ORIGINAL ARTICLE

Molecular cloning and biochemical characterization of threeConcord grape (Vitis labrusca) flavonol 7-O-glucosyltransferases

Dawn Hall • Kyung Hee Kim • Vincenzo De Luca

Received: 18 May 2011 / Accepted: 24 June 2011 / Published online: 14 July 2011

� Springer-Verlag 2011

Abstract Grapes berries produce and accumulate many

reactive secondary metabolites, and encounter a wide range

of pathogen- and human-derived xenobiotic compounds.

The enzymatic glucosylation of these metabolites changes

their reactivity, stability and subcellular location. Two ESTs

with more than 90% nucleotide sequence identity to three

full-length glucosyltransferases are expressed in several

grape tissues. The full-length clones have more than 60%

amino acid sequence similarity to previously characterized

flavonoid 7-O-glucosyltransferases, catechin O-glucos-

yltransferases and anthocyanin 5-O-glucosyltransferases.

In vitro, these enzymes glucosylate flavonols and the

xenobiotic 2,4,5-trichlorophenol (TCP). Kinetic analysis

indicates that TCP is the preferred substrate for these

enzymes, while expression analysis reveals variable tran-

scription of these genes in grape leaves, flowers and berry

tissues. The in vivo role of these Vitis labrusca glucos-

yltransferases is discussed.

Keywords Concord grape � Detoxification mechanism �Flavonol � Glucosyltransferase � Xenobiotic

Abbreviations

AF After flowering

DFCI Dana Farber Cancer Institute

GTs Glucosyltransferases

PMSF Phenylmethyl sulfonyl fluoride

PSPG Plant secondary product

glucosyltransferases

PVP-10 or PVPP Polyvinylpyrrolidone

Q3G Quercetin 3-O-glucoside

TCP Trichlorophenol

UDPG Uridine 50-diphosphoglucose

VLOGT Vitis labrusca O-glucosyltransferase

VLRSGT Vitis labrusca resveratrol

glucosyltransferase

Introduction

Plants encounter a wide range of external stimuli and envi-

ronmental conditions that trigger the production of diverse

secondary metabolites. These compounds accumulate within

the vacuole, in specialized cell types, or outside the plant in

cell walls, on the leaf or in the rhizosphere surrounding plant

roots. Enzymatic modification by hydroxylation, methylation,

acylation and glycosylation changes the physiological and

biochemical properties of these molecules and alters their role

and location within the plant.

Grapes have been cultivated for thousands of years for

the production of wine, and they remain culturally, eco-

nomically and agronomically important in many areas

around the world. The grape berry can be divided into the

seed, mesocarp (flesh), and exocarp (skin) tissues that

differ in their enzymatic and metabolite profiles throughout

development. A key stage of grape berry development,

Electronic supplementary material The online version of thisarticle (doi:10.1007/s00425-011-1474-0) contains supplementarymaterial, which is available to authorized users.

D. Hall � K. H. Kim � V. De Luca (&)

Department of Biological Sciences, Brock University,

500 Glenridge Ave, St. Catharines, ON L2S 3A1, Canada

e-mail: vdeluca@brocku.ca

Present Address:D. Hall

Michael Smith Laboratories, The University of British

Columbia, #301, 2185 East Mall, Vancouver, BC V6T 1Z4,

Canada

123

Planta (2011) 234:1201–1214

DOI 10.1007/s00425-011-1474-0

veraison, is characterized by berry softening, the rapid accu-

mulation of total soluble solids (�Brix) and the accumulation

of anthocyanins in the exocarp to produce red/blue colors.

Many of the berry-derived compounds contribute to the

organoleptic characteristics (color, flavour, astringency) of

wine (Polaskova et al. 2008), and their regular consumption is

associated with numerous human health benefits (Leifert and

Abeywardena 2008; Huntley 2007).

Glucosylated metabolites are detected in many grape-

vine tissues. Grape berries accumulate the 3-O-mono-

glucosides and the 3,5-O-diglucosides of the colorful

anthocyanins (Munoz-Espada et al. 2004; Wu and Prior

2005), the 3-O- and 7-O-glucosides of the UV-protectant

flavonols (Castillo-Munoz et al. 2007, 2009) and the 3-O/

40-O-glucosides of health beneficial stilbenes (Ali and

Strommer 2003). Glucosylated anthocyanins, flavonols

and stilbenes have also been identified in grape leaves and

flowers (Langcake and Pryce 1977; Moore and Giannasi

1994; Schneider et al. 2008).

Glucosyltransferases (GTs) catalyze the transfer of

glucose from uridine 50- diphosphoglucose (UDPG) to

acceptor molecules that include plant-derived and xenobi-

otic metabolites (Vogt and Jones 2000). In plants, GTs

exist as large, multigene families (Bowles 2002) and the

Arabidopsis genome currently has 121 Group 1 plant sec-

ondary product glucosyltransferases (PSPG; CaZy—

http://www.cazy.org/) which can be subdivided into 12

classes (A–N) (Lim et al. 2003). Although more than 150

GTs have been cloned from a number of plant species, only

4 GTs have been functionally identified in grape. These

Vitis GTs catalyze the glucosylation of the 3-O- position of

anthocyanins and flavonols (Ford et al. 1998), the 5-O-

positions of anthocyanin-3-O-glucosides (Janvary et al.

2009), and produce either the glucose esters of hydroxy-

cinnamic acids or stilbene diglucosides (Hall and De Luca

2007), however, GTs which glucosylate the 7-O-position of

flavonols have not been identified.

The present study details the homology-based molecular

cloning and biochemical characterization of three full-

length members of the 7GT-family from Vitis labrusca cv.

Concord that O-glucosylate flavonols, and a xenobiotic in

vitro. Substrate specificity, kinetic analysis, RT-PCR gene

expression studies, and in vivo enzyme activity profiles

suggest putative roles for these GTs in the modification of

xenobiotics and flavonols.

Materials and methods

Plant material

Grape berries (Vitis labrusca cv. Concord) were harvested

weekly from June until October, 2003–2006 at the 3rd St.

site of the G & H Wiley vineyard (St. Catharines, ON,

Canada). Vitis vinifera cv. Pinot Noir and Vitis vinifera cv.

Pinot Noir Droit were harvested from the Chateau des

Charmes Vineyard (Niagara on the Lake, ON, Canada) in

September of 2004. All stages of grape berry maturity are

reported as weeks after flowering (AF), and physiological

parameters of grape berry development were determined as

reported previously (Hall and De Luca 2007). Concord pre-

bloom flower buds were harvested June 12, 2006, grape

flowers were harvested June 16, 2006 and grapevine leaves

and stems were harvested June 21, 2006. All plant material

was stored at -80�C for future use. To separate grape berry

exocarp from mesocarp tissue, berries were briefly thawed;

the exocarp was removed by hand, scraped to remove

residual mesocarp tissue, and was quick frozen in liquid

nitrogen until future use.

Chemicals

All chemicals were purchased from Sigma (http://

www.sigmaaldrich.com) (Oakville, ON, Canada) or Indo-

fine Chemical (http://www.indofinechemical.com) (Hills-

borough, NJ, USA) and were prepared as reported previously

(Hall and De Luca 2007). Kaempferol, quercetin, and isorh-

amnetin 3-O-glucosides and quercetin 7-O-glucoside were

prepared enzymatically using the recombinant UDP-glucose:

flavonoid 3-O-glucosyltransferase (VL3GT) or the recombi-

nant UDP-glucose: resveratrol/hydroxycinnamic acid gluco-

syltransferase (VLRSGT), respectively (Hall and De Luca

2007).

Crude protein extraction and profiling of GT enzyme

activity throughout development

Vitis labrusca cv. Concord pre-bloom flower bud (1 g), leaf

(0.5 g), flower (1 g), week 1 whole berry (1 g), weeks 3, 6,

8, 10, 12, 14, and 16 AF and post-veraison (week 14 AF)

Vitis vinifera cv. Pinot Noir and cv. Pinot Noir Droit

exocarp (1 g) and mesocarp (2 g) tissue was ground to a

fine powder in a mortar and pestle with liquid nitrogen and

extracted in 4 ml of grape extraction buffer [500 mM Tris–

HCl, pH 8.0, 0.1% b-mercaptoethanol, 1% PVP-10 (poly-

vinylpyrrolidone), 5 mM sodium metabisulfite (Na2S2O5),

10% glycerol (v/v), 1 mM phenylmethyl sulfonyl fluoride

(PMSF), and 10% insoluble polyvinylpyrrolidone (PVPP)].

The extracts were centrifuged at 21,000g for 10 min at

4�C, and the supernatant was desalted on a PD-10 Sepha-

dex G-25 column (http://www.gehealthcare.com/caen), as

per the manufacturer’s instructions. For each enzyme

assay, 100 ll of protein was used with 100 lM quercetin

or cyanidin 3-O-glucoside and 9 mM UDPG in a final

reaction volume of 125 ll. The assays were incubated for

90 min at 30�C, stopped with 100% methanol, and

1202 Planta (2011) 234:1201–1214

123

processed prior to HPLC analysis as described below. All

specific activities are reported as picomoles of glucoside

produced in 1 s by 1 mg of total protein.

High performance liquid chromatography (HPLC)

analysis of enzyme assays

Enzyme assays were performed and analyzed by HPLC as

described previously (Hall and De Luca 2007). Addition-

ally, the O-glucosides of luteolin (350 nm), eriodictyol

(290 nm), myricetin (370 nm), isorhamnetin (370 nm),

dihydroquercetin (290 nm) and 3,6,20,30-tetrahydroxyflav-

one (320 nm) were monitored using a 2996 photodiode

array detector (Waters, Milford, MA, USA; http://www.

waters.com).

Flavonoid and anthocyanin analysis of mature grape

exocarp and mesocarp

The exocarp and mesocarp (with seeds removed) of mature

week 16 AF Concord grapes were harvested, dried, defatted

with 100% hexane and dried to produce 1 g dry weight of 3

biological replicates. Individual samples were extracted

three times in 10 ml of 15% acetic acid in methanol and the

combined supernatants from each extraction were centri-

fuged (7,250g, 4�C, 10 min) and the supernatants were

concentrated to dryness using a rotary evaporator at 40�C.

The acidic methanolic extracts were dissolved in 10 and

30 ml of water for mesocarp and exocarp fractions,

respectively, and then partitioned three times with 20 and

50 ml of ethylacetate for mesocarp and exocarp fractions,

respectively. After drying the combined organic fractions

using a rotary evaporator at 40�C, samples were reconsti-

tuted in 1 and 6 ml of 50% methanol for mesocarp and

exocarp fractions, respectively. The injection volume for

UPLC–MS analysis was 5 and 3 ll for mesocarp and exo-

carp fractions, respectively and analyses were performed

using an Aquity UPLC BEH C18 column with a particle size

of 1.7 lm and column dimensions of 1.0 9 50 mm and the

following parameters (Fig. S1, Table S1).

RNA extraction

RNA was extracted from grape berry exocarp and meso-

carp tissue as described previously (Hall and De Luca

2007), except that 1 ll of linear acrylamide (5 lg/ll) was

added to the RNA as a co-precipitant and the extracted

RNA was resuspended on ice. RNA (0.5–1 lg) was reverse

transcribed using either Superscript III reverse transcriptase

(Invitrogen) or the Takara RNA PCR kit version 3.0

(Fisher Scientific, http://www.fishersci.ca/) as per the

manufacturer’s instructions, and was diluted fivefold with

water. For RACE-ready cDNA, 5 lg of Vitis labrusca

week 12 AF berry exocarp RNA was extracted as above,

and prepared using the GeneRacer Kit (Invitrogen) as per

the manufacturer’s instructions.

Cloning of putative glucosyltransferase (GT) genes

from Vitis labrusca

At the time of cloning glucosyltransferases, the grape gen-

ome had yet to be released and 50 RACE was used to obtain

full length of glucosyltransferase sequences. A BLAST

search of Dana-Farber Cancer Institute (DFCI) Vitis vinifera

grape gene index EST database (http://compbio.dfci.

harvard.edu/tgi/cgi-bin/tgi/gimain.pl?gudb=grape) with the

(PSPG) box (WAPQVEVLAHPAVGCFVTHCGWNS-

TLESISAGVPMVAWPFFADQ) (Vogt and Jones 2000;

Jones and Vogt 2001) identified 141 expressed sequence tags

(ESTs) with sequence identity (E value \10-5) to GTs.

Three ESTs (TC75303, TC91700, NP9526770) with at least

60% sequence amino acid similarity to previously charac-

terized flavonoid GTs were identified, and their corre-

sponding full-length clones were obtained through two

rounds of 50 RACE.

Primers were designed to amplify the corresponding

full-length genes from Vitis labrusca week 12 AF exocarp-

enriched cDNA. VLOGT1 (EF533704) and VLOGT2

(EF533705) corresponding to the Vitis vinifera ESTs

NP9526770 and TC91700, respectively, were amplified

using FP3 (50 ATGGACAAACATCACTTCCTCTTAC 30)and RP6 (50 CTCAATGAGTCCTGACTTCTAAGC 30),and VLOGT3 (EF533606) corresponding to the EST

TC75303 was amplified with the primers followed by FP4

(50 GATCAGCACCCGGAAACGGAGAA 30) and RP14

(50 CTAAAAAATACTGCCTTGTCCCA 30). The prod-

ucts of these PCR reactions were cloned into the pGEM-T

easy TA-vector (Promega, http://www.promega.com) and

sequence verified.

Expression of recombinant VLOGT1, VLOGT2

and VLOGT3

Full-length VLOGT1, VLOGT2 and VLOGT3 were sub-

cloned into the pGEX-4T-1 GST-fusion protein expression

vector (GE Healthcare) and were transformed into Esche-

richia coli (E. coli) cells for recombinant protein expression.

Several E. coli strains, different induction conditions [(0.1 or

1 mM IPTG) (Bioshop, http://www.bioshopcanada.com)],

three growth temperatures following induction (16–18, 25,

or 37�C), and concurrent expression of the groES-groEL

chaperone proteins (to assist in GT folding) were tested

to obtain maximal amounts of soluble protein. For each

condition, soluble protein was extracted, assayed for GT

activity and 5–10 lg of protein was subjected to SDS-PAGE

Planta (2011) 234:1201–1214 1203

123

analysis to determine the optimal growth conditions for these

enzymes.

Expression in the pGEX-4T-1 N-terminal GST-fusion

vector in DE3 E. coli cells expressing chaperone protein,

induced with 0.1 mM IPTG and grown for 20 h at 16–18�C

was determined to yield the most active, soluble protein

and these conditions were used for all future analysis. The

soluble protein fraction was either affinity purified by

glutathione Sepharose 4B, or was desalted prior to use in

enzyme assays. One 50 ml bacterial culture produced

25–150 lg of purified recombinant (r)VLOGT2, 1–5 lg of

purified recombinant (r)VLOGT1 or 1–10 lg of purified

recombinant (r)VLOGT3. For this reason, only rVLOGT2

was purified prior to enzyme assays, while the desalted

soluble fraction was used in enzyme assays for rVLOGT3

and rVLOGT1 with the exception that rVLOGT3 was

purified prior to performing kinetic analyses.

RT-PCR analysis of VLOGT gene expression

Gene expression was monitored using the gene-specific

primers: VLOGT1,2—FP: 50 GCCCTTGGACACTATAT

GAATG 30, RP: 50 GGCGATGACTTTCCATCAGTC 30;VLOGT3—FP: 50 GGCGCTACAA GCTGTGGATAAGG

30, RP: 50 CTGCTGCTGTGTAGCAAAGC 30; VL3GT—

FP: 50TGCAGGGCCTAACTCACTCT 30, RP: 50 GCAGT

CGCCTTAGGTAGCAC 30; as well as the gene-specific

primers used in RT-PCR analysis for VLRSGT and VLActin

as reported previously (Hall and De Luca 2007).

VLOGT1,2; VLOGT3; and VLActin gene expression anal-

ysis was performed in duplicate and was quantified using

Multigauge ver 3.0 (Fujifilm, Tokyo, Japan; http://www.

fujifilm.ca), and the mean values were divided by the mean

actin gene expression value to obtain a relative value for

gene expression (% expression) in these tissues.

Template generated from RNA without reverse tran-

scription was used to check for genomic DNA contami-

nation. Tissue specificity of the RNA extracts was verified

using Vitis labrusca UDP-glucose: flavonoid 3-O-GT

(VL3GT) primers (which amplified a product from exocarp,

but not mesocarp tissue) and VLRSGT primers (which

amplified a product from mesocarp tissue but not exocarp

tissue) (Hall and De Luca 2007).

Gene specificity of the primers was verified by in-silico

analysis of the DFCI grape ESTs to the best of our

knowledge at the time (prior to release of the grape gen-

ome). Additionally, primer specificity was evaluated based

on amplification of PCR products from vectors containing

five Vitis labrusca GTs (VLOGT1, VLOGT2, VLOGT3,

VLRSGT, VL3GT). PCR using VLOGT1,2 and VLOGT3

gene-specific primers with identical amounts of plasmid

template and identical PCR cycling conditions generated

nearly identical amounts of amplified product.

In vivo biocatalysis of GT reactions

E. coli cultures expressing Vitis labrusca VLOGT1,

VLOGT2 and VLOGT3 proteins were grown as described

above. Concurrent with the induction of GT gene expres-

sion, substrates were added to a final concentration of

100 lM and the induced bacterial culture was grown for

20 h at 16�C, at which time the cultures were harvested by

centrifugation at 3,600g for 5 min. The supernatant was

decanted, acidified and extracted three times with ethyl

acetate. The ethyl acetate fractions were pooled, evapo-

rated to dryness, dissolved in 1 ml of methanol, and sub-

jected to HPLC.

Recombinant GT enzyme activity assays

Recombinant protein (rVLOGT1, rVLOGT2, rVLOGT3)

was isolated and rVLOGT2 was further purified by glu-

tathione Sepharose 4B (GE Healthcare) affinity chroma-

tography (as described in Hall and De Luca 2007). The

enzymes were assayed in 100 mM Tris, pH 8.0, with

0.1% b-mercaptoethanol in a final reaction volume of

125 ll, with 9 mM UDPG, 100 lM acceptor substrate

and 2–8 lg of purified rVLOGT2, 140–283 lg of desal-

ted total soluble rVLOGT1 protein or 119–242 lg of

desalted total soluble rVLOGT3 protein. Assays were

incubated for 2 h at 30�C, were stopped with an equal

volume of methanol or 0.12 N HCl in methanol (antho-

cyanidin substrates), were centrifuged, filtered and sub-

jected to HPLC analysis as described previously (Hall

and De Luca 2007).

The acceptor substrates quercetin, 2,4,5-tricholorphenol

(TCP), phenol, benzyl alcohol and phenylethanol were

assayed using the radioactive assay in a 100 ll reaction

volume containing 100 mM Tris–HCl, pH 8.0, 0.1% b-

mercaptoethanol, 100 lM (quercetin, TCP, phenol) or

20 mM (benzyl alcohol, phenylethanol) acceptor substrate,

25 mM gluconic acid lactone, 196 lM 14C-UDPG (specific

activity = 7.4 GBq/mmol), 804 lM cold UDPG and pro-

tein as described above. The assays were incubated for

30–120 min at 30�C, stopped with 10 ll of 1 N HCl,

extracted with 500 ll of ethyl acetate, centrifuged at

21,000g for 10 min and the organic phase was evaporated

to dryness. The products were resuspended in 10 ll of

methanol, spotted on a Polygram Sil/UV254 (Machery-

Nagel; Fisher Scientific) TLC plate and developed in the

solvent systems ethyl acetate:glacial acetic acid:formic

acid:water (100:11:11:20, by vol) (quercetin); ethyl ace-

tate:methanol:water (60:35:8, by vol) (TCP, phenol) or

ethyl acetate:acetone:dichloromenthane:methanol:water

(40:30:12:10:8, by vol) (benzyl alcohol, phenethyl alcohol

by vol) (Jones et al. 1999). The TLC plate was dried, and

exposed to a storage phosphor screen (GE Healthcare) for

1204 Planta (2011) 234:1201–1214

123

16–48 h prior to analysis with a Phosporimager FLA-3000

(Fujifilm) and MultiGuage ver 3.0.

Characterization of rVLOGT1, rVLOGT2

and rVLOGT3

Determination of pH optima, substrate specificities and

kinetic analysis were completed as described previously

(Hall and De Luca 2007) with the exceptions that the

typical buffer system used was 100 mM Tris–HCl, pH 8.0,

0.1% b-mercaptoethanol, and that the assay incubation

time was 2 h. To determine kinetic parameters with TCP, a

50 ll final reaction volume (100 mM Tris–HCl, pH 8.0,

0.1% b-mercaptoethanol), 50 lM C14UDPG, 250 lM of

cold UDPG 0.04 lg of purified rVLOGT2 and varying

concentrations of TCP were incubated for 90 min at 30�C,

were stopped with the addition of 5 ll of 0.12 N HCl and

extracted with 500 ll ethyl acetate, which was dried and

processed as described above. Non radioactive assays were

performed in order to determine the types of quercetin

glucosides produced by each GT reaction and reaction

products were analyzed and identified by HPLC–mass

spectrometry (HPLC-negative-ESI/MS) on an Agilent

1100 HPLC system equipped with a DAD detector and a

Bruker HCT ? ESI/MS.

The anthocyanin substrates cyanidin, cyanidin-3-O-

glucoside, cyanidin-3,5-O-diglucoside, peonidin, peonidin-

3-O glucoside, delphinidin, and malvidin; the stilbenes

trans-resveratrol, cis-resveratrol and trans-piceatannol; as

well as sapogenin, coumaric acid, sinapic acid, caffeic

acid, cinnamic acid, ferulic acid, chlorogenic acid, para-

hydroxybenzoic acid, benzoic acid, gallic acid, vanillin,

benzyl alcohol, phenylethanol, esculetin, catechin, epicat-

echin, phenol, abscisic acid, indole-3-acetic acid, salicylic

acid and UDP-glucuronic acid were not accepted by any of

the recombinant enzymes. UDP-galactose was not accepted

by rVLOGT2 with luteolin as an acceptor substrate, eri-

odictyol was not accepted by rVLOGT1 and 3, 6, 20, 30

tetrahydroxyflavone was not accepted by rVLOGT3.

In silico and phylogenetic analysis of Vitis labrusca

GTs

A BLAST search of the NCBI grape genome (Vitis vinifera

cv. Pinot Noir) database (http://www.ncbi.nlm.nih.gov/

genome/seq/BlastGen/BlastGen.cgi?taxid=29760) (PSPG)

box (WAPQVEVLAHPAVGCFVTHCGWNSTLESISAG

VPMVAWPFFADQ) (Vogt and Jones 2000) as a query

identified 273 sequences (e-5) with sequence identity

(E value \10-5) to plant secondary product glucos-

yltransferases. An additional BLAST search of the Vitis

vinifera genome with the nucleotide sequences of VLOGT1,

VLOGT2 and VLOGT3 identified the orthologous Vitis

vinifera sequences (LOC100248120; LOC100242998 and

LOC100246377, respectively). Sequence alignment and

phylogenetic analysis of the Vitis sp. GTs and other func-

tionally characterized GTs, were performed and visualized

using the Clustal algorithm of the CLC Bio Main Workbench

software (v.5.1).

Statistical analysis of the results

The data provided in Figs. 1, 2, 3 represents the mean of

3 or more biological replicates, while the relative gene

expression results provided in Fig. 4 represent the mean

real time RT-PCR values of two biological replicates.

The data provided in Fig. S1 represents the mean of 3

biological replicates. The data provided in Table S3 for

kinetic parameters of rVLOGT2 and rVLOGT3 repre-

sent mean values from at least three independent

experiments.

Results

Detection of cyanidin-3-O-glucoside-5-O-GT (Cy5OGT),

quercetin-30-O-GT (Q30OGT) and quercetin-7-O-GT

(Q7OGT) activities during grape berry development and in

several grape tissues.

Grape exocarp and mesocarp tissues prepared from

different stages of grape development were extracted and

desalted to produce crude protein extracts that were

assayed for Q30OGT and Q7OGT activity with quercetin as

glucose acceptor and for Cy5OGT activity using cyanidin

3-O-glucoside as glucose acceptor (Fig. 1). The different

quercetin glucosides including those substituted in posi-

tions 3, 7, 30 and 40 could be differentiated based on their

mass, HPLC retention times and absorbance maxima

(Table S2; Jurd 1962; Harborne 1967). Analyses by HPLC

of reaction products showed variable glucosylating activity

of each substrate depending on the tissue and stage of

development being analyzed (Fig. 1). While high levels of

Q7GT activity were clearly detected within grape meso-

carp (Fig. 1a) during grape ripening (veraison) between

weeks 12 and 16 AF, little of this activity could be detected

within grape exocarp in any stage of development

(Fig. 1b). Interestingly, there were low levels of Q30GT

activity in grape mesocarp but not in grape exocarp during

weeks 1–3 AF that disappeared between weeks 4–8 and

then reappeared during veraison between weeks 10–16 AF.

In contrast to these results, Cy5OGT activity was exclu-

sively found in grape exocarp, that it is the main site of

anthocyanin biosynthesis during grape development and

only with the onset of veraison between weeks 8–16

(Fig. 1b).

Planta (2011) 234:1201–1214 1205

123

Since maturing week 10–16 Concord grape tissues

consistently displayed higher Cy5OGT, Q7GT and

Q305GT activities (Fig. 1a, b), these results were com-

pared with the GT activities found in week 14 AF post-

veraison Vitis vinifera [Pinot Noir (PN) and Pinot Noir

Droit (PND)] grape mesocarp and exocarp (Fig. 1c).

While little Cy5OGT could be detected in either grape

variety and in either tissue, both PN and PND varieties

had variable levels of Q7GT and Q305GT activities. The

results suggest that while similar enzyme activities also

appear to occur in Vitis vinifera varieties and that their

activity profile through grape development should be

investigated.

Concord flower bud, leaf and flower tissues all had

variable levels of Cy5OGT, Q7GT and Q305GT (Fig. 1d).

Grape flowers displayed the highest levels of Q7GT

activity while the highest levels of Cy5OGT and Q305GT

activity were detected in leaf tissue (Fig. 1d). These results

suggest that Cy5OGT, Q7GT and Q305GT activities can be

detected in a number of grape tissues in addition to those

found within grape mesocarp and exocarp.

The flavonoid and anthocyanin content of mature week

16 AF grape berry mesocarp and exocarp

Flavonoid and anthocyanin extracts from mature berry

mesocarp were analyzed by UPLC MS (Fig. S1). Since

grape exocarp is the primary site of anthocyanin and fla-

vonoid accumulation, typical UPLC absorption profiles at

530 and 360 nm, respectively, are highlighted in Fig. S1.

The composition of anthocyanins and flavonoids were

characterized, based on their typical UPLC retention times,

their absorption spectra and their individual masses that

were correlated with the identical compounds from those

described previously (Fig. S1, Table insert; Park and Cha

2003). The results show that while small amounts of

cyanidin-3-O-glucoside, cyanidin 3-O-(6’’-O-coumaroyl)-

glucoside, myricetin 30-O-glucoside and quercetin-3-O-

glucoside could be identified in grape mesocarp, the

exocarp was the principal site of accumulation of 4

anthocyanins and 4 flavonoids. The exocarp anthocya-

nins included cyanidin, delphinidin and malvidin deriva-

tives, while flavonoids included quercetin-3-O-glucoside,

Fig. 1 Developmental profile of anthocyanin 5-O-, quercetin 7-O-

and 30- or 5-O- glucosyltransferase activity profile in Vitis labruscatissues. a Anthocyanin 5-O-glucosyltransferase (Cy5OGT), quercetin

30- or 5-O-glucosyltransferase (Q30OGT) and quercetin 7-O-gluco-

syltransferase (Q7OGT) activity in Vitis labrusca grape berry

mesocarp tissues throughout development. b Anthocyanin 5OGT,

quercetin 30OGT and quercetin 7OGT activity in Vitis labrusca grape

berry exocarp tissue throughout development. c Anthocyanin 5-OGT,

quercetin 30OGT and quercetin 7OGT activity in Vitis vinifera cv.

Pinot Noir (PN) and Pinot Noir Droit (PND) post-veraison (week 14

AF) mesocarp and exocarp tissues. d Anthocyanin 5OGT, quercetin

30OGT and quercetin 7OGT activity in Vitis labrusca leaf, flower and

pre-bloom bud tissue. Bars represent the mean of three independent

trials ± SD

1206 Planta (2011) 234:1201–1214

123

quercetin-3-O-glucuronide, myricetin-3-O-glucuronide,

myricetin 30-O-glucoside. It is important to note that

quercetin and myricetin 7-O-glucosides were not detected

in either exocarp or mesocarp tissues.

Identification and molecular cloning

of glucosyltransferase-like genes from grape berry

exocarp-enriched tissue

Initially, the corresponding full-length clones were isolated

from Vitis labrusca week 12 AF exocarp-enriched cDNA

using multiple rounds of 50RACE. During the course of

research, the grape genome was released, confirming the

presence of these genes within the Vitis vinifera genome.

A BLAST search of the DFCI grape index EST database with

the plant secondary product glucosyltransferase (PSPG) box

consensus sequence (Vogt and Jones 2000) identified 141

expressed Vitis sequences (E value \10e-5). Three Vitis

vinifera ESTs (TC75303, TC91700, NP9526770) with at

least 60% amino acid sequence similarity to previously

characterized flavonoid GTs were identified and pursued

based on their expression profile in several grape tissues,

including berries, flowers, and leaves, which glucosylate

quercetin at the 7-O-position (Fig. 1).

An alignment of TC75303, TC91700 and NP9526770

with the 273 Vitis vinifera genomic glucosyltransferases

suggested that these ESTs corresponded to three unique

genomic glucosyltransferase sequences located on chro-

mosome 5. TC75303 had 97% nucleotide sequence identity

to LOC100246377, TC91700 had 99% nucleotide

sequence identity to LOC100242998 and NP9526770 had

99% nucleotide sequence identity to LOC100248120.

Primers based on the predicted open reading frame of

these genes were used with Vitis labrusca week 12 AF

exocarp cDNA to amplify the corresponding transcripts

from Concord grape. The Concord clones (VLOGT1,

VLOGT2, VLOGT3) had 98–99% nucleotide sequence

identity and 97–98% amino acid sequence identity to their

Vitis vinifera equivalents (NP9526770, TC91700,

TC75303). VLOGT1 has 93 and 62% nucleotide sequence

identity to VLOGT2 and VLOGT3, respectively (1,344 bp)

and the three genes encode putative proteins of 447, 448

and 464 aa, respectively, with corresponding theoretical

molecular weights of 49.8, 49.7 and 51.3 kDa.

The presence of the PSPG box in the C-terminus of the

three Vitis labrusca clones (Fig. S2) placed these proteins

with more than 100 functionally characterized family 1 GTs

(Fig. S3). A BLAST search with VLOGT1, VLOGT2 and

VLOGT3 protein sequences revealed at least 60% amino

acid similarity to a Nicotiana tabacum GT (NtGT2) that

glucosylates the 3-O- position of coumarins and the 7-O-

position of flavonols (Taguchi et al. 2003; BAB88935), to a

Lycium barbarium catechin 40-O-GT which also glucosy-

lates the 7-O-position of flavonols (Noguchi et al. 2007;

BAG80544) to a Eucalyptus perriniana monoterpene O-GT

(Nagashima et al. 2004; BAD90934), to an anthocyanin

5-O-glucosyltransferase from Petunia x hybrida (Yamazaki

et al. 2002 (BAA89009), and to two flavonoid 7-O-GTs

(7GT)s from Maclura pomifera (Tian et al. 2006;

ABL85474) and Pyrus communis (Fischer et al. 2007;

AAY27090) (Fig. S3).

Biofermentation of kaempferol by rVLOGT1,

rVLOGT2, and rVLOGT3

Bacterial cultures (50 ml) expressing rVLOGT1, rVLOGT2,

rVLOGT3 or the pGEX empty vector were incubated with

cyanidin 3-O-glucoside (anthocyanin), kaempferol (flavonol)

and sapogenin (saponin) (Cascado and Heredia 1999) as

potential substrates for glucosylation. Cultures expressing the

empty vector did not glucosylate any of the substrates;

whereas cultures expressing rVLOGT1, rVLOGT2, and

Fig. 2 Effect of assay pH on rVLOGT2 and rVLOGT3 activity in

vitro. a Effect of assay pH on the production of quercetin 7-O-

monoglucoside (Q 7OMG), quercetin 30 or 5-O-monoglucoside

(Q 30OMG), and quercetin 40-O-monoglucoside (Q 40OMG) by

rVLOGT2 in vitro. b Effect of assay pH on the production of

isorhamnetin 7-O-monoglucoside (IsoR7OMG) by rVLOGT3 in

vitro. Each point represents the mean of three independent

trials ± SD

Planta (2011) 234:1201–1214 1207

123

rVLOGT3 produced 40.5 ± 11.6, 112.5 ± 14.0 and

26.8 ± 2.6 ng of kaempferol 7-O-glucoside/ml of cell culture

media [identified by co-chromatography with authentic stan-

dard and by UV spectral analysis (Hall and De Luca 2007)],

respectively, but none of the cultures were able to glucosylate

sapogenin or cyanidin-3-O-glucoside. The product of ka-

empferol glucosylation was identified as kaempferol 7-O-

glucoside by co-chromatography and UV spectral analysis

(Hall and De Luca 2007). These results confirmed that

rVLOGT1, rVLOGT2, and rVLOGT3 expressing lines were

biochemically active for further characterization using a cell-

free system.

Preliminary in vitro characterization of recombinant

GTs

E. coli cells expressing VLOGT1, VLOGT2 and VLOGT3

were induced for protein expression, and bacterial pellets

were harvested and extracted. SDS-PAGE analysis indi-

cated that approximately 95% of the GST-tagged recom-

binant proteins (molecular weight, 80 kDa) were insoluble.

Attempts to solubilise and refold an active protein from the

insoluble protein fractions of VLOGT1, VLOGT2 and

VLOGT3 were unsuccessful. When soluble active enzymes

were further purified by removal of the GST tag, they

Fig. 3 In vitro substrate

specificity of rVLOGT1,

rVLOGT2 and rVLOGT3.

a Substrate specificity of

rVLOGT2. b Substrate

specificity of rVLOGT1 and

rVLOGT3. The substrate in

each assay is identified,

followed by the product formed

in brackets (7OG,

7-O-glucoside; 30OG, 30 or

5-O-glucoside; 40OG,

40-O-glucoside; 7,40OdiG,

7,40-O-diglucoside; 7,3-O diG

- 7,3-O-diglucoside; 7,30OdiG,

7,30- or 5-O-glucoside. Each barrepresents the mean of at least

three independent trials ± SD

1208 Planta (2011) 234:1201–1214

123

displayed the same activity when they were GST tagged.

Crude desalted protein extracts of rVLOGT1 and

rVLOGT3 as well as GST purified rVLOGT2 were tested

with UDPG and the substrates quercetin, isorhamnetin,

dihydroquercetin, sinapic acid, cyanidin, cyanidin-3-O-

glucoside, para-hydroxybenzoic acid, trans-resveratrol,

and esculetin. Remarkably rVLOGT1 and rVLOGT3 con-

verted quercetin and isorhamnetin to monoglucosides as

determined by HPLC, while rVLOGT2 converted querce-

tin to three separate monoglucosides and kaempferol,

isorhamnetin and dihydroquercetin to monoglucosides. No

reaction products were formed with boiled enzymes or with

bacterial extracts expressing empty vector.

Since rVLOGT1 and rVLOGT3 seemed to have simi-

lar substrate specificities, further studies were performed

with rVLOGT2 and rVLOGT3 using quercetin to deter-

mine the pH optima of these 2 enzymes as described

previously (Hall and De Luca 2007). The pH optima of

rVLOGT2 was between pH 8.5 and 9.0 (Fig. 2a) while

that of rVLOGT3 was pH 8.5 (Fig. 2b). However

rVLOGT2 converted quercetin to one major (quercetin

7-O-monoglucoside; no hypsochromic shift of UV max

from 370 nm; Harborne 1967; Vogt et al. 1997; Kramer

et al. 2003; Hall and De Luca 2007) and 2 minor (quer-

cetin 30-O-monoglucoside and quercetin 40-O-monog-

lucoside; Jurd 1962; Vogt et al. 1997; Kramer et al. 2003)

products that were identified based on their different

HPLC retention times, their different absorbance maxima

and their identical masses (Table S2). In contrast,

rVLOGT3 converted quercetin to a single monoglucoside

that had the same HPLC, absorbance maxima and mass

properties of quercetin 7-O-glucoside.

Substrate specificity of rVLOGT1, rVLOGT2

and rVLOGT3

Substrate specificity of the recombinant proteins was

determined in vitro with UDP-glucose (UDPG) as a donor

substrate and 34 different acceptor substrates for

rVLOGT1, while 45 different acceptor substrates for tested

for rVLOGT2 and rVLOGT3. Recombinant rVLOGT2

produced the 7-O-glucoside of several flavonoids and

glucosylated the non-flavonoid model xenobiotic substrate

2,4,5-trichlorophenol (TCP). The 7-O-position of luteolin

Fig. 4 Relative gene expression of VlOGT1,2, and VlOGT3 as

compared to VLActin in berry mesocarp (a) and berry exocarp

(b) tissues of Vitis labrusca (VL) and in berry mesocarp and exocarp

tissues of Vitis vinifera cvs. Pinot Noir (PN) and Pinot Noir Droit

(PND) (c). d Relative gene expression of VlOGT1,2, and VlOGT3compared to VLActin gene expression in Vitis labrusca leaf, flower

bud and flower tissue. All values represent the mean of two

independent trials ± SD

Planta (2011) 234:1201–1214 1209

123

and eriodictyol were glucosylated by rVLOGT2 with the

highest specific activities (Fig. 3a) while the 30-O-gluco-

sides of luteolin and 3,6,20,30-tetrahydroxyflavone and the

40-O-glucosides of luteolin and quercetin were also pro-

duced at low levels. Recombinant VLOGT2 also converted

luteolin to 7,30-O-diglucoside and 7,40-O-diglucosides in

small amounts, while isorhamnetin and kaempferol could

also be converted to the corresponding 3, 7-O-diglucosides

(Fig. 3a).

Incubation of rVLOGT1 with isorhamnetin and UDPG

produced isorhamnetin 7-O-glucoside as the major product

and isorhamnetin 30-O- or 5-O-glucoside and isorhamnetin

7,30-O-diglucoside as minor products. rVLOGT1 also

glucosylated the 7-O -position of the flavonoids naringenin,

kaempferol, myricetin, luteolin, quercetin and dihydroqu-

eretin, while it was also able to glucosylate TCP at low

levels (Fig. 3b). rVLOGT3 glucosylated TCP and pro-

duced single products, identified as the 7-O-glucosides of

isorhamnetin, kaempferol, naringenin, eriodictyol, luteolin,

quercetin, myricetin and dihydroquercetin T (Fig. 3b).

Interestingly, VLOGT3 was able to use UDP-galactose

(UDP-Gal) as a donor with isorhamnetin as an acceptor at

low levels (Fig. 3b). Substrates which were not accepted by

any of the recombinant proteins are listed in ‘‘Materials and

methods.’’

Michaelis-Menton kinetic analysis of rVLOGT2

and rVLOGT3

Since recombinant enzymes appear to accept a broad range

of flavonoids as substrates, kinetic analyses were per-

formed with myricetin (for rVLOGT2) and isorhamnetin

(for rVLOGT3) as well as with TCP for both enzymes.

Kinetic analysis of rVLOGT2 calculated a 3.8-fold higher

affinity (Km) for myricetin than for TCP (48.4 and

184.5 lM) (Table S3). The turnover number (Kcat) with

TCP as a substrate for rVLOGT2 (Kcat = 105.7 s-1) is

more than 10,000-fold higher than with myricetin

(Kcat = 2.4 9 E-4 s-1) and the calculated catalytic effi-

ciency (Kcat/Km) is more than 100,000-fold higher with

TCP (Kcat/Km = 0.573 lM-1 s-1) than with myricetin

(Kcat/Km = 4.96 E-6 lM-1 s-1) (Table S3).

Michaelis-Menton kinetic analysis of rVLOGT3 with

varying concentrations of isorhamnetin yielded a Km of

30.2 lM, a Kcat of 8.11 E-7 s-1, and a Kcat/Km of

2.69 E-8 lM-1 s-1 (Table S3), similar to the kinetic

characteristics of rVLOGT2 with myricetin. Kinetic anal-

ysis of UDPG for rVLOGT2 (100 lM myricetin) and

rVLOGT3 (100 lM isorhamnetin) yielded Kms of 940,

and 886 lM, respectively; Kcats of 2.0 E-4 s-1 and

7.67 E-6 s-1, respectively, and Kcat/Kms of 2.13

E-7 lM-1 s-1 and 8.66 E-9 lM-1 s-1, respectively

(Table S3).

Gene expression profiling of VLOGT1 and VLOGT2

(VLOGT1,2) and VLOGT3 in grape berry, vegetative

and reproductive tissues

RNA was extracted from grape tissues and was screened

with one pair of primers to amplify both VLOGT1 and

VLOGT2, and a second pair of VLOGT3-specific primers.

Amplification with gene-specific primers corresponding to

the Vitis vinifera actin transcript was used to determine

baseline gene expression. The relative expression of the

VLOGT genes was analyzed in berry mesocarp (Fig. 4a)

and exocarp (Fig. 4b) to show that all 3 genes appeared to

be expressed to various extents throughout grape devel-

opment., with VLOGT1, 2 being expressed at higher levels

than VLOGT3 throughout development (Fig. 4a, b). While

the VLOGT3 transcript was variably detected throughout

grape development in both tissues, no clear peaks of gene

expression could be established to correlate with the 7- and

30/5-OGT enzyme activities observed throughout exocarp

and mesocarp grape development (Fig. 1a, b). In contrast

the relative expression of VLOGT1, 2 transcripts increases

in both mesocarp (Fig. 4a) and exocarp (Fig. 4b) tissues

with the onset of grape ripening from week 8 onwards that

may tentatively be correlated with the observed increases

of 7-OGT activities observed in mesocarp tissues (Fig. 1a).

VLOGT1, 2 also appeared to be expressed in both week

14 AF berry exocarp and mesocarp tissues of Vitis vinifera

cv. Pinot Noir and cv. Pinot Noir Droit, with lower relative

expression levels of the VLOGT3 transcript being detected

in berry exocarp tissue and no transcript being detected in

berry mesocarp tissue (Fig. 4c). Conversely, both

VLOGT1, 2 and VLOGT3 transcripts were detected in Vitis

labrusca leaf, flower bud and flower tissues (Fig. 4d).

These expression profiles are consistent with the corre-

sponding Vitis vinifera ESTs expression summaries from

the DFCI EST database in which these genes are expressed

in several tissues including flowers, pre-veraison berries,

and roots.

Phylogenetic analysis places VLOGT1, VLOGT2

and VLOGT3 in a clade with flavonoid 7GTs

A BLAST search of the NCBI grape genome database

(http://www.ncbi.nlm.nih.gov/genome/seq/BlastGen/Blast

Gen.cgi?taxid=29760) with the three VLOGT1, VLOGT2,

and VLOGT3 queries identified three corresponding geno-

mic sequences from Vitis vinifera cv. Pinot Noir (LOC

100248120, LOC100242998 and LOC100246377, respec-

tively) located on chromosome 5.

A phylogenetic tree places the Vitis sp. GTs identified in

this study in a clade with enzymes which glucosylate

terpenes, catechins, the 5-O-position of anthocyanins and

the 7-O-position of flavonols (Fig. S3; Yamazaki et al. 2002;

1210 Planta (2011) 234:1201–1214

123

Taguchi et al. 2003; Nagashima et al. 2004; Tian et al. 2006;

Fischer et al. 2007; Noguchi et al. 2007). GTs which accept

TCP as a substrate (Loutre et al. 2003; Meßner et al. 2003),

other 7GTs (Hirotani et al. 2000; Kramer et al. 2003; Jones

et al. 2003; Willits et al. 2004) and other 5GTs (Imayama et al.

2004; Ogata et al. 2005; Tohge et al. 2005) are located

throughout the phylogenetic tree (Fig. S3).

Discussion

The Vitis vinifera genome contains more than 140

genes with identity to glucosyltransferases

More than 150 plant group 1 glucosyltransferases have

been cloned from various plant sources, expressed in

recombinant organisms and functionally characterized in

vitro. This includes the recombinant expression of 107 GTs

from Arabidopsis which have been screened for activity

towards terpenes (Caputi et al. 2008), cytokinins (Hou et al.

2004), ABA (Lim et al. 2005), and quercetin (Lim et al.

2003). A BLAST search of the Vitis vinifera cv. Pinot noir

genome sequence with the PSPG box (Vogt and Jones

2000) identified 243 group 1 glucosyltransferases and a

BLAST search of the Vitis vinifera DFCI grape gene index

with the PSPG box as a query identified 141 Vitis sp. group

1 glucosyltransferases, which are expressed in grape, sug-

gesting biological roles for many of these enzymes.

Despite the prevalence of expressed GT sequences, only

six of the Vitis sp. enzymes have been functionally char-

acterized. Glucosyltransferases involved in the biosynthe-

sis of the viticulturally important anthocyanins (Ford et al.

1998; Janvary et al. 2009), a glycosyltransferase involved

in formation of flavonol 3-glucoronate (Ono et al. 2010), a

bifunctional glycosyltransferase involved in the formation

of flavonol-3-glucoside or galactoside (Ono et al. 2010)

and a bifuncional grape GT, which glucosylates hydroxy-

cinnamic acids and the human health beneficial stilbenes

(Hall and De Luca 2007) have characterized at the

molecular and biochemical level. In addition to anthocya-

nins, hydroxycinnamic acids and stilbenes, grapes biosyn-

thesize and accumulate the 3-O-, 7-O-, and 30O-glucosides

of many flavonols including quercetin, kaempferol, my-

ricetin and isorhamnetin that may involve GTs, which have

yet to be characterized. For example, recent studies (Park

and Cha 2003) using spot densities of flavonoids separated

by cellulose TLC combined with HPLC analysis identified

Q3G as a major flavonoid in grape exocarps while Q3G

and Q3G-7-O-glucuronide were the major flavonoids of

grape leaves and isorhamnetin-3-O–G and kaempferol

3,7-O–G were found in minor quantities.

The current study describes the homology-based cloning

and biochemical characterization of three full-length

(VLOGT1, VLOGT2, VLOGT3) members of the Vitis

labrusca cv. Concord flavonoid 7 GT gene family. The

deduced protein sequences of these genes are highly

identical to the 7 GTs from other plant species and group

phylogenetically with GTs that glucosylate monoterpenes,

phenolic acids, flavonols and dihydroflavonols in vitro

(Fig. S3; Lim et al. 2002; Taguchi et al. 2003; Nagashima

et al. 2004; Tian et al. 2006; Fischer et al. 2007; Noguchi

et al. 2007).

rVLOGT1, rVLOGT2, and rVLOGT3 glucosylate

flavonoids and the xenobiotic TCP

Biocatalysis with bacterial cultures expressing rVLOGT1,

rVLOGT2, and rVLOGT3 converted kaempferol to the

7-O-glucoside. In vitro cell-free extracts of these enzymes

glucosylated the model xenobiotic TCP as well as several

flavonoids to produce mono- and di-glucosides (Fig. 3).

rVLOGT2 predominantly glucosylated the 7-O-position in

vitro, with the flavone luteolin and the flavanone eri-

odictyol (differing only in B-ring stereochemistry) yielding

the two highest specific activities. Enzyme assays with

luteolin yielded 3 mono- and 2 di-glucosylated products,

whereas only the 7-O-glucoside was detected with eri-

odictyol as a substrate, suggesting that B-ring orientation is

essential for secondary and tertiary glucosylations. This

broad regio-specificity is consistent with previously char-

acterized flavonoid 7GTs, which can glucosylate multiple

positions on several flavonoid substrates in vitro (Kramer

et al. 2003; Tian et al. 2006).

VLRGT 1, 2, 3 gene expression as it relates to GT

activities and flavonoid/anthocyanin profiles

The predominant anthocyanins within Concord grape berry

are anthocyanidin 3-O-glucosides and quercetin 3-O-

glucosides (Park and Cha 2003; Wu and Prior 2005),

respectively. While the present analysis of mature Concord

berry tissues (Fig. S1) confirms the prevalence in the

exocarp of the 3-O-glucosides of different anthocyanidins-

and of quercetin, this tissue also contains the considerable

amounts of quercetin 3-O-glucuronide, myricetin 3-O-

glucoside and myricetin 3-O-glucuronide. The exocarp

tissue of several grape cultivars (Park and Cha 2003;

Amico et al. 2008; Castillo-Munoz et al. 2009) have also

been shown to accumulate quercetin 3-O-rhamnoglucoside

as well as the 3-O-galactosides, 3-O-glucuronides and 3-O-

glucosides of quercetin, kaempferol, myricetin, isorham-

netin, laricitrin and syringetin. Consistent with the preva-

lence of 3-O-glucosylated metabolites and the lack of 7-O-;

30-O-; or 40-O-glucosides in grape berries, enzyme assays

for quercetin-3-OGT showed up to 100-fold higher specific

activity (250 pkat/mg protein) than those obtained for

Planta (2011) 234:1201–1214 1211

123

Q7GT (2.5 pkat/mg protein) or Q30GT (0.5 pkat/mg pro-

tein) (Fig. 1) found in grape berry tissues. Low levels of

Q30GT activity were detected biphasically in Concord

berry mesocarp tissue, whereas both Q30GT and Q7GT

activity were detected during veraison in Vitis labrusca and

Vitis vinifera exocarp tissue (Fig. 1), consistent with

VLOGT1,2 and VLOGT3 gene expression in pre- and post-

veraison mesocarp and exocarp (Fig. 4). While the low

reaction rates observed in vitro (rVLOGT2 Kcat/Km

(myricetin) = 4.96 M-1 s-1; rVLOGT3 Kcat/Km(isorham-

netin) = 0.0269 M-1 s-1) are consistent with the lack of

accumulation of these metabolites within the grape berry,

these observations do not preclude a role for these genes in

flavonol modification in planta. However, these low reac-

tion rates may also suggest that the correct substrates for

VLOGT1,2,3 remain to be determined as well as the

pathways that they are involved with.

Within Vitis labrusca flower and bud tissue, Q7GT

activity is detected (Fig. 1c), and VLOGT1,2 and VLOGT3

are expressed (Fig. 4d). In Vitis labrusca leaves, the

expression of VLOGT1,2 and VLOGT3 (Fig. 4c) and the

detection of Q7GT activity is consistent with the accu-

mulation of kaempferol 3,7-O-diglycoside (Park and Cha

2003) and luteolin 7-O-glucoside (Moore and Giannasi

1994).

Broad substrate specificity and kinetic analysis suggests

a possible biological role for VLOGT1, VLOGT2

and VLOGT3 in the plant stress response to xenobiotics

In plants, GTs exist as large multigene families (Li et al.

2001; Bowles 2002; Lim et al. 2003) and have broad, but

regio-selective substrate specificity (Vogt and Jones 2000),

allowing the plant to glucosylate a wide range of natural

products and xenobiotics. Although in vitro characteriza-

tion of GTs may not reflect their in vivo function, over-

expression in heterologous systems facilitates the discovery

of several catalytic roles. The glucosylation of multiple

positions and functional groups on many metabolites in

addition to glucosylation of the endogenous substrate has

been demonstrated (Vogt et al. 1999; Kramer et al. 2003;

Nagashima et al. 2004; Tian et al. 2006; Hall and De Luca

2007; Brazier-Hicks et al. 2007).

Broad substrate specificity of GTs is a proposed mech-

anism that plants use to respond to their constantly

changing environment, and serves as a mechanism to

detoxify and compartmentalize xenobiotics (Jones and

Vogt 2001; Brazier-Hicks et al. 2007). The 7GTs from

Maclura pomifera, Oryza sativa, Allium cepa and Scutel-

laria baicalensis are not regio-selective since they produce

different mono- and di-glucosides with various flavonoid

substrates (Kramer et al. 2003; Jones et al. 2003; Tian et al.

2006). Similarly, rVLOGT1, rVLOGT2 and rVLOGT3

exhibit broad substrate specificity and VLOGT2 also

demonstrates broad regio-specificity with flavonol sub-

strates (Fig. 3). The recombinant Vitis labrusca GTs

characterized in this study glucosylate the xenobiotic TCP

in vitro (Fig. 3), and kinetic analysis of rVLOGT2 reveals

that TCP is the preferred substrate for this enzyme (Kcat/Km

for TCP is 105-fold higher than with myricetin) (Table S3),

consistent with a possible role for these enzymes in

xenobiotic detoxification.

Many Arabidopsis thaliana GTs glucosylate plant

metabolites and are also active in glucosylating xenobiotics

with varying affinities and catalytic efficiencies (Meßner

et al. 2003). Similarly, many GTs with broad substrate

specificity (including 7GTs) are often rapidly induced by

the application of herbicide safeners (Loutre et al. 2003)

and by wounding (Hirotani et al. 2000). The induction of

GTs by treatment with methyl jasmonate (Imanishi et al.

1998; Taguchi et al. 2003) or with salicylic acid (Hirotani

et al. 2000; Taguchi et al. 2003) suggests a proposed

downstream role in systemic acquired resistance and the

plant defense response (Dong 2004). Simultaneous initia-

tion of the plant defense response and an increase of GT

gene expression would prepare the plant for the concomi-

tant increase of secondary metabolites produced by the

plant (which are glucosylated and stored in vacuoles) to

combat the pathogen, as well as to prepare the plant for

exposure to pathogen-derived metabolites that can be

detoxified by glucosylation. The 7GT-like gene family of

Vitis labrusca likely contributes to this model of plant

defense, since these genes are expressed spatially and

temporally in tissues frequently subjected to pathogen

attack, such as post-veraison berry exocarp, leaf, flower

and pre-bloom bud tissue (Fig. 4).

Conclusion

This study has used homology-based cloning to identify

three full-length members of the Vitis labrusca cv. Concord

7GT gene family that correlate with homologous genes

found in the Vitis vinifera genome. Biocatalysis and sub-

strate specificity studies indicate that the Vitis labrusca

GTs can glucosylate flavonols as well as the xenobiotic

TCP in vitro. While enzyme activity profiles with crude

extracts from Vitis labrusca show that both 7- and 30-OGT

activities can be found in grape berries, in leaves and in

flower tissues, relevant 7- or 30-O-glucosides (Fig. S1, apart

from myricetin 30-O-glucoside) do not appear to accumu-

late in grape berries. The 100-fold lower levels of 7- and 30-OGT activities observed compared with those Q3GT may

suggest that VLOGTs 1, 2 & 3 play very minor roles in

flavonoid-O-glucosylation during grape development.

Kinetic analysis indicates that TCP is the preferred

1212 Planta (2011) 234:1201–1214

123

substrate for this class of enzymes, and gene expression of

VLOGT1,2 and VLOGT3 in tissues which are commonly

exposed to pathogen attack and xenobiotics suggests a

possible second role for the Vitis labrusca 7GT gene family

in plant defense.

Acknowledgments This research was funded by a Natural Sciences

and Engineering Research Council of Canada Discovery grant to

V.D.L. and a Tier 1 Canada Research Chair in Plant Biotechnology to

V.D.L., as well as an NSERC post-graduate fellowship to D.H. We

thank G&H Wiley Ltd (St, Catharine’s, Ontario, Canada) for allowing

us to collect Concord grapes during the growing season. We thank

Tim Jones for performing HPLC-mass spectrometry on selected fla-

vonoid reaction products.

References

Ali A, Strommer J (2003) A simple extraction and chromatographic

system for the simultaneous analysis of anthocyanins and

stilbenes of Vitis species. J Agric Food Chem 51:7246–7251

Amico V, Chillemi R, Mangiafico S, Spatafora C, Tringali C (2008)

Polyphenol-enriched fractions from Sicilian grape pomace:

HPLC-DAD analysis and antioxidant activity. Bioresource

Technol 99:5960–5966

Bowles D (2002) A multigene family of glycosyltransferases in

a model plant, Arabidopsis thaliana. Biochem Soc Trans

30:301–306

Brazier-Hicks M, Edwards LA, Edwards R (2007) Selection of plants

for roles in phytoremediation: the importance of glucosylation.

Plant Biotechnol J 5:627–635

Caputi L, Lim E, Bowles D (2008) Discovery of new biocatalysts

for the glucosylation of terpenoid scaffolds. Chem Eur J

14:6656–6662

Cascado CG, Heredia A (1999) Structure and dynamics of reconsti-

tuted cuticular waxes of grape berry cuticle (Vitis vinifera L.).

J Exp Bot 50:175–182

Castillo-Munoz N, Gomez-Alonso S, Garcıa-Romero E, Hermosın-

Gutierrez I (2007) Flavonol profiles of Vitis vinifera red grapes

and their single cultivar wines. J Agric Food Chem 55:992–1002

Castillo-Munoz N, Gomez-Alonso S, Garcıa-Romero E, Gomez V,

Velders A, Hermosın-Gutierrez I (2009) Flavonol 3-O-glyco-

sides series of Vitis vinifera cv. Petit Verdot red wine grapes.

J Agr Food Chem 57:209–219

Dong X (2004) NPR1, all things considered. Curr Opin Plant Biol

7:547–552

Fischer T, Gosch C, Pfeiffer J, Halbwirth H, Halle C, Stich K,

Forkmann G (2007) Flavonoid genes of pear (Pyrus communis).

Trees 21:521–529

Ford CM, Boss PK, Høj PB (1998) Cloning and characterization of

Vitis vinifera UDP-Glucose:flavonoid 3-O-glucosyltransferase, a

homologue of the enzyme encoded by the maize Bronze-1 locus

that may primarily serve to glucosylate anthocyanidins in vivo.

J Biol Chem 273:9224–9233

Hall D, De Luca V (2007) Mesocarp localization of a bi-functional

resveratrol/hydroxycinnamic acid glucosyltransferase of Con-

cord grape (Vitis labrusca). Plant J 49:579–591

Harborne JB (1967) Comparative biochemistry of the flavonoids.

Academic Press, London

Hirotani M, Kuroda R, Suzuki H, Yoshikawa T (2000) Cloning and

expression of UDP-glucose: flavonoid 7-O-glucosyltransferase

from hairy root cultures of Scutellaria baicalensis. Planta

210:1006–1013

Hou B, Lim E, Higgins G, Bowles D (2004) N-Glucosylation of

cytokinins by glucosyltransferases of Arabidopsis thaliana.

J Biol Chem 279:47822–47832

Huntley A (2007) Grape flavonoids and menopausal health. Meno-

pause Int 13:165–169

Imanishi S, Hashizume K, Kojima H, Ichihara A, Nakamura K (1998)

An mRNA of tobacco cell, which is rapidly inducible by methyl

jasmonate in the presence of cycloheximide, codes for a putative

glycosyltransferases. Plant Cell Physiol 39:202–211

Imayama T, Yoshihara N, Fukuchi-Mizutani M, Tanaka Y, Ino I,

Yabuya T (2004) Isolation and characterization of a cDNA clone

of UDP-glucose: anthocyanin 5-O-glucosyltransferase in Irishollandica. Plant Sci 167:1243–1248

Janvary L, Hoffmann T, Pfeiffer J, Hausmann L, Topfer R, Fischer T,

Schwab W (2009) A double mutation in the anthocyanin 5-O-

glucosyltranferase gene disrupts enzymatic activity in Vitisvinifera L. J Agr Food Chem 57:3512–3518

Jones P, Vogt T (2001) Glycosyltransferases in secondary plant

metabolism: tranquilizers and stimulant controllers. Planta

213:164–174

Jones P, Møller B, Høj P (1999) The UDP-glucose:p-hydroxymand-

elonitrile-O-glucosyltransferase that catalyzes the last step in

synthesis of the cyanogenic glucoside dhurrin in Sorghumbicolor. J Biol Chem 274:35483–35491

Jones P, Meßner B, Nakajima JI, Schaffner AR, Saito K (2003)

UGT73C6 and UGT78D1, glycosyltransferases involved in

flavonol glycoside biosynthesis in Arabidopsis thaliana. J Biol

Chem 278:43910–43918

Jurd L (1962) Spectral properties of flavonoid compounds. In:

Geissmann TA (ed) The chemistry of flavonoid compounds.

Pergamon Press Inc, Oxford, pp 107–155

Kramer CM, Prata RTN, Willits MG, De Luca V, Steffens JC, Graser G

(2003) Cloning and regiospecificity studies of two flavonoid glucos-

yltransferases from Allium cepa. Phytochemistry 64:1069–1076

Langcake P, Pryce RJ (1977) A new class of phytoalexins from

grapevine. Experientia 33:151–152

Leifert W, Abeywardena M (2008) Cardioprotective actions of grape

polyphenols. Nutr Res 28:729–737

Li Y, Baldauf S, Lim EK, Bowles DJ (2001) Phylogenetic analysis of

the UDP-glycosyltransferase multigene family of Arabidopsisthaliana. J Biol Chem 276:4338–4343

Lim E, Doucet CJ, Li Y, Elias L, Worrall D, Spencer SP, Ross J,

Bowles DJ (2002) The activity of Arabidopsis glycosyltransfer-

ases toward salicylic acid, 4-hydroxybenzoic acid and other

benzoates. J Biol Chem 277:586–592

Lim E, Baldauf S, Li Y, Elias L, Worrall D, Spencer SP, Jackson RG,

Taguchi G, Ross J, Bowles DJ (2003) Evolution of substrate

recognition across a multigene family of glycosyltransferases in

Arabidopsis. Glycobiology 13:139–145

Lim E, Doucet C, Hou B, Jackson R, Abrams S, Bowles D (2005)

Resolution of (?)-abscisic acid using an Arabidopsis glucosyl-

transferase. Tetrahedron Assymetr 16:143–147

Loutre C, Dixon DP, Brazier M, Slater M, Cole DJ, Edwards R (2003)

Isolation of a glucosyltransferase from Arabidopsis thalianaactive in the metabolism of the persistent pollutant 3, 4-dichlo-

roaniline. Plant J 34:485–493

Meßner B, Thulke O, Schaffner AR (2003) Arabidopsis glucos-

yltransferases with activities toward both endogenous and

xenobiotic substrates. Planta 217:138–146

Moore MO, Giannasi DE (1994) Foliar flavonoids of eastern North

American Vitis (Vitaceae) north of Mexico. Plant Syst Evol

193:21–36

Munoz-Espada A, Wood K, Bordelon B, Watkins B (2004) Antho-

cyanin quantification and radical scavenging capacity of Con-

cord, Norton, and Marechal Foch grapes and wines. J Agr Food

Chem 52:6779–6786

Planta (2011) 234:1201–1214 1213

123

Nagashima S, Tomo S, Orihara Y, Yoshikawa T (2004) Cloning and

characterization of glucosyltransferase cDNA from Eucalyptusperriniana cultured cells. Plant Biotech 21:343–348

Noguchi A, Saito A, Homma Y, Nakao M, Sasaki N, Nishino T,

Takahashi S, Nakayama T (2007) A UDP-glucose: isoflavone

7-O-glucosyltransferase from the roots of soybean (Glycine max)

seedlings. J Biol Chem 282:23581–23590

Ogata J, Kanno Y, Itoh Y, Tsugawa H, Suzuki M (2005) Plant

biochemistry: anthocyanin biosynthesis in roses. Nature

435:757–758

Ono E, Homma Y, Horikawa M, Kunikane-Doi S, Imai H, Takahashi

S, Kawai Y, Ishiguro M, Fukui Y, Nakayama T (2010)

Functional differentiation of the glycosyltransferases that con-

tribute to the chemical diversity of bioactive flavonol glycosides

in grapevines (Vitis vinifera). Plant Cell 22:2856–2871

Park HJ, Cha HC (2003) Flavonoids from leaves and exocarps of the

grape Kyoho. Korean J Biol Sci 7:327–330

Polaskova P, Herszage J, Ebeler S (2008) Wine flavor: chemistry in a

glass. Chem Soc Rev 37:2478–2489

Schneider E, von der Heydt H, Esperester A (2008) Evaluation of

polyphenol composition in red leaves from different varieties of

Vitis vinifera. Plant Med 74:565–572

Taguchi G, Ubukata T, Nobuaki H, Yamamoto H, Okazaki M (2003)

Cloning and characterization of a glucosyltransferase that reacts

on the 7-hydroxyl group of flavonol and the 3-hydroxyl group

of coumarin from tobacco cells. Arch Biochem Biophys

420:95–102

Tian L, Blount JW, Dixon RA (2006) Phenylpropanoid glycosyl-

transferases from osage orange (Maclura pomifera) fruit. FEBS

Lett 580:6915–6920

Tohge T, Nishiyama Y, Hirai MY, Yano M, Nakajima J, Awazuhara

M, Inoue E, Takahashi H, Goodenow DB, Kitayama M, Noji M,

Yamazaki M, Saito K (2005) Functional genomics by integrated

analysis of metabolome and transcriptome of Arabidopsisplants over-expressing an MYB transcription factor. Plant J

42:218–235

Vogt T, Jones P (2000) Glycosyltransferases in plant natural product

synthesis: characterization of a supergene family. Trends Plant

Sci 5:380–388

Vogt T, Zimmermann E, Grimm R, Meyer M, Strack D (1997) Are

the characteristics of betanidin glucosyltransferases from cell-

suspension cultures of Dorotheanthus bellidiformis indicative of

their phylogenetic relationship with flavonoid glucosyltransfer-

ase? Planta 203:349–361

Vogt T, Grimm R, Strack D (1999) Cloning and expression of a

cDNA encoding betanidin 5-O-glucosyltransferase, a betanidin-

and flavonoid-specific enzyme with high homology to inducible

glucosyltransferases from the Solanaceae. Plant J 19:509–519

Willits MG, Giovanni M, Prata RTN, Kramer CM, De Luca V,

Steffens JC, Graser G (2004) Bio-fermentation of modified

flavonoids: an example of in vivo diversification of secondary

metabolites. Phytochemistry 65:31–41

Wu X, Prior RL (2005) Systematic identification and characterization of

anthocyanins by HPLC–ESI–MS/MS in common foods in the

United States: fruits and berries. J Agric Food Chem 53:2589–2599

Yamazaki M, Yamagishi E, Gong Z, Fukuchi-Mizutani M, Fukui Y,

Tanaka Y, Kusumi T, Yamaguchi M, Saito K (2002) Two

flavonoid glucosyltransferases from Petunia hybrida: molecular

cloning, biochemical properties and developmentally regulated

expression. Plant Mol Biol 48:401–411

1214 Planta (2011) 234:1201–1214

123