrocchio F, Verweij W, Kroon A, Spelt C, Mol J, and Koes

R, Plant Cell, 18, 1274–1291 (2006).

56) Matus JT, Loyola R, Vega A, Pena-Neira A, Bordeu E,

Arce-Johnson P, and Alcalde JA, J. Exp. Bot., 60, 853–867

(2009).

57) Mehrtens F, Kranz H, Bednarek P, and Weisshaar B, Plant

Physiol., 138, 1083–1096 (2005).

58) Dubos C, Gourrierec JL, Baudry1 A, Huep G, Lanet E,

Debeaujon I, Routaboul J-M, Alboresi A, Weisshaar B, and

Lepiniec L, Plant J., 55, 940–953 (2008).

59) Matsui K, Umemura Y, and Ohme-Takagi M, Plant J., 55, 954–

967 (2008).

60) Morishita T, Kojima Y, Maruta T, Nishizawa-Yokoi A, Yabuta

Y, and Shigeoka S, Plant Cell Physiol., 50, 2210–2222 (2009).

61) Winkel BSJ, Annu. Rev. Plant Biol., 55, 85–107 (2004).

62) Oren-Shamir M, Plant Sci., 177, 310–316 (2009).

63) Katsumoto Y, Mizutani M, Fukui Y, Brugliera F, Holton T,

Karan M, Nakamura N, Yonekura-Sakakibara K, Togami J,

Pigeaire A, Tao G-Q, Nehra N, Lu C-Y, Dyson B, Tsuda S,

Ashikari T, Kusumi T, Mason J, and Tanaka Y, Plant Cell

Physiol., 48, 1589–1600 (2007).

64) Boase MR, Lewis DH, Davies KM, Marshall GB, Patel D,

Deroles S, and Schwinn KE, BMC Plant Biol., 10, 107 (2010).

65) Meyer P, Heidemann I, Forkmann G, and Saedler H, Nature,

330, 677–678 (1987).

66) Elomaa O, Helariutta Y, Griesbach RJ, Kotilainen M, Seppanen

P, and Teeri TH, Mol. Gen. Genet., 248, 645–649 (1995).

67) Okinaka Y, Shimada Y, Nakano-Shimada R, Ohbayashi M,

Kiyokawa S, and Kikuchi Y, Biosci. Biotechnol. Biochem., 67,

161–165 (2003).

68) Togami J, Tamura M, Ishiguro K, Hirose C, Okuhara H,

Ueyama Y, Nakamura N, Yonekura-Sakakibara K,

Fukuchi-Mizutani M, Suzuki K, Fukui Y, Kusumi T, and

Tanaka Y, Plant Biotechnol., 23, 5–11 (2006).

69) Nakatsuka T, Abe Y, Kakizaki Y, Yamamura S, and Nishihara

M, Plant Cell Rep., 26, 1951–1959 (2007).

70) Nakatsuka T, Mishiba K, Kubota A, Abe Y, Yamamura S,

Nakamura N, Tanaka Y, and Nishihara M, J. Plant Physiol.,

167, 231–237 (2009).

71) Ueyama Y, Katsumoto Y, Fukui Y, Fukuchi-Mizutani M,

Ohkawa H, Kusumi T, Iwashita T, and Tanaka Y, Plant

Biotechnol., 23, 19–24 (2006).

72) Meyer P, Linn F, Heidemann I, Meyer H, Niedenhof I, and

Saedler H, Mol. Gen. Genet., 231, 345–352 (1992).

73) Waterhouse PM, Graham MW, and Wang M-B, Proc. Natl.

Acad. Sci. USA, 95, 13959–13964 (1998).

74) Ono E, Fukuchi-Mizutani M, Nakamura N, Fukui Y,

Yonekura-Sakakibara K, Yamaguchi M, Nakayama T, Tanaka

T, Kusumi T, and Tanaka Y, Proc. Natl. Acad. Sci. USA, 103,

11075–11080 (2006).

75) Nakamura N, Fukuchi-Mizutani M, Suzuki K, Miyazaki K, and

Tanaka Y, Plant Biotechnol., 23, 13–18 (2006).

76) Nakamura N, Fukuchi-Mizutani M, Fukui Y, Ishiguro K, Suzuki

K, and Tanaka Y, Plant Biotechnol., in press.

77) Kumar S and Thompson WF, ‘‘Plant Genome Engineering

Using Zinc Finger Nucleases,’’ eds. Mohan JS and Brar DS,

Springer, New York, pp. 579–590 (2010).

78) Comai L, Moran P, and Maslyar D, Plant Mol. Biol., 15, 373–

381 (1990).

79) Mitsuhara I, Ugaki M, Hirochika H, Ohshima M, Murakami T,

Gotoh Y, Katayose Y, Nakamura S, Honkura R, Nishimiya S,

Ueno K, Mochizuki A, Tanimoto H, Tsugawa H, Otsuki Y, and

Ohashi Y, Plant Cell Physiol., 37, 49–59 (1996).


Documents

Flower Color