Sintesis flavonoles

12
110 Synthesis of flavonols & flavonol synthase gene expressionAustralian Journal of Grape and Wine Research 9, 110–121, 2003 Introduction Flavonols are products of the flavonoid biosynthetic pathway, which also gives rise to anthocyanins and to condensed tannins in grapevines (Stafford 1990, Darne 1993). Flavonols in plants have physiological functions ranging from microbial interactions (Koes et al. 1994) to pollen fertility (Taylor 1995) and free radical scavenging (Markham et al. 1998). However, their most widespread roles appear to be as UV protectants (Flint et al. 1985, Smith and Markham, 1998), and as copigments in flow- ers and fruit (Asen et al. 1972, Scheffeldt and Hrazdina 1978). Copigmentation is an association between flavonols and the anthocyanin pigments that confers stability on the coloured form of the anthocyanin molecule resulting in increased colour or altered hue (Hoshino et al. 1980, Osawa 1982). Of the three main flavonols kaempferol, quercetin and myricetin, mainly quercetin-3-O-glucoside and -3-O- glucuronide are found in grape berries (Cheynier and Rigaud 1986, Price et al. 1995). Flavonols have also been reported in the leaves (Hmamouchi et al. 1996) and stems of grapevines (Souquet et al. 2000). Flavonols occur in the upper epidermis of plant organs, consistent with their role in UV protection (Flint et al. 1985, Beggs et al. 1987), while anthocyanins are present in the cells of the lower epidermis (Asen 1975). However, for co- pigmentation to occur, anthocyanins and flavonols would have to be present in the same cells. Hrazdina and Moskowitz (1980, Moskowitz and Hrazdina 1981) identi- fied both flavonols and anthocyanins in vacuoles of the outer epidermis of the grape, but were not convinced these were associated as copigments. Nevertheless, flavonols may play a more significant role as copigments in wine. The colour of young red wines is due to anthocyanin pigments extracted from the grape skin, but these com- pounds are unstable and much of the initial colour is lost during fermentation and maturation (Somers and Evans Synthesis of flavonols and expression of flavonol synthase genes in the developing grape berries of Shiraz and Chardonnay (Vitis vinifera L.) MARK O. DOWNEY 1,2 , JOHN S. HARVEY 1,3 and SIMON P. ROBINSON 1,3,4 1 CSIRO Plant Industry, Horticulture Unit, PO Box 350 Glen Osmond, SA 5064, Australia 2 Department of Horticulture,Viticulture and Oenology,The University of Adelaide,Waite Campus, PMB 1, Glen Osmond, SA 5064 Australia 3 CRC for Viticulture, PO Box 154, Glen Osmond, SA 5064 Australia 4 Corresponding author: Dr Simon P. Robinson, facsimile +61 8 83038601, email [email protected] Abstract Flavonols were determined in Shiraz and Chardonnay grapes throughout berry development. The predominant flavonols were quercetin-3-glycosides with trace amounts of kaempferol-3-glycosides detected in Shiraz flowers but not in developing berries. Flavonols were present in the skin of ripening grapes but were not detected in seeds or flesh. Flavonols were also present in buds, tendrils, inflorescences, anthers and leaves. The concentration of flavonols in flowers (mg/g fresh weight) was high and decreased between flowering and berry set then remained relatively constant through berry development. The total amount of flavonols in berries (mg/berry) was low until pre-veraison then increased during berry development, particularly before veraison, the onset of ripening, in Chardonnay and during ripening in Shiraz. Two cDNA fragments with homology to genes encoding the enzyme flavonol synthase (FLS) were isolated from Shiraz flowers. In the overlapping region of the two cDNAs, they had 80% sequence identity at the nucleotide level and both had high homology to FLS genes from other plants. VvFLS1 was expressed in leaves, tendrils, pedicels, buds and inflorescences as well as in developing grapes. Expression was highest between flowering and fruit set then declined, increasing again during ripening coincident with the increase in flavonols per berry. Expression of VvFLS2 was much lower than for VvFLS1 and did not change during berry development. The results indicate that two distinct periods of flavonol synthesis occur in grapes, the first around flowering and the second during ripening of the developing berries. Keywords: Vitis vinifera, Shiraz, Chardonnay, grape, grape skin, leaves, berry development, flavonols, quercetin, flavonol synthase, gene expression

Transcript of Sintesis flavonoles

Page 1: Sintesis flavonoles

110 Synthesis of flavonols & flavonol synthase gene expressionAustralian Journal of Grape and Wine Research 9, 110–121, 2003

IntroductionFlavonols are products of the flavonoid biosynthetic pathway, which also gives rise to anthocyanins and tocondensed tannins in grapevines (Stafford 1990, Darne1993). Flavonols in plants have physiological functionsranging from microbial interactions (Koes et al. 1994) topollen fertility (Taylor 1995) and free radical scavenging(Markham et al. 1998). However, their most widespreadroles appear to be as UV protectants (Flint et al. 1985,Smith and Markham, 1998), and as copigments in flow-ers and fruit (Asen et al. 1972, Scheffeldt and Hrazdina1978). Copigmentation is an association betweenflavonols and the anthocyanin pigments that confersstability on the coloured form of the anthocyanin molecule resulting in increased colour or altered hue(Hoshino et al. 1980, Osawa 1982).

Of the three main flavonols kaempferol, quercetinand myricetin, mainly quercetin-3-O-glucoside and -3-O-glucuronide are found in grape berries (Cheynier and

Rigaud 1986, Price et al. 1995). Flavonols have alsobeen reported in the leaves (Hmamouchi et al. 1996) andstems of grapevines (Souquet et al. 2000). Flavonolsoccur in the upper epidermis of plant organs, consistentwith their role in UV protection (Flint et al. 1985, Beggset al. 1987), while anthocyanins are present in the cells ofthe lower epidermis (Asen 1975). However, for co-pigmentation to occur, anthocyanins and flavonols wouldhave to be present in the same cells. Hrazdina andMoskowitz (1980, Moskowitz and Hrazdina 1981) identi-fied both flavonols and anthocyanins in vacuoles of theouter epidermis of the grape, but were not convincedthese were associated as copigments. Nevertheless,flavonols may play a more significant role as copigmentsin wine.

The colour of young red wines is due to anthocyaninpigments extracted from the grape skin, but these com-pounds are unstable and much of the initial colour is lostduring fermentation and maturation (Somers and Evans

Synthesis of flavonols and expression of flavonol synthasegenes in the developing grape berries of Shiraz

and Chardonnay (Vitis vinifera L.)

MARK O. DOWNEY1,2, JOHN S. HARVEY1,3 and SIMON P. ROBINSON1,3,4

1 CSIRO Plant Industry, Horticulture Unit, PO Box 350 Glen Osmond, SA 5064, Australia2 Department of Horticulture,Viticulture and Oenology,The University of Adelaide,Waite Campus,

PMB 1, Glen Osmond, SA 5064 Australia3 CRC for Viticulture, PO Box 154, Glen Osmond, SA 5064 Australia

4 Corresponding author: Dr Simon P. Robinson, facsimile +61 8 83038601, email [email protected]

AbstractFlavonols were determined in Shiraz and Chardonnay grapes throughout berry development. Thepredominant flavonols were quercetin-3-glycosides with trace amounts of kaempferol-3-glycosidesdetected in Shiraz flowers but not in developing berries. Flavonols were present in the skin of ripeninggrapes but were not detected in seeds or flesh. Flavonols were also present in buds, tendrils,inflorescences, anthers and leaves. The concentration of flavonols in flowers (mg/g fresh weight) washigh and decreased between flowering and berry set then remained relatively constant through berrydevelopment. The total amount of flavonols in berries (mg/berry) was low until pre-veraison thenincreased during berry development, particularly before veraison, the onset of ripening, in Chardonnayand during ripening in Shiraz. Two cDNA fragments with homology to genes encoding the enzymeflavonol synthase (FLS) were isolated from Shiraz flowers. In the overlapping region of the two cDNAs,they had 80% sequence identity at the nucleotide level and both had high homology to FLS genes fromother plants. VvFLS1 was expressed in leaves, tendrils, pedicels, buds and inflorescences as well as indeveloping grapes. Expression was highest between flowering and fruit set then declined, increasing againduring ripening coincident with the increase in flavonols per berry. Expression of VvFLS2 was much lowerthan for VvFLS1 and did not change during berry development. The results indicate that two distinctperiods of flavonol synthesis occur in grapes, the first around flowering and the second during ripeningof the developing berries.

Keywords: Vitis vinifera, Shiraz, Chardonnay, grape, grape skin, leaves, berry development, flavonols, quercetin, flavonol synthase, gene expression

Page 2: Sintesis flavonoles

Downey, Harvey & Robinson Synthesis of flavonols & flavonol synthase gene expression 111

1979, Somers and Verette 1988). The colour of olderwines results from increasingly stable associationsbetween anthocyanins and other flavonoid compoundsto form polymeric pigments (Nagel and Wulf 1979,McCloskey and Yengoyan 1981, Somers and Verette,1988, Mateus et al. 2002). Polymeric pigments formslowly in the wine and anthocyanin-copigment complex-es are considered intermediates that not only maintainthe anthocyanins in the wine, but align them infavourable orientations for the formation of more stableassociations (Brouillard and Dangles 1994, Boulton2001). The most stable copigment associations occurbetween the flavonols quercetin and quercetin-3-O-glucoside and the main anthocyanin in red wines, mal-vidin-3-O-glucoside (Baranac et al. 1997, Lambert 2002).These observations suggest that the critical factor indetermining the extent of colour in mature wine is theconcentration of anthocyanins and copigments at harvest(Lambert 2002). Thus, any process that alters the antho-cyanin or flavonol content in grapes may influence winequality.

Previous investigations into grape flavonols havelargely focused on the flavonol composition of the ripefruit (Cheynier and Rigaud 1986) and effects of sun exposure (Price et al. 1995). Haselgrove et al. (2000) examined flavonols earlier in berry development, report-ing relatively high levels (0.15 mg/berry) at veraison, theonset of ripening, with little change towards harvest.Furthermore, while many genes involved in flavonoidbiosynthesis have been cloned in Vitis vinifera (Sparvoli etal. 1994), this has not been the case for flavonol synthase(FLS), the gene that encodes the biosynthetic enzymeconverting dihydroflavonols to flavonols (Spribille andForkmann, 1984). This gene has been cloned in othergenera including Petunia, Arabidopsis, Solanum andCitrus (Holton et al. 1993, Pelletier et al. 1997, van Eldiket al. 1997, Moriguchi et al. 2002). Analysis of flavonolaccumulation and flavonol synthase expression through-out berry development would greatly expand our under-standing of flavonoid biosynthesis in grapevines.

In this paper, we report the cloning and characteris-ation of cDNAs encoding flavonol synthase fromgrapevine together with a study of flavonol accumulationin grapes during berry development.

Materials and methods

Grapevine tissue collectionGrapevine tissues of the Vitis vinifera L. cultivars Shirazand Chardonnay were collected from a commercial vine-yard in Willunga, South Australia (34° 46’ south, 138°32’ east). The region has a maritime climate with themean daily temperature of 23°C during berry develop-ment. Shiraz samples were collected at weekly intervalsfrom flowering until harvest during the 1999–2000 and2001–2002 seasons and from bud-burst until harvest inthe 2000–2001 season. Samples of Chardonnay fruitwere collected from flowering until harvest during the1999–2000 season.

Bunches were collected randomly from 120 vines onmodified Scott-Henry trellises spread across four rows

and ten panels of a single block each of Chardonnay andShiraz, with a two-row and two-panel buffer to avoidend effects. A minimum of twenty bunches were collect-ed on each date and the berries removed and bulked. Asub-sample of approximately 100 berries was taken formeasurement of berry weight and total soluble solids. Theremaining berries were dissected in the field and im-mediately frozen in liquid nitrogen and stored at –80°C for subsequent analysis. During the first season(1999–2000), whole berries were collected, while in thesubsequent seasons (2000–2001 and 2001–2002), seedsand skin were collected. In 2000–2001, whole flowerswere collected nine, eight and seven weeks pre-veraison.Samples for six and five weeks pre-veraison compriseddeveloping berries with the seeds removed. From fourweeks pre-veraison onwards it was possible to separatethe skin from developing berries in the field.

Shiraz leaves were collected at five stages of develop-ment. The first stage was a newly emerged leaf less than2 cm2 and 0.09 g, densely villous with pigmented leafmargins. The second stage was the next oldest leaf afterStage 1, approximately 5–6 cm2 and 0.22 g, becomingglabrous on the dorsal surface of the leaf lobes. Shiraz leafStage 3 was approximately 15 cm2 and 0.44 g, withnoticeably reduced trichome density and a faint trace ofanthocyanin pigment remaining in leaf epidermis. Stage4 was an expanded leaf around 40 cm2 and 0.99 g withglabrous dorsal, but villous ventral surface with a trans-lucent appearance and no visible anthocyanins. The finalleaf stage, Stage 5, was a fully expanded mature leafapproximately 160 cm2 in area, weighing 3.62 g and nolonger translucent. Tendrils and pedicels were collectedfrom Shiraz vines when berries were collected at fruit-set.Tendrils were removed at the axil; pedicels were separatedfrom the bunch rachis and the developing fruit. Sampleswere immediately frozen in liquid nitrogen and stored at–80°C for later analysis.

In addition, early vegetative stages were collected. Atbudburst the emerging buds were excised from the canes.On successive weeks the developing shoots were collected,the growth stages of the developing shoots were deter-mined according to the modified Eichhorn-Lorenz (E-L)scheme (Coombe 1995) and corresponded to E-L Stage 4(buds), 13 (inflorescence) and 17 (flowers). Antherswere collected at flowering by shaking bunches over anopen container to collect the anthers as they dropped off.This material was then sorted to remove flowers, flowercaps and other extraneous matter. All samples wereimmediately frozen in liquid nitrogen and stored at –80°Cfor later analysis.

Roots of Shiraz vines were collected from glasshouse-grown potted vines. Plants were removed from pots andthe soil shaken free from the root ball. The roots werethen washed to remove the remainder of the soil andblotted dry. Root samples of 1–2 cm in length were cutfrom the root tips, frozen immediately in liquid nitrogenand stored at –80°C.

HPLC analysesFor HPLC analysis of flavonols, the frozen samples were

Page 3: Sintesis flavonoles

112 Synthesis of flavonols & flavonol synthase gene expressionAustralian Journal of Grape and Wine Research 9, 110–121, 2003

ground to a fine powder under liquid nitrogen. Threeseparate aliquots of 0.1 g were extracted in 1.0 mL ofacidulated ethanol (Price et al. 1995) for 5 min at 70°Cfollowed by 20 min at room temperature with occasionalvortexing. Extracts were then centrifuged (15 min at13,000 × g) and 200 µL of the supernatant transferred toHPLC auto-sampler vials.

Samples were analysed using a Hewlett Packard HP1100 high-performance liquid chromatograph (HPLC)with two Merck (Germany) Chromolith analyticalcolumns in series protected by an SGE (Australia) C-18guard column. The separation used 0.2% acetic acid(solvent A) with a methanol (solvent B) gradient (gradi-ent of solvent B: zero min, 24%; 10 min, 28%; 10.1 min,100%; 13 min, 100%; 13.1 min 24%) with a flow rate of5 mL/min. The major flavonol peak was identified bycomparison of the elution time and absorbance spectrawith a commercial standard of quercetin-3-O-glucosideobtained from Extrasynthese (France). In addition, themain flavonol peak was collected and analysed by massspectroscopy with electrospray ionisation (ESI-MS). Theintegrated absorbance at 353 nm was used to determinethe concentration of flavonol glycosides in each sample(25 µL injection) expressed as quercetin-3-O-glucosideequivalents. Where present, kaempferol-3-O-glucosidewas identified by comparison of elution time andabsorbance spectra with a kaempferol-3-O-glucoside stan-dard from Extrasynthase (France). Concentrations ofquercetin-3-O-glucoside and kaempferol-3-O-glucosideequivalents were calculated from a standard curve prepared from commercial standards.

Cloning of flavonol synthase genesFor the cloning and expression of flavonol synthasecDNAs and gene fragments, grapevine material frozen inliquid nitrogen in the field and stored at –80oC wasground to a fine powder in liquid nitrogen using a pre-frozen mortar and pestle. Genomic DNA was extractedfrom frozen Stage 2 leaf tissue following the method ofThomas et al. (1993).

Total RNA was extracted from 1.0 g of the frozenpowdered tissue in 7.0 mL of hot borate buffer asdescribed by Wan and Wilkins (1994). Final pelletedRNAs were resuspended in Tris-EDTA (TE) and stored at–40oC in a solution containing 1/10th volume sodiumacetate and 3 volumes of 100% ethanol. Two 5.0 µgaliquots of total RNA for each sample were pelleted bycentrifugation at 13,000 × g for 20 minutes at 4oC, thepellet was rinsed in 70% cold ethanol and allowed to airdry. First strand cDNA synthesis was performed using the3’-RACE adapter primer and Superscript II (InvitrogenLife Technologies, Australia) using methods recommend-ed by the supplier in a final reaction volume of 25 µL.After heat inactivation of the reverse transcriptase, tworeactions were combined to give a working 50 µL stockand stored at –20oC.

cDNA fragments encoding flavonol synthase (FLS)were obtained by PCR of cDNA prepared from RNA isolated from Shiraz flowers. Degenerate primers weredesigned based on FLS sequences for other dicotyledon-

ous plants in the GENBANK database. The forward primerwas 5’-GCGAATTCGAA/GAAIGAA/GCAA/GCCIGC-3’ (FLS9),designed to the protein sequence EN/KEQPA and thereverse primer was 5’-ACIGGCCAIC/GA/TCATICG/TIGTC/TTT-3’ (FLS6), designed to the protein sequence KTRM-SWPV. The PCR was carried out according to the methodof Frohman et al. (1988) and a predicted band of approx-imately 800 bp was obtained and cloned into the pGEM-T Easy vector from Promega (Australia). A number ofindependent clones were sequenced and one clone(FLS36) was sequenced in both directions and found toencode a 797 bp fragment with high homology to FLSsequences in the database. This sequence was named aspVvFLS1 (GenBank Accession No. AY257978).

A second gene fragment encoding FLS was identifiedby PCR from grapevine genomic DNA. The equivalentcDNA was amplified from the flower cDNA using a gene-specific primer based on the genomic sequence. FlowercDNA was amplified by PCR using the specific primer 5’-AATCCTCCTTCTTACAGGGATGCG-3’ (FLS10) and the oligo-dT adapter primer B26 described by Frohman et al.(1988). This yielded a band of approximately 700 bp thatwas cloned and sequenced as described above to yield a706 bp cDNA fragment (FLS26), also encoding FLS. Thissecond sequence was named pVvFLS2 (GenBankAccession No. AY257979).

Expression of flavonol synthase genesGene expression was determined by real-time PCR analy-sis. cDNA syntheses for all RNA samples to be analysedwere performed simultaneously. Reagents required forthe reverse transcription (RT) reactions were made as amaster mix and aliquoted into each reaction in a singlestep to avoid pipetting errors that may have affected theefficiency of the RT reactions in different samples.

Quantitative real-time PCR was performed on aRotor-Gene 2000 (Corbett Research, Australia) real-timePCR machine using SYBR green as the system of detec-tion for double stranded PCR products. Primers homolo-gous to the grapevine cDNAs encoding VvFLS1, VvFLS2and VvUbiquitin1 were designed using primer 3 (http://www-genome.wi.mit.edu/cgi-bin/primer/primer3_www.cgi). The primer pairs VvFLS1 Forward: 5’-CAGGGCTTG-CAGGTTTTTAG-3’, VvFLS1 Reverse: 5’-GGGTCTTCTC-CTTGTTCACG-3’ and VvFLS:2 Forward 5’-TGAGCTGGC-TATAGGTCCTC-3’ and VvFLS:2 Reverse 5’-TGCATGTA-CACTGGAAAAGG-3’ were used to generate PCR productsof 154 bp and 195 bp respectively. The VvUbiquitin1transcript was detected using the primers VvUbiquitinReverse: 5’-AACCTCCAATCCAGTCATCTAC-3’ and Vv-Ubiquitin Forward: 5’-GTGGTATTATTGAGCCATCCTT-3’which amplify a product of 182 base pairs.

All primer pairs were used under identical conditions,95oC for 5 minutes followed by 35 rounds of amplifica-tion (95oC/30sec, 57oC/30sec, 72oC/30sec). Each PCRreaction contained 1 × SYBR Green PCR Master Mix(Applied Biosystems, UK), 30 nmoles of each primer and1/50th volume of cDNA stock in a final reaction volumeof 25 µL. The VvFLS1 and VvFLS2 mRNAs were quantifiedas a proportion of VvUbiquitin1 using the Rotor Gene

Page 4: Sintesis flavonoles

Downey, Harvey & Robinson Synthesis of flavonols & flavonol synthase gene expression 113

2000 software. Expression of VvFLS1 and VvFLS2 mRNAis presented as a ratio of expression of VvFLS1 and VvFLS2to expression of VvUbiquitin1and therefore has no units.

Southern and Northern blot analysesSouthern and Northern blot analysis was performed asdescribed by Sambrook et al. (1989) using 0.4 M and 50mM NaOH respectively as transfer buffer. Nucleic acidswere transferred onto Zeta-Probe nylon membranes (Bio-Rad, Australia) and fixed as recommended by the suppli-er. Membranes were pre-hybridised in 7% SDS, 0.25 MSodium Phosphate (pH 7.0), 1.0 mM EDTA at 65oC for 2hours. DNA fragments specific for VvFLS1 and VvFLS2were excised from pGEM T-easy (Promega, Australia)clones by digestion with EcoRI according to the instruc-tions of the supplier. Vector and insert fragments wereresolved on 1% agarose gels and the insert bandsremoved under UV light after staining with ethidium bro-mide. Insert DNA was purified from the gel fragmentusing a QIAquick gel extraction kit (QIAGEN, Australia)according to the manufacturers instructions. Inserts wereradioactively labelled using a Rediprime II labelling kit(Amersham Biosciences, Australia), added to the pre-hybridisation solution after heat denaturation andhybridised overnight at 65oC. Membranes were rinsed in2 × SSC, 0.1% SDS at room temperature before beingwashed in 1.0 × SSC, 0.1% SDS for 20 minutes at 65oCand 0.1 × SSC, 0.1%SDS for 20 minutes. Membraneswere then rinsed in 2 × SSC at room temperature beforebeing exposed to Biomax (MS) film (Eastman Kodak Co.,USA) at –80oC with appropriate intensifying screens.

Results

Cloning of flavonol synthase genesThe flavonoid pathway produces a range of secondarycompounds including anthocyanins, flavonols and con-densed tannins. The first step in the branch of this path-way leading to synthesis of flavonols is catalysed by theenzyme flavonol synthase (FLS). To determine expressionlevels of the genes encoding FLS in grapevine tissues, wefirst isolated cDNA clones encoding FLS to provide geneprobes. RNA extracted from Shiraz flowers was used tomake cDNA by reverse transcription and this was ampli-fied using the polymerase chain reaction (PCR) and theDNA fragments obtained were cloned and sequenced inboth directions. This yielded two distinct grapevine FLSclones: pVvFLS1, a 797 bp fragment of FLS encoding 265amino acids in the open reading frame of the gene, andpVvFLS2, a 706 bp fragment encoding 185 amino acids atthe C-terminal end of the protein plus a 3’-untranslatedregion of 151 bp. The two fragments shared a 425 bpoverlap and pVvFLS1 and pVvFLS2 had 79% sequenceidentity at the nucleotide level in this overlapping region.The two grapevine FLS sequences have been deposited inthe GenBank database (Accession numbers AY257978and AY257979). The two grape FLS cDNA sequences hadhigh protein sequence homology to other FLS genes inthe two conserved dioxygenase domains (data notshown), indicating that they are part of the dioxygenasefamily (van Eldik et al. 1997, Moriguchi et al. 2002).

Both grape sequences had high homology to FLS genes inthe databases and the nearest matches at the amino acidlevel were to Petunia FLS (Z22543; 73% identity topVvFLS1, 74% identity to pVvFLS2) and Citrus FLS(BAA36554; 73% identity to pVvFLS1, 70% identity topVvFLS2). A number of other degenerate primersdesigned to plant FLS sequences were used in PCR ofcDNA from grape leaves and flowers but no additionalFLS genes were identified.

Southern analysis performed on Shiraz genomic DNAdigested with HindIII showed that the VvFLS1 and VvFLS2probes bound specifically to distinct bands. Observation ofthe hybridisation pattern showed that each probe alsocross hybridised with the target genomic sequence of theopposing clone but that the cross hybridising bands wereremoved by high stringency washing (results not shown).No significant hybridisation to additional bands wasdetected.

Identification of flavonols in grapevinesFlavonols in the Vitis vinifera L. cultivars Chardonnay andShiraz were analysed by HPLC and were identified initially by comparison of their elution times andabsorbance spectra with commercially available standardsof quercetin-3-O-glucoside or kaempferol-3-O-glucoside.Figure 1a shows the separation of these standards usingthe above HPLC conditions described in the Materials andmethods (HPLC analyses). Typically, a HPLC separation ofgrape tissue extracts (Figure 1b) showed two peakscorresponding to the quercetin-3-O-glucoside andkaempferol-3-O-glucoside standards, based on their elution times and absorbance spectra. In all samples, thepeak corresponding to quercetin-3-glucoside accountedfor most of the flavonols. The quercetin aglycone, whichwas separable from the glycosides by HPLC, was notdetected in any of the grape samples. Figure 1b shows theHPLC profile for Shiraz flowers at eight weeks pre-veraison, the only time during Shiraz berry developmentwhen the kaempferol glycoside could be detected.Kaempferol derivatives were not detected in Chardonnayberries or flowers. The major flavonol peak, corres-ponding to quercetin-3-O-glucoside was collected andanalysed by mass spectrometry. The mass spectrum fromthis peak (Figure 1c) showed three ionisation productspresent, corresponding to the aglycone quercetin,quercetin-3-O-glucoside and quercetin-3-O-glucuronidewith quercetin likely arising from hydrolysis of the glyco-sides in the MS ion beam.

Berry development, flavonol accumulation and FLS expressionin Chardonnay berriesTo determine when flavonols are synthesised duringberry development, samples of Chardonnay grapes col-lected during the 1999–2000 growing season wereanalysed for flavonol content by HPLC. The pattern ofberry development and the accumulation of flavonolsand expression of VvFLS1are presented in Figure 2.Flowering occurred in the week of 21 October 1999 andveraison, the onset of ripening, occurred nine weekslater, around 23 December 1999 (Figure 2a). Six weeks

Page 5: Sintesis flavonoles

114 Synthesis of flavonols & flavonol synthase gene expressionAustralian Journal of Grape and Wine Research 9, 110–121, 2003

after veraison, the Chardonnay was harvested at23.2°Brix on 03 February 2000. Berry weight increasedslowly from flowering until six weeks pre-veraison, how-ever this represented a 12-fold increase in berry weight(Figure 2a). From six weeks pre-veraison berry weightincreased steadily until one week pre-veraison, thenremained static until veraison when berry growth recom-menced for a further three weeks. During this time thelevel of sugar in the berry, measured as total soluble solids(°Brix), increased steadily (Figure 2a). Berry weight andtotal soluble solids were constant for the next two weeks,while in the final week before harvest, there was adecrease in berry weight and an increase in total soluble

solids. This observation was consistent with water lossfrom the ripe fruit.

In Chardonnay berries, only glycosides of quercetinwere detected. At flowering there was a high concentra-tion of flavonols (1.46 ± 0.13 mg/g fresh weight), how-ever as the berry began to develop there was a 5-folddecrease in flavonol concentration (Figure 2b). Thisdecline continued until two weeks pre-veraison, afterwhich time the level of flavonols per gram of tissueremained relatively constant until harvest (0.03 ± 0.005mg/g fresh weight). During this time, the berry wasincreasing in size suggesting an overall increase in totalflavonols in the berry. This was the pattern observed on a

Quercetin-3-glucoside

Kaempferol-3-glucosidea

0 2 4 6 8 10 120 2 4 6 8 10 12

0

20

40

60

80

Elution time (minutes)

Dete

ctor

resp

onse

(mAU

)De

tect

or re

spon

se (m

AU)

0 2 4 6 8 10 12

0

20

40

60

80

0 2 4 6 8 10 12

Elution time (minutes)

b

303.2

479.0

465.2

200 300 400 500 600 700

m/z

2

4

6

8

10

12

14

16

Ion

inte

nsity

(x 1

0–3)

Quercetin

Quercetin glucoside

Quercetin glucuronide

c

OH

O

O

OO

O

O

O

OO

O

H

H H

H

H

H HH

H

H

O

HO

HO

HO

HO

HO

HO

OH

OHOH

OHOH

OHOH

OH

OHOH

OH

OHOH

OH

Figure 1. Identification of flavonols in Vitis vinifera by high-performance liquid-chromatography (HPLC with detection at 353nm) and electrospray ionisation mass spectrometry (ESI-MS). (a)HPLC separation of commercial standards of quercetin-3-glucosideand kaempferol-3-glucoside (25 µL injection, 50 µg/mL). (b) HPLCseparation of extract from Shiraz flowers (8 weeks pre-veraison, 16November 2000; 25µL injection). (c) Mass spectrum of the majorflavonol peak in Vitis vinifera tissues corresponding to quercetin-glucoside; three ionisation products are apparent, quercetin (M+H+,m/z 303.2), quercetin-3-O-glucoside (M+H+, m/z 465.2) andquercetin-3-O-glucuronide (M+H+, m/z 479.0).

0.00

0.40

0.80

1.20

1.60

-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6Weeks post-veraison

Flav

onol

s (m

g/g)

b

0.00

0.01

0.02

0.03

0.04

0.05

-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6Weeks post-veraison

Flav

onol

s (m

g/be

rry) c

0.00

0.25

0.50

0.75

1.00

1.25

1.50

-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6Weeks post-veraison

Berr

y w

eigh

t (g)

0

10

20

30

°Brix

a

Berry weight

°Brix

0.0

0.1

0.2

0.3

0.4

-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6Weeks post-veraison

VvFL

S1 e

xpre

ssio

n d

Figure 2. Grape berry development of Vitis vinifera L. cv. Chardonnaygrown in Willunga, South Australia in the 1999–2000 season.Flowering occurred during the week of 21 October 1999 (9 weekspre-veraison) with veraison around 23 December 1999 andcommercial harvest on 03 February 2000 (6 weeks post-veraison).(a) Chardonnay berry development, berry weight (�) and totalsoluble solids expressed as °Brix (�). (b) Quercetin glycosides(mg/g fresh weight of berry ±SEM). (c) Quercetin glycosides(mg/berry ±SEM). (d) Ratio of the expression of the flavonolsynthase gene (VvFLS1) in developing grape berries relative to theexpression of VvUbiquitin1 determined every second week after 8weeks pre-veraison.

Page 6: Sintesis flavonoles

Downey, Harvey & Robinson Synthesis of flavonols & flavonol synthase gene expression 115

per berry basis (Figure 2c), where total quercetin glyco-sides per berry were relatively constant for the first fewweeks (0.009 mg/berry) followed by a two-fold increaseat five weeks pre-veraison (0.019 mg/berry). This levelwas maintained until one week pre-veraison whenflavonols began to accumulate once more. At one weekpost-veraison, quercetin glycosides increased again tothe maximum level observed in the berry duringdevelopment (0.035 mg/berry). From one week post-veraison the flavonol level remained relatively constantuntil harvest.

The expression of a grapevine gene encoding theflavonol synthase biosynthetic enzyme (VvFLS1) inChardonnay berries during the 1999–2000 season isshown in Figure 2d relative to an internal control(VvUbiquitin1) whose expression was constant through-out berry development (data not shown). In flowers,nine weeks pre-veraison, a high level of VvFLS1 expres-sion was observed, which declined markedly after flow-ering and remained low until veraison. After veraison,the level of VvFLS1 expression increased steadily towardsharvest with a 10-fold increase in VvFLS1 expression dur-ing this time. The final level of VvFLS1 expression ingrape berries at harvest was similar to that at flowering.The expression of VvFLS2 was not determined inChardonnay berries.

Berry development, flavonol accumulation and FLS expressionin Shiraz berriesThe results with Chardonnay were extended to the black grape, Shiraz to determine whether the presence of anthocyanins influenced flavonol accumulation.Primarily, results from the 2000–2001 season are pre-sented here with the previous and successive seasonsincluded for comparison. Shiraz grapes were collectedover three seasons; 1999–2000, 2000–2001 and 2001–2002. Table 1 shows the flowering and veraison dates aswell as the date on which berries attained 24°Brix forShiraz grapes over the three successive seasons. Berryweight, skin weight and flavonol levels at 24°Brix arealso indicated. The patterns of berry development,flavonol accumulation and expression of VvFLS1 duringthe 2000–2001 season are presented in Figure 3. Berrygrowth during the 2000–2001 season was measured asberry weight and showed a six-fold increase from seven

0.00

0.01

0.02

0.03

0.04

0.05

-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9

Flav

onol

s (m

g/be

rry)

c

0 .0

0 .1

0 .2

0 .3

0 .4

0 .5

-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9

VvF

LS1

expr

essi

on d

0.0

0.3

0.6

0.9

1.2

1.5

1.8

-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9

Flav

onol

s (m

g/g) b

0 .0

0 .4

0 .8

1 .2

1 .6

-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9

Weeks post-veraison

Weeks post-veraison

Weeks post-veraison

Weeks post-veraison

Berr

y w

eigh

t (g)

0

1 0

2 0

3 0

°Brix

a

Berry weight

°Brix

Figure 3. Grape berry development of Vitis vinifera L. cv. Shirazgrown in Willunga, South Australia in the 2000–2001 season.Flowering occurred around 16 November 2000 (8 weeks pre-veraison) with veraison around 11 January 2001 and commercialharvest on 13 March 2001 (9 weeks post-veraison). (a) Shiraz berrydevelopment, berry weight (�) and total soluble solids expressed as°Brix (�). (b) Quercetin glycosides (mg/g fresh weight of skin±SEM). (c) Quercetin glycosides (mg/berry ±SEM). (d) Ratio of theexpression of the flavonol synthase gene (VvFLS1) in developinggrape skins relative to the expression of VvUbiquitin1.

Table 1. Variation in the timing of flowering, veraison, ripening and in some berry parameters for Vitis vinifera L. cv.Shiraz grown in three successive seasons, at Willunga, SA.

Season Date of Date of a Date berries Berry weight Skin weight b Flavonols at 24°Brixflowering veraison reached approx at at

24°Brix 24°Brix 24°Brix(g) (g) mg/g skin mg/berry

1999–2000 04 Nov 99 06 Jan 00 17 Feb 00 1.36 —c —c 0.046 ± 0.0012000–2001 16 Nov 00 11 Jan 01 01 Mar 01 1.27 0.39 0.086 ± 0.002 0.033 ± 0.0012001–2002 23 Nov 01 31 Jan 02 08 Mar 02 1.16 0.44 0.071 ± 0.009 0.026 ± 0.003

a The grapes reached 24°Brix six weeks post-veraison in 1999–2000, seven weeks post-veraison in 2000–2001, and five weeks post-veraison in 2001–2002. b As quercetin glycosides, ± SEM, n = 3.c Shiraz grape skin was not separated from the berry during the first season, 1999–2000.

Page 7: Sintesis flavonoles

116 Synthesis of flavonols & flavonol synthase gene expressionAustralian Journal of Grape and Wine Research 9, 110–121, 2003

weeks pre-veraison (fruit-set) until four weeks pre-veraison when berry growth slowed (Figure 3a). Berryweight reached a maximum five weeks post-veraisonthen declined until commercial harvest at nine weekspost-veraison. From veraison, sugars in the berry, measured as total soluble solids (°Brix), also increasedsteadily reaching 29.3°Brix at commercial harvest. Thisincrease in total soluble solids from five weeks post-veraison coincided with a decrease in berry weight,which was consistent with previous reports of berryshrivel in ripe Shiraz (McCarthy 1999).

The concentration of quercetin glycosides duringberry development was highest in the unopened Shirazflowers, nine weeks pre-veraison (1.60 ± 0.050 mg/gfresh weight, Figure 3b). At flowering (50% cap-fall) thefollowing week, the level had fallen to 1.31 ± 0.069 mg/gand in the following week to fruit-set, a substantialdecrease was observed to 0.17 ± 0.005 mg/g, followed bya smaller decline in the next week. From this time untilharvest the flavonol concentration remained relativelyconstant (0.08 ± 0.004 mg/g fresh weight of skin). On aper berry basis (Figure 3c), there was an initial increase intotal quercetin glycosides in the developing berries fromfruit-set (0.011 ± 0.001 mg/berry) until five weeks pre-veraison. Although some fluctuations were observed,this level was maintained until veraison (0.013 ± 0.001mg/berry). At veraison, flavonols in the Shiraz berrybegan to increase and continued to do so until six weekspost-veraison (0.049 ± 0.001 mg/berry). For the follow-ing two weeks flavonols appeared to decline, however inthe final week, total flavonols in the berry increasedagain.

The expression of the gene encoding the flavonol synthase biosynthetic enzyme (VvFLS1) during berrydevelopment in the 2000–2001 season is shown in Figure3d relative to the expression of VvUbiquitin1 whoseexpression remained constant throughout berry develop-ment (real-time PCR data not shown). In flowers prior tocap-fall, a high level of VvFLS1 expression was observed,which declined markedly at flowering (eight weeks pre-veraison) and during the following week, remaining at alow level from six weeks pre-veraison until veraison.After veraison, the level of VvFLS1 expression began toincrease slowly, with a noticeable increase three weekspost-veraison. This was followed by an additionalincrease from four to six weeks post-veraison. From twoto six weeks post-veraison there was a 10-fold increase inVvFLS1 expression. This was followed by a decline andthen a further increase over the final two weeks of berrydevelopment until harvest. The final level of VvFLS1expression was marginally higher than that seen inunopened flowers.

VvFLS2 was constitutively expressed in all the devel-oping berry skin samples but at levels 104–105 times lowerthan that of VvFLS1 in corresponding samples. Theexpression of VvFLS2 did not vary in any of the skin sam-ples that were assayed and was thus incongruent withthe observed pattern of flavonol accumulation. Theseresults were confirmed by Northern blot analysis (datanot presented).

Seasonal variation in flavonol accumulation in Shiraz As seasonal variation can be significant in grapes, measure-ments of flavonol content were carried out over threeseasons for Shiraz grapes in the same vineyard. Figure 4shows the variation in total soluble solids (°Brix) andflavonols (quercetin glycosides) across three seasons,1999–2000, 2000–2001 and 2001–2002 in Vitis vinifera L.cv. Shiraz berries. In the 1999–2000 season, floweringoccurred during the week of 04 November 1999 withveraison nine weeks later around 06 January 2000 andcommercial harvest seven weeks later on 24 February2000 (Table 1). In the 2000–2001 season, floweringoccurred two weeks later than in the 1999–2000 season,during the week of 16 November 2000. Veraisonoccurred eight weeks after flowering, around 11 January2001, with commercial harvest after a further nine weekson 13 March 2001. During the 2001–2002 season flower-ing occurred later than in either of the preceding seasons,around 23 November 2001. In addition, the timebetween flowering and veraison was longer than in theprevious seasons at ten weeks, veraison occurring at theend of January (31 January 2002). However, in the2001–2002 season the time between veraison and com-mercial harvest was the shortest recorded during theseinvestigations at six weeks. Shiraz grapes were harvestedat 26.4 °Brix on 15 March 2002 (Table 1).

Ripening of the fruit progressed in similar fashioneach season with sugar accumulation (°Brix) increasingfrom veraison towards harvest, reaching 24°Brix betweenfive and seven weeks post-veraison (Table 1, Figure 4a).The pattern of flavonol (quercetin glycoside) accumula-tion that was observed in Shiraz grape skin during the2000–2001 season was observed over all three of the

0

5

10

15

20

25

30

-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9Weeks post-veraison

°Brix

1999-2000

2000-20012001-2002

a

0.00

0.01

0.02

0.03

0.04

0.05

0.06

-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9

Weeks post-veraison

Flav

onol

s (m

g/be

rry)

1999-20002000-20012001-2002

b

Figure 4. Seasonal variation in the accumulation of sugars (°Brix)and quercetin glycosides in Shiraz berry skin across threeexperimental seasons, 1999–2000, 2000–2001 and 2001–2002. (a)Sugar accumulation (°Brix). (b) Quercetin glycoside accumulation.

Page 8: Sintesis flavonoles

Downey, Harvey & Robinson Synthesis of flavonols & flavonol synthase gene expression 117

seasons studied here (Figure 4b). When examined on aper berry basis, total quercetin glycosides per berryincreased steadily from flowering/fruit-set through veraison towards harvest. In addition, there was a signif-icant increase in flavonol accumulation commencing ataround two to six weeks post-veraison.

Towards harvest, there was some variation in the levels of flavonols between seasons. When considered atsimilar levels of ripeness (24°Brix), the highest level offlavonols per berry was observed in the first season(1999–2000), while the lowest level was observed in thefinal season (2001–2002). However, it should be notedthat berry weight at this time was also highest in the firstseason and lowest in the third season (Table 1).

In the two seasons where skin was collected (2000–2001 and 2001–2002), the level of flavonols (mg/g freshweight of skin) was higher in 2000–2001 than in the following season. To some extent this is reflected in thelevel of flavonols per berry (Table 1), but it should also benoted that in the earlier season skin weight as a per-centage of berry weight was lower (30.1%) than in2001–2002 (37.9%).

Flavonol accumulation and FLS expression in other tissues ofShiraz Flavonols are also present in other plant tissues so theflavonol content and composition of different grapevinetissues were determined together with expression ofVvFLS1. Figure 5 shows the comparative levels offlavonols present in a range of different tissues of thegrapevine and expression of VvFLS1 in the same tissues.Immediately after bud-burst, at the beginning of thegrowing season for grapevines (E-L stage 4), a sample ofthe emerging buds was taken with successive develop-mental stages of young shoots made at weekly intervals.All of these tissues had high levels of quercetin glycosides,but kaempferol glucoside was not detected. The develop-ing inflorescence (E-L stage 13) had the highest flavonolconcentration of any stage in berry development(2.111 ± 0.063 mg/g) and, accordantly, the highest levelof VvFLS1 expression (Figure 5b). The decline observed inthe level of quercetin glycosides in the developing fruit(Figure 3b) was more dramatic when the initial level inthe inflorescence is considered. A high level of flavonolswas also detected in the anthers of Shiraz flowers(1.200 ± 0.025 mg/g). Compared with the vegetative tissues, the levels of quercetin derivatives in the berrythree weeks post-veraison and at harvest, were low(Figure 5a). There were no flavonols detected in eitherthe seeds or the flesh of Shiraz berries sampled threeweeks post-veraison, although a trace of VvFLS1 expres-sion could be detected in flesh, while there were sig-nificant levels in the skin throughout development asdescribed above. Grapevine tendrils and the pedicel of theberry also contained high levels of flavonols, withkaempferol glucoside also detected in these tissues,around 5% of total flavonols in tendril and 2% in thepedicel. The level of flavonols in newly opened buds, tendrils, pedicels and developing inflorescences was significantly higher than at any time during berry

development from fruit-set until harvest. Although therewere clearly detectable levels of VvFLS1 in emergingbuds, tendrils and developing flowers, very little VvFLS1expression was detected in the pedicel or root, despite thepresence of flavonols in these tissues. There was nodetectable expression of VvFLS2 in any of the vegetative tissues analysed here.

Shiraz leaves had a high concentration of flavonols,around 2.5 times greater than that observed in the flowers.In berries, almost all of the flavonols were quercetin glycosides, but Shiraz leaves contained both quercetinand kaempferol glucosides (Figure 5a). In leaves, approximately 10% of total flavonols were kaempferol glucoside. A high level of flavonols was observed innewly emergent Shiraz leaves (quercetin glycoside3.160 ± 0.270 mg/g; kaempferol glycoside 0.390 ± 0.040mg/g fresh weight). As the leaf expanded this leveldeclined as seen in the second leaf stage (Figure 5a). Thelevel of flavonols increased again as the leaf continued to

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Buds

Tend

ril

Pedi

cel

Inflo

resc

ence

Flow

er

Anth

ers

Post

-Ver

aiso

n Se

ed

Post

-Ver

aiso

n Fl

esh

Post

-Ver

aiso

n Sk

in

Harv

est S

kin

Leaf

Sta

ge #

1

Leaf

Sta

ge #

2

Leaf

Sta

ge #

3

Leaf

Sta

ge #

4

Leaf

Sta

ge #

5

Root

Flav

onol

s (m

g/g)

none

det

ecte

d

none

det

ecte

d

a

0.0

0.2

0.4

0.6

0.8

1.0

1.2

Buds

Tend

ril

Pedi

cel

Inflo

resc

ence

Flow

er

Anth

ers

Post

-Ver

aiso

n Se

ed

Post

-Ver

aiso

n Fl

esh

Post

-Ver

aiso

n Sk

in

Harv

est S

kin

Leaf

Sta

ge #

1

Leaf

Sta

ge #

2

Leaf

Sta

ge #

3

Leaf

Sta

ge #

4

Leaf

Sta

ge #

5

Root

VvFL

S1 e

xpre

ssio

n

not d

eter

min

ed

none

det

ecte

d

b

Figure 5. Flavonol concentration and expression of a gene encodingflavonol synthase in tissues of Vitis vinifera L. cv. Shiraz. (a) Levelsof quercetin (�� ) and kaempferol (� ) glucosides in Shiraz tissues(mg/g fresh weight of berry ±SEM). (b) Expression of VvFLS1relative to the expression of VvUbiquitin1 in grapevine tissues. Note:Expression of VvFLS1 in anthers was not determined due to difficultyisolating RNA from that tissue. Very low levels of VvFLS1 expressionwere detected in pedicel (0.005) and post-veraison seed (0.001).No expression of VvFLS1 was detected in grapevine roots. Post-veraison samples: 01 February 2001 (3 weeks post-veraison);Harvest: 13 March 2001 (9 weeks post-veraison). Developmentalstages defined according to the Eichhorn-Lorenz (E-L) system(Coombe 1995), buds (E-L 4), expanding inflorescence (E-L 13; 14weeks pre-veraison), flowers pre-capfall (E-L 17; 10 weeks pre-veraison)

Page 9: Sintesis flavonoles

118 Synthesis of flavonols & flavonol synthase gene expressionAustralian Journal of Grape and Wine Research 9, 110–121, 2003

expand reaching a level similar to that in the youngestleaf. However, the level of kaempferol glucoside in theolder leaves was lower than that observed in theyoungest leaf. Roots from glasshouse grown Shiraz vinescontained low levels of flavonols compared to other vegetative tissues, but similar levels to the skin (0.069 ±0.004 mg/g).

The VvFLS1, but not VvFLS2, mRNA was detected atall stages of leaf development with the transcript becom-ing gradually more abundant as the leaves matured. Thislevel of transcript accumulation was commensurate withthe ongoing accumulation of flavonols in the expandingleaf.

DiscussionFlavonols and anthocyanins are predominantly syn-thesised in the skin of grape berries whereas condensedtannins are present in both the seeds and skin (Mazzaand Miniati 1993, Souquet et al. 1996, Downey et al.2003). No flavonols were detected in the seeds, nor in theflesh, of Shiraz grapes (Figure 5) suggesting that flavonolswere only present in the skin, which was consistent withprevious reports (Souquet et al. 1996).

In Vitis vinifera L. cv. Shiraz grapes at harvest overthree seasons, the major flavonols detected werequercetin-3-glucoside and quercetin-3-glucuronide(Figure 1). In Chardonnay grapes the major flavonolswere also quercetin glycosides (data not shown). In neither Chardonnay or Shiraz were flavonols other thanquercetin glycosides detected at harvest, althoughkaempferol glucosides were detected in Shiraz at flower-ing. We did not detect the flavonol aglycones quercetinand kaempferol in any of the grape berry samplesanalysed. Ribereau-Gayon (1964) reported the mainflavonols in grapes as being the 3-glucosides of myricetin,quercetin and kaempferol, as well as the 3-glucuronide ofquercetin. In addition to these, Cheynier and Rigaud(1986) also reported the glucuronides of kaempferol andmyricetin, the glucosylgalactoside and glucosylxyloside ofquercetin, and the galactoside and glucosylarabinoside ofkaempferol, as well as the 3-glucoside of isorhamnetin, a3’-methylated quercetin derivative. Of these compounds,the glucoside and glucuronide of quercetin representedby far the greater proportion of flavonols in the fruit ofVitis vinifera L. cv. Cinsault (Cheynier and Rigaud 1986).Macheix et al. (1990) later quantified the level ofquercetin derivatives in grape berries, reporting levels torange from less than 0.01 to 0.1 mg/g fresh weight ofskin. Price et al. (1995) also reported quercetin glucosideand glucuronide as the largest flavonol components ingrape skin, with around 0.065 mg/g fresh weight of skinin Vitis vinifera L. cv. Pinot Noir at harvest (23°Brix).Subsequently, Haselgrove et al. (2000) reported levels of around 0.065 mg/g fresh weight of berry (0.056mg/berry) in Shiraz at harvest (24°Brix).

The results presented here for Chardonnay and Shirazberries at commercial ripeness are comparable with previous reports. In the 1999–2000 season, the level ofquercetin glycosides in Shiraz (0.034 ± 0.001 mg/g ofberry fresh weight; 0.046 ± 0.001 mg/berry; n = 3) and

Chardonnay (0.031 ± 0.003 mg/g of berry fresh weight;0.047 ± 0.005 mg/berry; n = 3) were not significantly different on either a per gram (p = 0.133) or a per berry(p = 0.478) basis. These levels are slightly lower thanthose reported by Haselgrove et al. (2000) for wholeShiraz berries. The level of flavonols in Shiraz skin rangedfrom 0.071–0.086 mg/g fresh weight of skin at approxi-mately 24°Brix, which was slightly higher than reportedby Price et al. (1995) in Pinot Noir, but falling within thegeneral range reported by Macheix et al. (1990). Overthree seasons, there were similar levels of flavonol glyco-sides in Shiraz berries (mg/berry) at a similar level ofripeness (24°Brix, Table 1).

In both Chardonnay and Shiraz berries, similar patternsof accumulation were observed. High flavonol concen-trations (mg/g of berry fresh weight) were recordedaround flowering, followed by a decrease as the grapeberry increased in size. However, after the initial decreasefollowing flowering and fruit-set, the per gram level ofquercetin glycosides in the fruit remained constant untilharvest despite substantial increases in berry size. Thisindicated ongoing biosynthesis of flavonols during berrydevelopment and this was confirmed by the increase inflavonols per berry. The pattern of flavonol accumulationper berry suggests two main periods of flavonol synthesis,the first occurring around flowering and the secondoccurring after veraison. The post-veraison phase repre-sented the greatest increase in flavonols per berry, with asubstantial increase observed in all three seasons com-mencing three to four weeks post-veraison (Figure 4).Despite minor variations in timing and amplitude, thesimilarity across seasons establishes a convincing patternfor flavonol accumulation in grape berries that has notpreviously been reported. The increase observed in thefinal weeks of berry development was not observed byHaselgrove et al. (2000), although McDonald et al. (1998)reported that the levels of flavonols were higher in winesmade from very ripe fruit, which is consistent with what we have observed here. Towards the end of berrydevelopment there was a notable increase in flavonolcontent during the final weeks of ripening. The post-veraison increase in flavonols occurred after the mainperiod of anthocyanin biosynthesis, which occurred inthe 2–3 weeks immediately post-veraison (data notshown). This increase in flavonols post-veraison might berelated to copigmentation or to the role of flavonols asUV protectants, or involved in preventing the photo-bleaching of anthocyanins reported by Yamasaki et al.(1996).

At flowering, significant amounts of flavonols werepresent in both Chardonnay and Shiraz grapes. This ledus to examine the flavonol content in developing in-florescences and other vegetative tissues of Vitis vinifera L.cv. Shiraz. The level of flavonols in flowers was 10–15times higher than that observed in the skin of ripegrapes, but was highest in the developing inflorescence(Figure 5a). Such high levels of flavonols would be con-sistent with a role for flavonols in UV protection and ofparticular importance in reproductive tissues (Lodish etal. 1995). High levels of flavonols were also observed in

Page 10: Sintesis flavonoles

Downey, Harvey & Robinson Synthesis of flavonols & flavonol synthase gene expression 119

the anthers collected from Shiraz flowers and this mightaccount for a considerable portion of the flavonols in theflower. The high level of quercetin glycosides in antherssuggests that flavonols might also play a role in grapevinepollen fertility as reported in tobacco and petunia (Ylstraet al. 1992, Ylstra et al. 1994).

The bunch rachis of Vitis vinifera cv. Merlot has previously been shown to contain primarily quercetin-3-glucoside, around 0.2 mg/g fresh weight, with quercetinglucuronide less than 10% of that amount and smallamounts of the myricetin and kaempferol glycosides(Souquet et al. 2000). While the relative proportionsobserved here are similar to those reported in Merlotstems by Souquet et al. (2000), the level of flavonols inShiraz pedicels was around two-fold greater (0.43 mg/g)and in tendrils, six-fold higher (1.23 mg/g). Given theproposed role of these compounds in UV protection, theEuropean (France) location of the previous work mightwell account for lower flavonol content compared tovines grown in South Australia.

Previously, the flavonol composition of grapevineleaves has also been studied. Mature grapevine leavesgrowing in Morocco were shown to contain all threeflavonols, myricetin, quercetin and kaempferol, withquercetin derivatives accounting for 80–100% of theflavonol content (Hmamouchi et al. 1996). Leaves of thefour Vitis cultivars, Alicante, Carignan, Cinsault andGrenache Noir, were examined with leaves of the lattertwo containing only trace amounts of myricetin andkaempferol. Total flavonols in grapevine leaves rangedfrom 8.1 mg/g dry weight in Grenache Noir, to 38.6 mg/gdry weight in Alicante (Hmamouchi et al. 1996). InShiraz, we observed both quercetin and kaempferolglycosides with total flavonols around 3.5 mg/g freshweight (Figure 5a). Based on a water content of grapeleaves of 90% or greater (Patakas et al., 1997), the levelsreported here in Shiraz are comparable with those report-ed in cultivars grown in Morocco.

The flavonol synthase (FLS) enzyme forms a branch-point in the flavonoid pathway and genes encoding FLShave now been isolated from a number of plants. Holtonet al. (1993) confirmed the function of the petunia FLSgene by expressing it in yeast and demonstrating thatextracts of the yeast catalysed the conversion of dihydro-quercetin to quercetin in the presence of the appropriatecofactors. The grape cDNAs VvFLS1 and VvFLS2 both hadhigh sequence homology to the petunia and other FLSgenes indicating that they most likely encode FLS ingrapevine. Previously, expression of genes encoding FLShas been detected in flowers but there have been fewstudies of flavonol synthesis in fruit. The petunia FLSgene was highly expressed in early developmental stagesof flowers, but was not detected in leaves of petunia(Holton et al. 1993). In potato, van Eldik et al. (1997)also found FLS expressed during flower development,particularly in the pistil, anther and petal tissues.Moriguchi et al. (2002) detected citrus FLS expression inflowers and young leaves but also in the juice sacs andpeel of the fruit throughout fruit development. While little expression of VvFLS2 was detected in any grapevine

tissues, VvFLS1 was expressed in the buds, leaves andtendrils of the grapevine as well as in the flowers, devel-oping fruit and the skin of ripening berries of bothChardonnay and Shiraz.

The pattern of FLS gene expression in bothChardonnay and Shiraz berries between flowering andveraison was commensurate with the pattern of flavonolaccumulation in the berry during that time. Expression ofthe VvFLS1 gene at a high level at flowering suggests aperiod of intense flavonol biosynthesis, coinciding withthe high concentrations of flavonols detected in these tissues. The concentration of flavonols was observed todecrease substantially after flowering, due to a 20-foldincrease in berry size by weight, and VvFLS1 expressionalso decreased after flowering.

The concentration of flavonols (mg/g fresh weight)was relatively constant after veraison, but the amount offlavonols (mg/berry) increased, suggesting continuedsynthesis during the second phase of berry growth.Expression of VvFLS1 increased around veraison andparticularly during the period from two weeks post-veraison until ripening was complete. This expression ofVvFLS1 during ripening coincided with the increase inflavonols per berry in Shiraz (Figure 3d), although thiswas less noticeable in ripening Chardonnay berries(Figure 2d).

The level of flavonols in newly opened buds, tendrils,pedicels and developing inflorescences of Shiraz was significantly higher than at any time during berry devel-opment from fruit-set until harvest. Although there wereclearly detectable levels of VvFLS1 in emerging buds, ten-drils and developing flowers, very little VvFLS1 expressionwas detected in the pedicel or root. No expression ofVvFLS2 was detected in these or any other vegetative tissue and it seems most likely that at the time of sampling, flavonol synthesis was complete in the grapepedicel and the root. Therefore, to determine expressionof VvFLS1 in these tissues, samples would need to betaken at different times during their development. Analternative, albeit less likely, explanation for the absenceof VvFLS1 expression in these tissues is that another FLSgene exists in grapevines that we are yet to identify andthat expression of this gene accounts for flavonol syn-thesis in these tissues. This seems less likely as we wereunable to amplify additional gene homologues by PCRwith a range of degenerate primers and only two bandshomologous to the VvFLS1 and VvFLS2 probes weredetected by Southern blot analysis.

Studies of anthocyanin biosynthesis in grapevinesshowed that genes encoding biosynthetic enzymes formuch of the anthocyanin pathway were expressed post-veraison when anthocyanin synthesis occurs, but alsoearly in berry development (Boss et al. 1996a). This wasconsidered particularly interesting, as expression of thepathway was not anticipated pre-veraison when antho-cyanins were not being produced. Furthermore, Boss etal. (1996b) observed that much of the pathway was alsoexpressed in other grapevine tissues that did not normal-ly accumulate anthocyanins. An explanation consistentwith these observations would be that flavonoid com-

Page 11: Sintesis flavonoles

120 Synthesis of flavonols & flavonol synthase gene expressionAustralian Journal of Grape and Wine Research 9, 110–121, 2003

pounds other than anthocyanins were being synthesisedin these tissues. We have previously shown substantialquantities of flavan-3-ols and condensed tannins (pro-anthocyanidins) to be present in Shiraz berries from theearliest phase of berry development and increase towardsveraison (Downey et al. 2003). It is now clear that a significant level of flavonol biosynthesis is also occurringduring berry development. Together, the spatial and tem-poral pattern of flavonol and tannin accumulation ingrapes and expression of the FLS gene form the strongestindication yet that the expression of genes involved inflavonoid biosynthesis reported by Boss et al. (1996a,1996b) is related to flavonol and condensed tanninbiosynthesis in grapevines.

To date, this is the most comprehensive study offlavonols in grapevines in that it has quantified flavonolsthroughout berry development. The pervasive nature offlavonols in the grapevine, and particularly the veryhigh levels in juvenile and reproductive tissue, as well asthe high levels in leaves are strong indications that therole of grapevine flavonols is UV protection. Such beingthe case, and the correlation between sun exposure andflavonol content reported by Price et al. (1995) andHaselgrove et al. (2000), viticultural practices thatincrease the sun exposure might significantly increase theflavonol content of the ripe fruit and the subsequentcopigment pool in the wine.

The level of flavonols in grapes is relatively low compared to the anthocyanins and other potential co-pigments such as proanthocyanidins, with flavonols comprising as little as one to ten percent of totalflavonoids depending on variety, site and season. The roleof flavonoids in contributing to colour stability and theorganoleptic properties of wine has yet to be established.However, our studies suggest that two periods of flavonolsynthesis occur during berry development, around flow-ering and in the latter phases of ripening. Manipulationof conditions in the vineyard at those times, such asbunch exposure to light, may have significant effects onflavonol synthesis and wine quality.

AcknowledgmentsThe authors would like to thank John and Di Harvey ofSlate Creek Vineyard, for generously providing grapesamples for this research. We would also like to expressour appreciation to Karin Sefton for her continued assis-tance and good humour. Thanks also to Dr GrahamJones of Adelaide University for providing the MerckChromolith columns used in this analysis. This researchwas funded by an Australian Post-graduate Award,CSIRO Plant Industry, the Grape and Wine Research andDevelopment Corporation. This project was supported bythe Commonwealth Cooperative Research CentresProgram and conducted by the CRC for Viticulture.

References Asen, S., Stewart, R.N. and Norris, K.H. (1972) Co-pigmentation of

anthocyanins in plant tissues and its effect on color. Phytochemistry11, 1139–1144.

Asen, S. (1975) Factors affecting flower colour. Acta Horticulturae 41,57–68.

Baranac, J.M., Petranovic̆, N.A. and Dimitric̆-Markovic̆, J.M. (1997)Spectroscopic study of anthocyanin copigmentation reactions. 2.Malvin and the nonglycosidized flavone quercetin. Journal ofAgricultural and Food Chemistry 45, 1694–1697.

Beggs, C.J., Kuhn, K., Böcker, R. and Wellmann, E. (1987)Phytochrome-induced flavonoid biosynthesis in mustard (Sinapisalba L.) cotyledons. Enzymic control and differential regulation ofanthocyanin and quercetin formation. Planta 172, 121–126.

Boss, P.K., Davies, C. and Robinson, S.P. (1996a) Analysis of theexpression of anthocyanin pathway genes in developing Vitisvinifera L. cv. Shiraz grape berries and the implications for pathwayregulation. Plant Physiology 111, 1059–1066.

Boss, P.K., Davies, C. and Robinson, S.P. (1996b) Expression ofanthocyanin biosynthesis pathway genes in red and white grapes.Plant Molecular Biology 32, 565–569.

Boulton, R. (2001) The copigmentation of anthocyanins and its rolein the color of red wine: a critical review. American Journal ofEnology and Viticulture 52, 67–87.

Brouillard, R. and Dangles, O. (1994) Anthocyanin molecular inter-actions: the first step in the formation of new pigments during wineaging? Food Chemistry 51, 365–371.

Cheynier, V. and Rigaud, J. (1986) HPLC separation and characteri-zation of flavonols in the skins of Vitis vinifera var. Cinsault.American Journal of Enology and Viticulture 37, 248–252.

Coombe, B.G. (1995) Adoption of a system for identifying grapevinegrowth stages. Australian Journal of Grape and Wine Research 1,104–110.

Darné, G. (1993) New hypotheses on anthocyanin biosynthesis inberries and leaves of the grapevine. Vitis 32, 77–85.

Downey, M.O., Harvey, J.S. and Robinson, S.P. (2003) Analysis oftannins in seeds and skins of Shiraz grapes throughout berrydevelopment. Australian Journal of Grape and Wine Research 9,15–27.

Flint, S.D., Jordan, P.W. and Caldwell, M.M. (1985) Plant protectiveresponse to enhanced UV-B radiation under field conditions: leafoptical properties and photosynthesis. Photochemistry andPhotobiology 41, 95–99.

Frohman, M.A., Dush, M.K. and Martin, G.R. (1988) Rapid pro-duction of full-length cDNAs from rare transcripts: amplificationusing a single gene-specific oligonucleotide primer. Proceedings ofthe National Academy of Sciences of the United States of America85, 8998–9002.

Haselgrove, L., Botting, D., van Heeswijck, R., Høj, P.B., Dry, P.R.,Ford, C. and Iland, P.G. (2000) Canopy microclimate and berrycomposition: the effect of bunch exposure on the phenolic compo-sition of Vitis vinifera L cv. Shiraz grape berries. Australian Journal ofGrape and Wine Research 6, 141–149.

Hmamouchi, M., Es-Safi, N., Lahrichi, M., Fruchier, A. and Essassi,E.M. (1996) Flavones and flavonols in leaves of some MoroccanVitis vinifera cultivars. American Journal of Enology and Viticulture47, 186–192.

Holton, T.A., Brugliera, F. and Tanaka, Y. (1993) Cloning and expres-sion of flavonol synthase from Petunia hybrida. Plant Journal 4,1003–1010.

Hoshino, T. Matsumoto, U. and Goto, T. (1980) The stabilizing effectof the acyl group on the co-pigmentation of acylated anthocyaninswith C-glucosylflavones. Phytochemistry 19, 663–667.

Hrazdina, G. and Moskowitz, A.H. (1980) Subcellular status ofanthocyanins in grape skins. In: Proceedings University ofCalifornia, Davis, Grape and Wine Centennial Symposium. Ed. A.D.Webb (University of California Press, Berkeley) pp. 245–253.

Koes, R.E., Quattrocchio, F. and Mol, J.N.M. (1994) The flavonoidbiosynthetic pathway in plants: function and evolution. BioEssays16, 123–132.

Lambert, S.G. (2002) Copigmentation and its impact on the stabili-sation of red wine pigments. PhD Thesis (University of Adelaide:Adelaide).

Lodish, H., Baltimore, D., Berk, A., Zipursky, S.L., Matsudaira, P. andDarnell, J. (1995) ‘Molecular cell biology’. (W. H. Freeman and Co.:New York)

Macheix, J.J., Fleuriet, A. and Billot, J. (1990) ‘Fruit phenolics’. (CRCPress Inc.: Boca Raton, FA)

Page 12: Sintesis flavonoles

Downey, Harvey & Robinson Synthesis of flavonols & flavonol synthase gene expression 121

Markham, K.R., Tanner, G.J., Caasi-Lit, M., Whitecross, M.I.,Nayudu, M. and Mitchell, K.A. (1998) Possible protective role for3’,4’-dihydroxyflavones induced by enhanced UV-B in a UV-tolerant rice cultivar. Phytochemistry 49, 1913–1919.

Mateus, N., Silva, A.M.S., Santos-Buelga, C., Rivas-Gonzalo, J.C. andde Freitas, V. (2002) Identification of anthocyanin-flavanol pig-ments in red wines by NMR and mass spectrometry. Journal ofAgricultural and Food Chemistry 50, 2110–2116.

Mazza, G. and Miniati, E. (1993) Grapes. In: ‘Anthocyanins in fruits,vegetables, and grains’. Eds G. Mazza and E. Miniati, (CRC Press:Boca Raton, FA) pp. 149–199.

McCarthy, M.G. (1999) Weight loss from ripening berries of Shirazgrapevines (Vitis vinifera L. cv. Shiraz). Australian Journal of Grapeand Wine Research 5, 10–16.

McCloskey, L.P. and Yengoyan, L.S. (1981) Analysis of anthocyaninsin Vitis vinifera wines and red color versus aging by HPLC and spectrometry. American Journal of Enology and Viticulture 32,257–261.

McDonald, M.S., Hughes, M., Burns, J., Lean, M.E.J., Matthews, D.and Crozier, A. (1998) Survey of the free and conjugated myricetinand quercetin content of red wines of different geographical origins.Journal of Agricultural and Food Chemistry 46, 368–375.

Moriguchi, T., Kita, M., Ogawa, K., Tomono, Y., Endo, T. and Omura,M. (2002) Flavonol synthase gene expression during citrus fruitdevelopment. Physiologia Plantarum 114, 251–258.

Moskowitz, A.H. and Hrazdina, G. (1981) Vacuolar contents of fruitsubepidermal cells from Vitis species. Plant Physiology 68, 686–692.

Nagel, C.W. and Wulf, L.W. (1979) Changes in the anthocyanins,flavonoids and hydroxycinnamic acid esters during fermentationand aging of Merlot and Cabernet Sauvignon. American Journal ofEnology and Viticulture 30, 111–116.

Osawa, Y. (1982) Copigmentation of anthocyanins. In:‘Anthocyanins as food colours.’ Ed. P. Markakis. (Academic Press:New York) pp. 41–68.

Patakas, A., Noitsakis, B. and Stavrakas, D. (1997) Water relationparameters in Vitis vinifera L. in drought period. Effects of leaf age.Agronomie 17, 129–138.

Pelletier, M.K., Murrell, J.R. and Shirley, B.W. (1997) Characterizationof flavonol synthase and leucoanthocyanidin dioxygenase genes inArabidopsis. Further evidence for differential regulation of ‘early’and ‘late’ genes. Plant Physiology 113, 1437–1445.

Price, S.F., Breen, P.J., Valladao, M. and Watson, B.T. (1995) Clustersun exposure and quercetin in Pinot noir grapes and wine.American Journal of Enology and Viticulture 46, 187–194.

Ribereau-Gayon, P. (1964) Les composés phénoliques du raisin et duvin. Institut National de la Recherche Agronomique, Paris.

Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) ‘Molecular cloning:a laboratory manual’. (Cold Spring Harbour Laboratory Press: NewYork)

Scheffeldt, P. and Hrazdina, G. (1978) Co-pigmentation of antho-cyanins under physiological conditions. Journal of Food Science 43,517–520.

Smith, G.J. and Markham, K.R. (1998) Tautomerism of flavonol glu-cosides: relevance to plant UV protection and flower colour. Journalof Photochemistry and Photobiology A: Chemistry 118, 99–105.

Somers, T.C. and Evans, M.E. (1979) Grape pigment phenomena:interpretation of major colour losses during vinification. Journal ofthe Science of Food and Agriculture 30, 623–633.

Somers, T.C. and Vérette, E. (1988) Phenolic composition of naturalwine types. In: ‘Wine analysis’. Eds H.-F. Linskens and J.F. Jackson(Springer-Verlag: Berlin) pp. 219–257.

Souquet, J.-M., Cheynier, V., Sarni-Manchado, P. and Moutounet, M.(1996) Les composés phénoliques du raisin. Journal Internationaldes Sciences de la Vigne et du Vin Special Issue, 99–107

Souquet, J-M., Labarbe, B., le Guernevé, C., Cheynier, V. andMoutounet, M. (2000) Phenolic composition of grape stems.Journal of Agricultural and Food Chemistry 48, 1076–1080.

Sparvoli, F., Martin, C., Scienza, A., Gavazzi, G. and Tonelli, C. (1994)Cloning and molecular analysis of structural genes involved inflavonoid and stilbene biosynthesis in grape (Vitis vinifera L.). PlantMolecular Biology 24, 743–755.

Spribille, R. and Forkmann, G. (1984) Conversion of dihydro-flavonols to flavonols with enzyme extracts from flower buds ofMatthiola incana R. Br. Zeitschrift für Naturforschung 39c, 714–719.

Stafford, H.A. (1990) ‘Flavonoid metabolism’. (CRC Press, Inc.: BocaRaton, FA)

Taylor, L.P. (1995) Flavonols: effects on fertility and fecundity. CropScience 35, 1521–1526.

Thomas, M.R., Matsumoto, S., Cain, P. and Scott, N.S. (1993)Repetitive DNA of grapevine: classes present and sequences suitablefor cultivation identification. Theoretical and Applied Genetics 86,173–180.

van Eldik, G.J., Reijnen, W.H., Ruiter, R.K., van Herpen, M.M.A.,Schrauwen, J.A.M. and Wullems, G.J. (1997) Regulation offlavonol biosynthesis during anther and pistil development, andduring pollen tube growth in Solanum tuberosum. Plant Journal 11,105–113.

Wan, C-Y. and Wilkins, T. (1994) A modified hot borate method sig-nificantly enhances the yield of high-quality RNA from cotton(Gossypium hirsutum L.). Analytical Biochemistry 223, 7–12.

Yamasaki, H., Uefuji, H. and Sakihama, Y. (1996) Bleaching of the red anthocyanin induced by superoxide radical. Archives ofBiochemistry and Biophysics 332, 183–186.

Ylstra, B., Busscher, J., Franken, J., Hollman, P.C.H., Mol, J.N.M. andvan Tunen, A.J. (1994) Flavonols and fertilization in Petunia hybrida:localization and mode of action during pollen tube growth. PlantJournal 6, 201–212.

Ylstra, B., Touraev, A., Moreno, R.M.B., Stöger, E., van Tunen, A.J.,Vicente, O., Mol, J.N.M. and Heberle-Bors, E. (1992) Flavonolsstimulate development, germination and tube growth of tobaccopollen. Plant Physiology 100, 902–907.

Manuscript received: 18 December 2002Revised manuscript received: 26 March 2003