New flavanol O-glycosides in grape and wine

Article abstract of DOI:10.1016/j.foodchem.2018.06.019

The presence of monomeric and dimeric flavan-3-ol monohexosides was investigated in grapes and wines. Polyphenol extracts were prepared from grape seeds and skins (Syrah, Merlot, and Cabernet-Sauvignon) sampled at three different developmental stages. Different wines (Tannat, Alicante, Syrah, Merlot, and Grenache) were also studied. The different samples obtained were analyzed by UPLC-ESI-IT-MS. Specific molecular ions corresponding to flavan-3-ol hexosides were detected by using targeted molecular ions with specific m/z values: 451 for (epi)catechin hexosides, and 739 for procyanidin dimer hexosides. 4′-O-β-glucosyl-(+)-catechin and 7-O-β-glucosyl-(+)-catechin were obtained by using a glucosyl transferase from apple, UGT71A15, and their structures determined by NMR. These glucosylated monomers and the dimers were identified in all analyzed grape seed and several skin extracts at the different stages and in wines made from different varieties.

Full text of DOI:10.1016/j.foodchem.2018.06.019

Accepted Manuscript
New flavanol O-glycosides in grape and wine
Marie Zerbib, Jean-Paul Mazauric, Emmanuelle Meudec, Christine Le
Guernevé, Alexander Lepak, Bernd Nidetzky, Véronique Cheynier, Nancy
Terrier, Cédric Saucier
PII:
S0308-8146(18)30970-1
DOI:
https://doi.org/10.1016/j.foodchem.2018.06.019
FOCH 22980
Reference:
To appear in:
Food Chemistry
Received Date:
Revised Date:
Accepted Date:
22 December 2017
1 June 2018
3 June 2018
Please cite this article as: Zerbib, M., Mazauric, J-P., Meudec, E., Le Guernevé, C., Lepak, A., Nidetzky, B.,
Cheynier, V., Terrier, N., Saucier, C., New flavanol O-glycosides in grape and wine, Food Chemistry (2018), doi:
https://doi.org/10.1016/j.foodchem.2018.06.019
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

New flavanol O-glycosides in grape and wine
a
b
b
Marie ZERBIB , Jean-Paul MAZAURIC , Emmanuelle MEUDEC , Christine Le
b
d
d
b
GUERNEVÉ , Alexander LEPAK , Bernd NIDETZKY , Véronique CHEYNIER ,
c
a*
Nancy TERRIER , Cédric SAUCIER
a
SPO, Univ Montpellier, INRA, Montpellier Supagro, Montpellier, France
marie.zerbib@umontpellier.fr and cedric.saucier@umontpellier.fr
b
SPO, INRA, Montpellier Supagro, Univ Montpellier, Montpellier, France
emmanuelle.meudec@inra.fr, christine.le-guerneve@inra.fr, veronique.cheynier@inra.fr
c
AGAP, INRA, CIRAD, Montpellier SupAgro, Univ Montpellier, Montpellier, France
nancy.terrier@inra.fr
d
Institute of Biotechnology and Biochemical Engineering, Graz University of Technology,
Graz, Austria
alexander.lepak@tugraz.at and bernd.nidetzky@tugraz.at
*
corresponding author : Cédric Saucier: cedric.saucier@umontpellier.fr
1

Abstract:
The presence of monomeric and dimeric flavan-3-ol monohexosides was investigated in
grapes and wines. Polyphenol extracts were prepared from grape seeds and skins (Syrah,
Merlot, and Cabernet-Sauvignon) sampled at three different developmental stages. Different
wines (Tannat, Alicante, Syrah, Merlot, and Grenache) were also studied. The different
samples obtained were analyzed by UPLC-ESI-IT-MS. Specific molecular ions corresponding
to flavan-3-ol hexosides were detected by using targeted molecular ions with specific m/z
values: 451 for (epi)catechin hexosides, and 739 for procyanidin dimer hexosides. 4’-O--
glucosyl-(+)-catechin and 7-O--glucosyl-(+)-catechin were obtained by using a glucosyl
transferase from apple, UGT71A15, and their structures determined by NMR. These
glucosylated monomers and the dimers were identified in all analyzed grape seed and several
skin extracts at the different stages and in wines made from different varieties.
Keywords: wine, grape, flavan-3-ol glycoside, mass spectrometry
1. Introduction
Flavonoids are a category of polyphenols present in a wide variety of natural products such as
grape, tea, and cocoa (Quideau, Deffieux, Douat-Casassus, & Pouysegu, 2011). They have
health properties and are involved in organoleptic properties such as the color and the taste of
wine and play a role in plant defense mechanisms (Corder, Mullen, Khan, Marks, Wood,
Carrier, et al., 2006). The common molecular backbone of these secondary metabolites is a
C6-C3-C6 type. Flavan-3-ols, called flavanols here, are an important subclass of flavonoids.
They are present in grape and wine as monomers or polymers called proanthocyanidins (PAs)
or condensed tannins. The monomeric units are linked through C
4
-C
6
or C
4
-C linkages (B
8
type) with sometimes additional C2-O-C5 or C2-O-C7 bond (A type). In Vitis vinifera berries,
four principal units of PAs are usually found: (+)-catechin, (+)-gallocatechin, (-)-epicatechin,
2

(
-)-epigallocatechin and (-)-epicatechin gallate. They are synthesized in skins and seeds at the
earlier developmental stages after flowering, and the monomer and polymer concentration
remains stable after veraison, the onset of ripening (Downey, Harvey, & S.P., 2003 ; Terrier,
Ollé, Verriès, & Cheynier, 2009).
The common flavonoid biosynthesis sequence is well described in the literature(Tanner &
Kristiansen, 1993 ; Xie, Sharma, Paiva, Ferreira, & Dixon, 2003 ; Fujita, Soma, Goto-
Yamamoto, Shindo, Kakuta, Koizumi, et al., 2005 ; Bogs, Downey, Harvey, Ashton, Tanner,
&
Robinson, 2005). However, the chemical or biochemical mechanism of the polymerization
steps generating PAs remains unknown. In 2008, Dixon and collaborators published a work
about the expression of a glucosyltransferase, UGT72L1, in the seed coat of Medicago
truncatula. This enzyme catalyzes the formation of epicatechin 3’-O-glucoside, named here
E3’OG, in the cytoplasm and its over-expression results in increased accumulation of PA-like
compounds in the vacuole (Pang, Cheng, Huhman, Ma, Peel, Yonekura-Sakakibara, et al.,
2
013). Further studies have suggested that E3’OG rather than free epicatechin is the substrate
for the tonoplast-localized MATE, transporters MATE1 from M. truncatula and TT12 from
A. thaliana, involved in PA biosynthesis (Zhao & Dixon, 2009). This glucosylated flavan-3-ol
was postulated to be an intermediate in flavan-3-ol polymerization in Medicago truncatula
(
Pang, Peel, Sharma, Tang, & Dixon, 2008).
Polyphenols substituted with different hexoses (glucose, galactose, mannose,..), including
stilbenes (Versari, Parpinello, Tornielli, Ferrarini, & Giulivo, 2001 ; Ali & Strommer, 2003),
phenolic acids (Monagas, Bartholomé, & Gomez-Cordovés, 2007), flavonols and flavones
(
Castillo-Muñoz, Gómez-Alonso, García-Romero, Gómez, Velders, & I., 2009 ; Hmamouchi,
Es-Safi, Lahrichi, Fruchier, & Essassi, 1996) have been described in the literature. Also,
several flavan-3-ol glucosides were found in a wide range of plants : (+)-catechin-7-O-β-
glucoside in elm (Kim, Yeom, Kim, Shim, Lim, & Lee, 2004), buckwheat (Smaali, Maugard,
3

Limam, Legoy, & Marzouki, 2007 ; Watanabe, 1998), cowpea seeds (Cui, Song, Shrestha,
Chung, Kim, Jeong, et al., 2016), barley and malt (Friedrich & Galensa, 2002), (+)-catechin
3
5
-O-glucoside in lentils (Dueñas, Sun, Hernández, Estrella, & Spranger, 2003), (+)-catechin
-O-, 7-O-, 3’-O-, and 4’-O-glucosides in Chinese peony (Tanaka, Kataoka, Tsuboi, &
Kouno, 2000), (+)-catechin 7-O- β- and 4′-O-β-glucosides in Douglas-fir bark (Foo &
Karchesy, 1989), (+)-catechin 5-O-β-glucoside and (−)-catechin 7-O-β-glucoside in rhubarb
and Rhaphiolepis umbellata bark (Nonaka, Ezaki, Hayashi, & Nishioka, 1983). The
characterization of flavan-3-ol hexosides was investigated in literature, and to this aim, some
of these compounds were produced. Indeed, different strategies have been used to obtain
these standards: organic synthesis (Raab, Barron, Arce Vera, Crespy, Oliveira, & Williamson,
2
010), enzymatic reactions (Smaali, Maugard, Limam, Legoy, & Marzouki, 2007 ; Vic,
Biton, Le Beller, Michel, & Thomas, 1994). Some purifications from plants or foods may be
performed as well. Raab and collaborators developed the chemical synthesis of (+)-catechin
3
’-O-β-D-glucopyranoside, (+)-catechin 5-O-β-D-glucopyranoside, and (+)-catechin 3-O-β-
D-glucopyranoside. They also obtained (+)-catechin 4’-O-β-D-glucopyranoside and (+)-
catechin 7-O-β-D-glucopyranoside by using enzyme extracts from lentil and barley (Raab,
Barron, Arce Vera, Crespy, Oliveira, & Williamson, 2010). Another work on a glucosyl-
transferase from apple called UGT71A15 has some affinity for flavonoids such as phlorizin
(
phloretin 2′-O-glucoside) (Lepak, Gutmann, Kulmer, & Nidetzky, 2015). In addition,
unpublished preliminary assays showed that this enzyme has some activity on flavanols.
The presence of hexosylated monomers of flavanols was already reported in Merlot grape
seeds and wine but not in grape skins (Delcambre & Saucier, 2012). By using an ESI/Q-
TOF/MS with a targeted MS/MS strategy, 14 flavanol hexosides were detected. MS
fragmentation patterns indicated that these molecules are O-hexosides of the four major
monomers found in grape (Simirgiotis, Silva, Becerra, & Schmeda-Hirschmann, 2012).
4

However, the nature and position of the hexoside substituents have not been determined and
whether these compounds can be found as constitutive units of proanthocyanidins is
unknown.
The main objective of the present work was to detect by mass spectrometry new flavan-3-ol
hexosides compounds in grapes and wines. Also, grape tissues, grape stages of development
and varieties influences were investigated. Different grapes varieties at three different stages
of development and wine were selected.
In this publication, we wanted to test the following hypotheses:
-
-
-
Several isomers of dimeric flavanol hexosides are present in different grape varieties
and wines
The content of these compounds is influenced by the grape variety, the tissue (skin or
seed) and the grape development stage
Some of the compounds are  glucosides of flavanols
2. Materials and Methods
2
.1 Reagents and samples
Deionized water was purified with a Milli-Q purification system (Millipore, Molsheim,
France). Acetone and diethyl ether were obtained from Analytic Lab (St Mathieu de Treviers,
France). Formic acid and HPLC grade methanol were purchased from Sigma Aldrich (St
Louis, MO, USA).
Grapes from Unité Expérimentale de Pech Rouge in Gruissan (France) were picked at three
different dates corresponding to different development stages.
Merlot: 31/07/2015 (3.4 °Brix), 14/08/2015 (20.2 °Brix) ; 09/09/2015 (23.1 °Brix)
Syrah: 31/07/2015 (2.9 °Brix), 14/08/2015 (11.1 °Brix) ; 09/09/2015 (20.4 °Brix)
5

Cabernet Sauvignon (CS): 14/08/2015 (3.2 °Brix), 09/09/2015 (15.1 °Brix) ; 24/09/2015 (22.9
Brix).
°
The first development stage sampling point of each grape is named phase I (P-I), the second
one, phase II (P-II), and the last one is at maturity, called phase III (P-III). The samples were
stored at -80°C before analysis.
Languedoc-Roussillon red wines selected for a screening were: a Tannat press wine from cave
de Crouseilles, Madiran-2016, an Alicante Bouschet from Chateau de Pinet-2015, a Merlot
from Mas Pellier-2014, a Syrah from Mas de Valbrune-2013 and a Grenache from Château
Flaugergues, Montpellier-2016. Red wines were stored at -80°C before analysis.
2.2 Preparation of grape tannin extracts
For each grape variety, seeds and skins were removed by hand on ice from 100 berries. Ten
grams of each tissue were extracted using 200 mL of acetone/water 7:3, under nitrogen with
mechanical agitation, overnight. Solutions were filtered on filter paper, evaporated under
reduced pressure at 37°C. The residues were dissolved in deionized water and lyophilized.
Tannin extract powders were stored at -20°C.
2
.3. UPLC-DAD-ESI-IT-MS method
All analyses were performed on a Waters Acquity UPLC-DAD system (Milford, MA)
coupled with a Bruker Daltonics Amazon (Bremen, Germany) mass spectrometer equipped
with an electrospray source and an ion trap mass analyzer. Separation was achieved using a
C18 column (Waters, Milford, BEH, 1.7 μm, 1x150 mm) at temperature of 35°C. The binary
mobile phase consisted of deionized water with 1 % of formic acid (A) and methanol with 1%
of acid formic (B). The 45-minutes-long gradient was as follows: 0-1 min, isocratic, 2% B ; 1-
1
0 min, 2-30% B ; 10-12 min, isocratic, 30% B ; 12-25 min, 30-75% B ; 25-30 min, 75-90%
6

-
B ; 30-35 min, isocratic, 90% B ; 35-45 min, isocratic 2% B. The flow rate was 0.08 mL.min
1
.
For UPLC-DAD-ESI-IT-MS analysis, grape extracts were prepared in methanol/H2O 1:1 to a
-1
final concentration of 1 mg.mL .
Red wines were all centrifuged at 13500 rpm, during 15 minutes at 4°C. The supernatants
were recovered and injected.
The MS instrument was operated in the negative mode with fragment ion scan mode focused
on specific precursor ions at m/z 451 and 739 (end plate off set: -500V; temperature, 200°C;
-
1
nebulizer gas: 10 psi and dry gas, 5 L.min ; capillary voltage, 4.5kV). Collision energy for
2
fragmentation used for MS experiments were set at 1.
A UGT714A15 reaction kinetic study was done, as shown on Appendix 1. The proportions of
(
+)-catechin 4’-O-β-D-glucoside and (+)-catechin 7-O-β-D-glucoside were monitored with
the assumption that they had the same UV absorption coefficient as (+)-catechin at 280 nm.
Monomeric and dimeric flavanol hexosides relative intensities were measured based on the
peak height of their associated m/z Extract Ion Chromatogram (EIC). Each sample was
analyzed in triplicate. Samples were analyzed in several batches within a 2-month period and
the average intensities and standard deviations (error bars) calculated.
2
.4. Heterologous production of UGT71A15 and enzymatic reactions
Cloning and transformation of the glucosyltransferase UGT71A15 (GenBank: DQ103712)
have been described elsewhere (Gosch, Halbwirth, Schneider, Hölscher, & Stich, 2010). The
heterologous expression of N-terminal Strep-tagged UGT71A15 in E. coli and purification by
Strep-affinity chromatography were performed as described earlier (Lepak, Gutmann,
Kulmer, & Nidetzky, 2015).
7

UGT71A15 was assayed in a reaction volume of 200 µL containing 50 mM HEPES pH 8.0,
0 mM KCl, 13 mM MgCl , 50 µg/mL enzyme with different acceptor reagents, (+)-catechin
5
2
and (-)-epicatechin, and UDPG concentrations (1 mM, 2.5 mM and 5 mM) at 30°C. To obtain
a sufficient amount for structural characterization by NMR and MS, the reaction was scaled
up to 4 ml using the same buffer conditions. (+)-catechin 4’-O--glucoside was then purified
by preparative HPLC (940 LC Varian) using a C18 reversed phase (Microsorb C18 100-3,
1
00*21.4 mm (L*ID), Dynamax Agilent). Samples were filtered on a 0.2 µm regenerated
cellulose membrane filter and 1 mL was injected. The binary mobile phase consisted of
deionized water with 1 % of formic acid (A) and acetonitrile, water and acid formic (80:19:1)
(
B). The 16-minutes-long gradient was as follows: 0-1 min, 10% B; 1-10 min, 16% B; 10-12
-1
min, 80% B; 12-14 min, 10% B; 14-16 min, 10% B. The flow rate was 10 mL.min .
2
.6 NMR spectroscopy experiments
All experiments were performed on an Agilent 500 MHz DD2 NMR spectrometer (Agilent
Technologies, Santa Clara, CA, USA) operating at 500.05 MHz for 1H and 125.75 MHz for
1
3
C and equipped with a 5mm indirect detection Z-gradient probe. The chemical shifts were
reported to that of methanol-d4, the organic solvent used for the experiments (3.31ppm and
1
13
4
9.1ppm for H and C respectively). All data were collected at 25°C with active sample
temperature regulation.
Identification and structure characterization of products were performed using both 1D and
1
1
13
2
D NMR, homonucluear H (DOSY, COSY, TOCSY and ROESY) and heteronuclear H/ C
(
HSQC and HMBC) experiments. Data were processed and analyzed using VNMRJ software.
2
.7.1. (+)-catechin-4’-O--glucoside (C4’OG)
1
H NMR (25°C, MetOD, ppm) δ 4.629 (d, J = 7,3, C2) ; 3.982 (m, C3), 2.516 (dd, 7.9 ; 16.5,
C4), 2.817 (dd, 5.5 ; 16.5, C4), 5.932 (d, 2.2, A6), 5.868 (d, 2.2, A8), 6.906 (d, 2.0, B2’),
8

7
.183 (d, 8.4, B5’), 6.831 (d, 2.0; 8.4, B6’), 4.771 (d, 7.5, Glc1), 3.492 (m, Glc2), 3.474 (m,
Glc3), 3.408 (m, Glc4), 3.410 (m, Glc5), 3.716 (m, Glc6,6’), 3.895 (bd, 11.9 Hz, Glc6,6’).
1
3
C NMR (25°C, MetOD) δ 82.45 (C2), 68.69 (C3), 28.35 (C4), 157.59 (A5), 96.33 (A6),
1
57.80 (A7), 95.45 (A8), 156.73 (A9), 100.68 (A10), 136.39 (B1’), 115.79 (B2’), 148.25
(
B3’), 146.51 (B4’), 118.58 (B5’), 119.88 (B6’), 104.29 (Glc1), 74.90 (Glc2), 77.51 (Glc3),
7
1.32 (Glc4), 78.26 (Glc5), 62.39 (Glc6,6’).
2
.7.2. (+)-catechin-7’-O--glucoside (C7’OG)
1
H NMR (25°C, MetOD, ppm) δ 4.659 (d, J = 7,4, C2) ; 3.996 (m, C3), 2.537 (dd, 8.2 ; 16.1,
C4), 2.817 (dd, 5.2 ; 16.5, C4), 6.201 (d, 2.4, A6), 6.157 (d, 2.4, A8), 6.829 (d, 1.8, B2’),
6
.758 (d, 8.1, B5’), 6.711 (d, 1.8 ; 8.1, B6’), 4.815 (d, 7.6, Glc1), 3.35-3.45 (m, Glc2, Glc3,
Glc4, Glc5), 3.696 (dd, 4.5 ; 12.0, Glc6,6’), 3.884 (dd, 2.0 ; 12.0, Glc6,6’).
3. Results and discussion
3
.1. Detection of hexosylated flavanols by targeted UPLC-DAD-ESI-IT-MS
The presence of hexosylated flavanol monomers and dimers was investigated in analyzed
wine and Merlot grape extracts. Specific molecular ions corresponding to (epi)catechin and
procyanidin dimer hexosides were targeted by ion-trap LC-MS. Respectively, five and seven
peaks eluted at different retention times were detected in the extracted ion chromatograms at
m/z 451 and m/z 739 (Appendix 5). These isomers may correspond to different substitution
positions on the flavanols (catechin or epicatechin), by different hexosides or different
conformations. Indeed, the sugar can be bound on any of the flavanol hydroxyl positions or
on the C6 or C8 position.
3
.1.1. Theoretical MS fragmentation of flavanol monomer hexosides
9

Specific MS/MS fragmentations were also searched beside the molecular ions to confirm the
monomeric flavanol hexosides structures. According to the hexose position, some theoretical
fragmentation can be described as illustrated in Figure 1. The different possible
fragmentations are then retro-diels Alder (RDA), Heterocyclic ring fission (HRF) (Gu, Kelm,
Hammerstone, Beecher, Holden, Haytowitz, et al., 2003) and different bond cleavages. The
most common detected fragment is (epi)catechin aglycone detected at m/z 289, corresponding
to the loss of hexose attached on a hydroxyl group (-162), represented by red arrows. This
fragment ion is present for each possible hexose position. In addition, other fragments from
different configurations may have the same mass and not permit to conclude on the hexose
position. For example, fragments at m/z 299 and 151 and m/z 137 and 313 obtained from
RDA and OC9/C2C3 cleavage (respectively orange and purple arrows) can be detected if the
hexose is bound on A, B or C ring (Delcambre & Saucier, 2012). NMR analysis and
molecular standards are then necessary to confirm the nature of the hexose and its position on
the molecule.
3
.1.2. Theoretical MS fragmentation of flavanol dimer hexosides
Figure 2 presents the theoretical fragmentations of a dimeric flavanol glycoside with a hexose
for instance a glucose) on the position 5 of the upper unit A ring. The loss of the glucose will
(
yield the proanthocyanidin dimer aglycone ion (m/z 577) for any position of glucose. RDA
fragmentation can occur twice on the C ring of the dimer upper unit and on the C’ ring of the
lower unit, yielding an ion fragment at m/z 587 and 435 or 587 and 273 with the loss of both
B rings (-152 x2). This fragmentation is specific of the hexose on the A ring. HRF may also
take place on the upper unit to yield a fragment ion at m/z of 451 and 287. The last classical
presented fragmentation is the quinone methide (QM) cleavage also called interflavan
cleavage. If the hexose is bound on the upper unit, like on figure 2, fragments ions at m/z of
4
49 Da and at 289 Da are formed. If the hexose is on the lower unit, ions at m/z 451 and 287
10

2
are obtained. Depending on the hexose position, other MS fragmentations are summarized in
Appendix 2. Other fragments such as m/z 587 and 721 are common for all flavanol dimer
hexosides.
3
.1.3. Characterization of flavanol dimer and monomer hexosides in grape and wine
using MS fragmentation
2
The use of MS fragmentation has allowed (Appendix 5) determination of the sugar position
on the upper or the lower unit of proanthocyanidin hexosides but not the exact hexose
position. In the case of monomeric flavanol hexosides, the fragment at m/z 289 was observed
from all isomers (Figure 3.a.) and did not indicate the hexose position. To better identify the
monomeric flavanol hexosides structures and for further semi quantification studies, flavanol
glucosides enzymatic hemisynthesis pathways were investigated.
Concerning observed dimer isomers, as presented in Figure 3.b., the principal patterns
detected in our samples were m/z 721 (-18, loss of water 18), 587, 569 (dehydration from
5
87), 451, 435, 417 (dehydration from 435), which are common for all dimer hexoside
configurations, and also at 449, 339 and 289 which are more specific of some hexose position.
2
Indeed, MS fragments at m/z 449 and 289 would be obtained from a QM, indicating that the
3
hexose position would be attached on the upper unit. MS fragment at 339 represents the loss
of the B cycle (- 110) from the previous pattern at 449 showing that the hexose was not on it.
The other possibility was that the hexose was attached on the C ring (position 3). But
fragments from dehydrations were detected and commonly occur on the C3 (Li H.-J. & M.L.,
2
007), then the hexose is attached on other hydroxyl. To conclude, the fragmentation
observed is in favor of a hexose attached on the upper unit and on the A ring (position 5 or 7)
as illustrated in Figure 2.
11

3
.2. Enzymatic production and characterization of flavan-3-ol glycosides: (+)-catechin
’-O--glucoside and catechin-7-O--glucoside
4
A high-efficiency single-step hemisynthesis of C4’OG and C7’OG, catalyzed by the
UGT71A15 glucosyl transferase, is reported. (Figure 4. a.) Previous work of Nidetsky and
Lepak showed that the UGT71A15 (GenBank: DQ103712) from apple catalyzes site-selective
5
-O--glucosylation of phlorizin, using UDPG as the glucose donor (Lepak, Gutmann,
Kulmer, & Nidetzky, 2015). Similar experimental conditions were tested for glucosylation of
+)-catechin and (-)-epicatechin. Like resveratrol, (+)-catechin was an efficient acceptor
(
substrate showing a high level of conversion after 24 hours. Two signals on the UV
chromatogram at 280 nm were observed which corresponded to a molecular ion mass at m/z
4
51 with a m/z at 289 MS/MS fragment for the two retention times. NMR analysis, in
accordance with literature (Raab, Barron, Arce Vera, Crespy, Oliveira, & Williamson, 2010),
allowed us to determine the molecular configuration of (+)-catechin 4’-O--glucoside
(
C4’OG) and catechin-7-O--glucoside (C7OG) showed in Appendix 3. Chemical shifts and
coupling constants are in accordance with those find in the literature (Raab, Barron, Arce
Vera, Crespy, Oliveira, & Williamson, 2010 ; Vihakas, Tahtinen, Ossipov, & Salminen, 2012).
On the contrary, (-)-epicatechin was not converted by the enzyme to its corresponding
glucoside. A kinetic study was undertaken to optimize the (+)-catechin concentration: 1 mM,
2
.5 mM and 5 mM concentrations were compared (Appendix 1). The reaction with 2.5 mM
reached a plateau at 24 hours for C7OG but not for C4’OG. The initial concentration of 1
mM was more effective than others based on the yield obtained. Indeed, the C4’OG amount
compared is around 40% of initial catechin concentration in the reaction medium for 1 mM
against 28% and 10% for respectively 2.5 mM and 5 mM catechin. The reaction containing 1
mM and 5 mM of catechin reached a plateau after 24 hours for C4’OG and C7’OG. The same
result is observed for C7OG yield: 25% for 1 mM and 13% and 4.5% for respectively, 2.5
12

mM and 5 mM. The higher the reagent concentration is, the lower the reaction yield is. We
can hypothesize that too much (+)-catechin can inhibit the enzyme activity. A substrate
concentration at 1 mM was then chosen for further synthesis. On figure 4.b., we compared
the overlaid extracted ion chromatograms (EIC) at m/z 451 of the enzymatic solution
containing C4’OG and C7OG and of the merlot seed extract in P-II. Based on the retention
2
times and mass spectrometry results, our standards seem to be naturally present. Indeed, MS
fragmentation depicted on figure 4.c. and e. and on figure 4.d. and f. show similar
fragmentation patterns with aglycone formation at m/z 289 for both of our compounds, even if
a coelution of different isomers cannot be completely ruled out. These monomers are the most
polar of flavanol hexosides and they are the major isomers. The less polar isomer could be
epicatechin hexosides derivatives but unfortunately, (-)-epicatechin was not a substrate for
UGT71A15.
3
.3. Study of Merlot grape composition at three different stages of development
Several isomers of monomeric and dimeric flavanol hexosides were detected in the skin and
seed extracts of the three grape varieties, Merlot, Cabernet-Sauvignon and Syrah. Comparison
is based on the assumption that compounds showing the same retention time and MS data are
identical although some isomers might co-elute. In figure 5 and Appendix 4, histograms
represent the ion intensities of targeted flavanol hexosides in seed and skin extracts at three
different stages of grape development. Since the standards for these molecules are not
available, these analyses are allow qualitative comparisons between samples rather than
quantification.
Merlot grape seed and skin extracts were first investigated. In seed, an increase of all
monomeric flavanol hexosides was observed during development; these include the
compounds eluted at 7.3 and 8.4 min, tentatively attributed respectively to C4’OG and C7OG
13

based on retention time and MS/MS fragmentation (loss of glucose), and also those eluted at
0.4, 11.3 and 11.8 min. The EIC intensities of the five monomeric flavanol hexosides
1
4
5
6
6
increased from 10 order of magnitude in P-I, to around 10 to 10 in P-II, and reached 4.10
at maturity, P-III, for the major isomer. In Merlot skins, a small amount of monomer
hexosides were detected in P-I. Both isomers of monomeric flavanol hexosides at 7.3 min and
8
.4 min were distinguished at the earlier stage. The isomer attributed to C4’OG (7.3 min) was
still present in P-II and finally no more detected at maturity.
Four flavanol dimer hexosides were detected in Merlot seed extracts and their distribution
monitored during ripening. The most abundant one, eluted at 13.2 min, accumulated in seeds
until P-II and was detected in skins in P-I and in P-III. Another isomer eluted at 8.6 min was
present in higher amount in the seeds in P-I. The other ones were detected in trace amounts in
seeds and skins in P-I (elution times at 12.2 and 13.2 min).
To compare, in Merlot seeds, five isomers of monomer hexosides and four isomers of dimer
hexosides are detected, and in Merlot skin, two monomers and three dimers in small amount.
For CS seeds, five monomers and three dimers were detected and one monomer and three
dimers in CS skin. Finally, five monomers and three dimers were observed in Syrah seed
extracts and three dimers and no monomer in skin (Appendix 3).
All the monohexosylated flavanols were detected in the different grape varieties except the
flavanol dimer hexoside eluted at 7.7 min which was only detected in our Merlot sample. To
compare tissues, in seeds, an accumulation of all flavanol monomer hexosides is measured
during maturation and the isomer at 10.4 min is the most abundant in all grape varieties. The
isomer of monomeric flavanol hexoside at 7.3 min is the second most present in CS and Syrah
seed extracts but not in merlot seed. In skin, the targeted hexosides were detected in trace
amounts with a major quantity of isomers detected during the P-I in all grape varieties.
14

The presence of flavanol hexoside monomers in Merlot seeds in P-III is in accordance with
previous studies (Delcambre & Saucier, 2012) and was also evidenced in two other grape
varieties in this work. In addition, dimeric flavanol hexosides were detected in grape seed and
skin for the first time in our experiments. Additionally, the presence of (epi)gallocatechin (m/z
4
67) and epicatechin gallate (m/z 603) hexosides were also investigated, but, their
concentrations were below their detection levels in our samples.
.4. Study of five red wine from different grape varieties
3
Analysis of five red wines from Tannat, Alicante, Merlot, Syrah and Grenache grape varieties
showed that they contain similar levels of flavanol hexosides as shown on figure 6.
Concerning flavanol monomer hexosides, the wine from the Tannat variety had the highest
number of isomers eluted at 7.3, 8.4, 10.4, and 11.3 minutes. Three flavanol monomer
hexosides were detected in Syrah and Grenache wines at 7.3, 8.4 and 10.4 min. Also, Alicante
and Merlot wines included only two isomers, at 10.4 and 11.3 min. All these flavanol
hexosides were detected in grape seed at different levels. The flavanol monomeric hexoside at
1
0.4 min is one of the most abundant in all analyzed seed grapes and wines. On the contrary,
the monomer isomer at 11.3 min was one of the most abundant compounds in wine but was
not the major one in our grape seed extracts. Finally, the monomer hexoside detected at 11.8
min in seed extracts from all grape varieties seems to be absent in our wine samples.
The same three flavanol dimer hexosides at 12.2, 13.2 and 14 min were detected in the wines,
except for the Grenache wine in which the dimer detected at 14 min was not present. The two
dimeric flavanol hexosides detected at 12.2 and 13.2 min in wine were also present in Merlot,
CS and Syrah grape seeds. But, the isomer at 14 min was absent in all grape extracts. Finally,
we detected mostly a dimer at 13.2 in Syrah, Merlot and CS seeds in P-III and in wines.
15

To conclude, our results based on similarity of retention times and MS data suggest that the
majority of flavanol hexosides detected in wines have a grape origin. Other flavanol
hexosides (such as the dimer detected at 14 min) are present in wine but not in grape. Other
biochemical pathways (production or degradations) from microorganisms (yeasts and
bacteria) are also possible. Concerning the grapes, the genetic ability to produce these
compounds seems to be shared amongst the different varieties but they were detected in
different proportions which might be a result of different genetic expressions of some
glycosyl transferase genes.
4. Conclusion
The presence of monomeric flavanol hexosides was confirmed in grape and red wines by
mass spectrometry. Proanthocyanidin dimer hexosides were successfully detected in our
samples with specific MS/MS fragmentations. The flavanol monomer and dimer hexosides
are present in wines from different grape varieties (Tannat, Alicante, Merlot, Syrah, and
Grenache) and also in Merlot, Syrah and CS grape berries. One of the flavanol monomer
hexoside was found in grapes but not in wines. All (epi)catechin hexosides concentration
increases from P-I until P-III in seeds and they decrease or are absent in skin. Standards of
(
+)-catechin 4’-O--glucoside and catechin-7-O--glucoside were obtained through an
enzymatic hemisynthesis with the use of glucosyl transferase UGT71A15. The presence of
these compounds in analyzed grape seeds, skins and wines was evidenced based on HPLC-
MS analysis. The mechanisms of formation of those hexosylated flavanols and their possible
roles in proanthocyanidin biosynthesis in grape remain to be elucidated.
Funding: This work was supported in part by a PhD grant (M.Z.) from the University of
Montpellier (Bourse école doctorale GAIA)
16

References
Ali, A., & Strommer, J. (2003). A Simple Extraction and Chromatographic System for the Simultaneous
Analysis of Anthocyanins and Stilbenes of Vitis Species. Journal of agricultural and food
chemistry, 51(25), 7246-7251.
Bogs, J., Downey, M. O., Harvey, J. S., Ashton, A. R., Tanner, G. J., & Robinson, S. P. (2005).
Proanthocyanidin synthesis and expression of genes encoding leucoanthocyanidin reductase
and Anthocyanidin Reductase in Developing Grape Berries and Grapevine Leaves. American
Society of Plant Biologists, 139(2), 652-663.
Castillo-Muñoz, N., Gómez-Alonso, S., García-Romero, E., Gómez, M. V., Velders, A. H., & I., H.-G.
(
2009). Flavonol 3-O-Glycosides Series of Vitis vinifera Cv. Petit Verdot Red Wine Grapes.
Journal of agricultural and food chemistry, 57(1), 209-219.
Corder, R., Mullen, W., Khan, N. Q., Marks, S. C., Wood, E. G., Carrier, M. J., & Crozier, A. (2006).
Oenology: Red wine procyanidins and vascular health. Nature, 444, 566.
Cui, E.-J., Song, N.-A., Shrestha, S., Chung, I.-S., Kim, J.-Y., Jeong, T.-S., & Baek, N.-I. (2016). Flavonoid
glycosides from cowpea seeds (Vigna sinensis K.) inhibit LDL. Food Science and
Biotechnology, 21(2), 619-624.
Delcambre, A., & Saucier, C. (2012). Identification of new flavan-3-ol monoglycosides by UHPLC-ESI-
Q-TOF in grapes and wine. Journal of Mass Spectrometry, 47(6), 727-736.
Downey, M. O., Harvey, J. S., & S.P., R. (2003). Analysis of tannins in seeds and skins of Shiraz grapes
throughout berry development. Australian Journal of Grape and Wine Research, 9(1), 15-27.
Dueñas, M., Sun, B., Hernández, T., Estrella, I., & Spranger, M. I. (2003). Proanthocyanidin
Composition in the Seed Coat of Lentils (Lens culinaris L.). Journal of agricultural and food
chemistry, 51(27), 7999–8004.
Foo, L. Y., & Karchesy, J. J. (1989). Polyphenolic glycosides from Douglas fir inner bark.
Phytochemistry, 28(4), 1237-1240.
Friedrich, W., & Galensa, R. (2002). Identification of a new flavanol glucoside from barley (Hordeum
vulgar L.) and malt. European Food Research and Technology, 214(5), 388-393.
Fujita, A., Soma, N., Goto-Yamamoto, N., Shindo, H., Kakuta, T., Koizumi, T., & Hashizume, K. (2005).
Anthocyanidin Reductase Gene Expression and Accumulation of Flavan-3-ols in Grape Berry.
American Journal of Enology and Viticulture, 56(4), 336-342.
Gosch, C., Halbwirth, H., Schneider, B., Hölscher, D., & Stich, K. (2010). Cloning and heterologous
expression of glycosyltransferases from Malus x domestica and Pyrus communis, which
convert phloretin to phloretin 2'-O-glucoside (phloridzin). Plant Science, 178(3), 299-306.
Gu, L., Kelm, M. A., Hammerstone, J. F., Beecher, G., Holden, J., Haytowitz, D., & Prior, R. L. (2003).
Screening of foods containing proanthocyanidins and their structural characterization using
LC-MS/MS and thiolytic degradation. Journal of agricultural and food chemistry, 51(25),
7513-7521.
Hmamouchi, M., Es-Safi, N., Lahrichi, M., Fruchier, A., & Essassi, E. M. (1996). Flavones and Flavonols
in Leaves of Some Moroccan Vitis vinifera Cultivars. American Journal of Enology and
Viticulture, 47(2), 186-192.
Kim, H.-J., Yeom, S.-H., Kim, M.-K., Shim, J.-G., Lim, H.-W., & Lee, M.-W. (2004). Nitric oxide and
prostaglandin E2 synthesis inhibitory activities of flavonoids from the barks of Ulmus
macrocarpa. Natural Product Sciences, 10(6), 344-346.
Lepak, A., Gutmann, A., Kulmer, S. T., & Nidetzky, B. (2015). Creating a Water‐Soluble
Resveratrol‐Based Antioxidant by Site‐Selective Enzymatic Glucosylation. A European Journal
of Chemical Biology, 16(13), 1870-1874.
Li H.-J., & M.L., D. (2007). Tandem Mass Spectrometry for Sequencing Proanthocyanidins. Analytical
Chemistry, 79(4), 1739-1748.
Monagas, M., Bartholomé, B., & Gomez-Cordovés, C. (2007). Updated Knowledge About the
Presence of Phenolic Compounds in Wine. Crit Rev Food Sci Nutr, 45(2), 85-118.
17

Nonaka, G.-I., Ezaki, E., Hayashi, K., & Nishioka, I. (1983). Flavanol glucosides from rhubarb and
Rhaphiolepis umbellata. Phytochemistry, 22(7), 1659-1661.
Pang, Y., Cheng, X., Huhman, D. V., Ma, J., Peel, G. J., Yonekura-Sakakibara, K., Saito, K., Shen, G.,
Sumner, L., Tang, Y., Wen, J., Yun, J., & Dixon, R. A. (2013). Medicago glucosyltransferase
UGT72L1: potential roles in proanthocyanidin biosynthesis. Planta, 238, 139-154.
Pang, Y., Peel, G. J., Sharma, S. B., Tang, Y., & Dixon, R. A. (2008). A transcript profiling approach
reveals an epicatechin-specific glucosyltransferase expressed in the seed coat of Medicago
truncatula. PNAS, 105(37), 14210-14215.
Quideau, S., Deffieux, D., Douat-Casassus, C., & Pouysegu, L. (2011). Plant polyphenols: chemical
properties, biological activities, and synthesis. A Journal of the Gesellschaft Deutscher
Chemiker, 50(3), 586-621.
Raab, T., Barron, D., Arce Vera, F., Crespy, V., Oliveira, M., & Williamson, G. (2010). Catechin
Glucosides: Occurrence, Synthesis, and Stability. Food Chemistry, 58(4), 2138-2149.
Simirgiotis, M. J., Silva, M., Becerra, J., & Schmeda-Hirschmann, G. (2012). Direct characterisation of
phenolic antioxidants in infusions from four Mapuche medicinal plants by liquid
chromatography with diode array detection (HPLC-DAD) and electrospray ionisation tandem
mass spectrometry (HPLC-ESI–MS). Food Chemistry, 131(1), 318-327.
Smaali, I., Maugard, T., Limam, F., Legoy, M.-D., & Marzouki, N. (2007). Efficient synthesis of gluco-
oligosaccharides and alkyl-glucosides by. World Journal of Microbiology and Biotechnology,
23(1), 145-149.
Tanaka, T., Kataoka, M., Tsuboi, N., & Kouno, I. (2000). New monoterpene glycoside esters and
phenolic constituents of Paeoniae radix, and increase of water solubility of
proanthocyanidins in the presence of paeoniflorin. Chemical and Pharmaceutical Bulletin,
48(2), 201-207.
Tanner, G. J., & Kristiansen, K. N. (1993). Synthesis of 3,4-cis-[3H]Leucocyanidin and Enzymatic
Reduction to Catechin. Analytical Biochemistry, 209(2), 274-277.
Terrier, N., Ollé, D., Verriès, C., & Cheynier, V. (2009). biochemical and molecular aspects of flavan-3-
ol synthesis during berry development. Grapevine Molecular Physiology & Biotechnology,
365-388.
Versari, A., Parpinello, G. P., Tornielli, G. B., Ferrarini, R., & Giulivo, C. (2001). Stilbene Compounds
and Stilbene Synthase Expression during Ripening, Wilting, and UV Treatment in Grape cv.
Corvina. Journal of agricultural and food chemistry, 49(11), 5531-5536.
Vic, G., Biton, J., Le Beller, D., Michel, J. M., & Thomas, D. (1994). Enzymatic glucosylation of
hydrophobic alcohols in organic medium by the reverse hydrolysis reaction using
almond‐β‐D‐glucosidase. Biotechnology and Bioengineering, 46(2), 109-116.
Vihakas, M., Tahtinen, P., Ossipov, V., & Salminen, J.-P. (2012). Flavonoid Metabolites in the
Hemolymph of European Pine Sawfly ( Neodiprion sertifer) Larvae | SpringerLink. Journal of
Chemical Ecology, 38(5), 538-546.
Watanabe, M. (1998). Catechins as Antioxidants from Buckwheat (Fagopyrum esculentum Moench)
Groats. Journal of agricultural and food chemistry, 46(3), 839–845.
Xie, D.-Y., Sharma, S. B., Paiva, N. L., Ferreira, D., & Dixon, R. A. (2003). Role of Anthocyanidin
Reductase, Encoded by BANYULS in Plant Flavonoid Biosynthesis. Science, 299(5605), 396-
399.
Zhao, J., & Dixon, R. A. (2009). MATE transporters facilitate vacuolar uptake of epicatechin 3'-O-
glucoside for proanthocyanidin biosynthesis in Medicago truncatula and Arabidopsis. Plant
Cell, 21(8), 2323-2340.
Figures Captions
18

Figure 1. Fragmentation pathways of (epi)catechin hexosides (adapted from Delcambre &
Saucier, 2012).
Figure 2. Fragmentation patterns for a proanthocyanidin glucoside with the glucose attached
on the A cycle of the upper unit.
Figure 3. Examples of flavan-3-ols hexosides MS2/3 fragmentations in merlot grape seed
extract from P-III in the negative ion mode with a. Hexoside (epi)catechin (from m/z 451, at
1
0.4 min), b. Hexoside (epi)catechin dimer (from m/z 739, at 12.2 min).
Figure 4. a. Enzymatic hemisynthesis of catechin 4’-O--glucoside (C4’OG) and catechin 7-
O--glucoside by the glucosyltransferase UGT71A15 b. m/z 451 EIC of C4’OG (light green),
2
C7OG (dark green) and merlot seed (pink) at the end of veraison. UV chromatogram and MS
from m/z 451 EIC of c. C7OG, c. C4’OG and molecules e. at 7.5 min and f. at 8.6 min from
merlot seed at the end of veraison.
Figure 5. EIC intensities for hexosylated monomers in purple (m/z 451) and dimers in blue
(m/z 739) from grape polyphenols extracts from seed and skin of Merlot at three different
stages (P-I, P-II and P-III). Error bars correspond to MS standard deviation for triplicate
analyses.
Figure 6. EIC intensities for hexosides monomers in purple (m/z 451) and dimers hexosides
in blue (m/z 739) from five red wines from different grape varieties.
19

20

Figure 1.
21

Figure 2.
22

Figure 3.
23

Figure 4.
24

Figure 5.
25

Figure 6.
26



Monomeric and dimeric flavanol hexosides were detected in grapes and wines by LC-MS
2
3
Specific MS fragments were obtained for these hexosides by using MS an MS


The level of momeric hexosides rised during grape development
(+)-catechin 4’-O- and 7-O- hexosides were hemi-synthesized and used as standards
27