TETRAHEDRON
LETTERS
Tetrahedmn Letters 40 (1999) 5869-5872
Pergamon
Structure of a N e w Xanthyfium Salt Derivative
Nour-Eddine Es-Safi "t, Christine Le Gueraev&, Benogt Lab-rbe, H$1~ne Fulerand,
V&roaique Cheyuier sad Michel Moutouaet
ISW-INRA lnstitut des Preduits de la Vigne, Unit~de RecherchesBiopolymheset Ar0mm,
2 Place Viala, 34060Montpelliercedex2, France.
t l~coleNormaleSup~rietwe,~
de Chimie Organiqueet d'Etudes Physicochimiques,
AvenueOued Akreuch, B.P. 5118 Tak_~___mjmRabat, Marec.
Received 28 April 1999; accepted 15 June 1999
Almraet. A new polyphanoliccompoundwith a xanthylium skeleton has been synlhesised from (+)catechin and glyonylicacid. Its structural elucidationwas achievedby UV, MS and NMR spectrmcepies.
Its formation involvedglyoxylicacid-mediated dimerisation of (+)-~_t~in I 8ivin8 thus ~mpoued 2,
followed by cyclisation to a xanthene derivative 3 which by oxidation afforded the xanthylimn
compound 4. The detection and structmal detmnination of the xanthene compound confirmed the
postulated mechanic. © 1999 Published by Elsevier ScienceLtd. All rights reserved.
Fiavanols are polyphenolic compounds found in many plants, fruits and beverages such as fruit juices, beer
and wine. They have attracted much attention in relation to their physiological activities and their role has
become an important issue in the reiatiorLqhip between health and hmmn diet. During conservation, phenolic
compounds usually undergo progressive changes which affect sensorial properties like colour, taste end colloidal
stability 1'2. Various mechanisms have been suggested to explain such transformations. Processes involving either
direct condensation between fiavanols and anthocyanins giving rise to compounds with a yellow brown hue 3'4
5-11
(xanthylium salts) or reactions mediated by acetaldehyde with the formation of violet pigments
have been
studied in model solution systems. While the process involving acetaldehyde is well documented, little is known
about the formation and the structures of xanthyllum salts. However, it is believed that xanthylium nuclei might
occur in condensed tannins (flavanol polymers), and that such chromophores may contn'bute to the high
absorbanee o f wine pigments in the 400-500 ran region4'12'13. In this paper, we describe the synthesis, the
isolation and the chemical structure analysis of a new yellowish xanthyfium compound named NJ2 4 formed by
interaction between (+)-cateehin I and glyoxylie acid (figure 1).
OH
HO.. t-~.. L
_|/~,.
*
~OH
S
/OH
HO
°'T""
y
v
HO"
H
0
-o.
,
1'
5'
OH
1
I-I0" ~f~
OH
OH
OH HO
2
F~,ure 1
Fax. : 33 4 99612683
e-mail : msafi~msam.inra.fr
0040-4039/99/$ - see front matter © 1999 Published by Elsevier Science Ltd. All rights reserved.
S0040-4039(99)01156-9
PII:
OH
OH
4
5870
Incubation of (+)-catechin 1 with tartaric acid in iron catalytic medium yielded colourless compounds
exhibiting maxima at 280 nm and yellowish compounds with maxima in the region of 440 - 460 nm as reported
14"15
.
. .
14
.
earfier
. As previously indicated , the obtained colourless compounds consisted of (+)-catechin units
bridged by a methine carboxylic acid group (compound 2). Formation ofthe yellowish compounds was rchtcd to
the disappearance of the colourless ones suggesting that the former were probably formed by evolution and
further rearrangement of the latter. This prompted us to prepare them livm individually isolated colourless
products.
Thus, the major colourless dimer 2 was isolated by high performance liquid chromatography at the
semipreparative scale and was further incubated in pH 3.5 solution, and appearmr~ of two new compounds
referred as to NJ2 and NJ3, initially absent in the mixture were observed. Their retention times and UV-visl~ole
spectra showing absorption maxima around 440 and 460 rim, respectively are the same as those of pLements
formed in reaction between (+)-catechin and tartaric acid in iron catalysed medium15.
LC/ESI-MS analysis of the major yellowish compound NJ2 both in positive and negative ion modes
showed m/z values at 615 and 617, resp~ively. The fact that the molecular weight of compound NJ2 was 20
mass units lower than that of the coiourless dimer 2 ([M-H]" at m/z = 635), suggested that the yellowish
compound may be formed by a dehydration followed by an oxidation process. The dehydrated product (Mr -618) was actually detected by LC/MS analysis at m/z 617, in the negative ion mode. The loss of a water
molecule may be achieved either between the carboxyl group and a neighbouring hydroxyl group like those at
the 7 position of the A or A' rings yielding, after oxidation, a lactonised product, or between the two 7-OH
groups, giving a xanthylium salt 4 after an oxidation process.
The UV-vi~'ble spectrum of NJ2 showed two maxima at 273 and 444 ran in addition to a shoulder at 308
nm. These maxima were I~hocbromically shifted to 283, 496 and 325 nm by addition of NaOH (figure 2), as
reported earlier for xanthylium salts4'13'16.
2.5
i
it
2
1.5
1
0.5
0
250
300
350
400
Wavelength
450
500
550
600
(am)
Figure 2: UV-Visible spectra o f compounds NJ2, NJ3 and NJ2 + NaOH.
The fragment ions at 571,463 and 419 obtained by mass spectroscopy, in the negative ion mode, can be
atm'buted respectively to the loss of carboxyl residue (-44 mass units), the loss of a hydroxyvinylphenol group
(CffhO3)obtained by retro-Diels-Alder fission (-152 mass units) and finally the ion obtained after both
fragmentations (-196 mass units), in agreement with the proposed structure 4.
The structure of NJ2 was further elucidated fiom IH and I~C chemical shift a s s ~ n e n t s (Table 1).
Unlike that of its colourless precursor which was reported to give two catechin systems14, the H spectrum of
compound N J2 presented only 8 protons, corresponding to one cat3echin spin system, thus indicating the
existenceofsome symmetry in the structure. This was confirmed by 1D CNMRspcctrum, which showed only
5871
17 signals (15 corresponding to one catachin skeleton and 2 constituting the bridge moiety). Besid~l,14the
disappearance of the proton signal previously attributed to that of the mathine bridge was observed , in
agreement with the proposed oxidised structure.
Position
2C, 2C'
3C, 3C'
4¢xC, 44zC'
415C, 4[IC'
4aA, 4aA'
5A, 5A'
6A, 6A'
7A, 7A'
8A, 8A'
8aA, 8aA'
9D
10
~ IH (ppm); m; J (Hz)
5.16; broad d; J = 3.3
4.18; m
2.46; dd; J = 3.6, 17.35
2.66; dd; J = 3.3, 17.35
6.90; s
-
I'B, I'B'
-
2'B, 2'B'
3'B, 3'B'
4'B, 4'B'
5'B, 5'B'
6'B, 6'B'
6.66; broad s
6.65; broad d; J = 8.0
6.54; broad d; J = 8.0
5 13C
82.7
64.4
24.5
24.5
107.9
170.5
95.6
156.8
104.8
154.9
149.4
166.1
128.9
113.5
145.2
145.3
115.7
117.0
Table 1. IH and 13C assignements of compound 4 in DMSO-de--TFA (9:1)
The signals which resonate at 2.46, 2.66, 4.18 and 5.16 ppm could be readily assigned to H-4, H-3 and
H-2 of the C and C' ~
while those located at 6.65 and 6.54 ppm were assigned to H-5' and H-6' after COSY
exp~.
Assignem~ of the two remalnln~ signals (6.66 and 6.90 ppm) was achieved using TOCSY
expr~i,m~. The broad singiet located at 6.66 ppm with correlates with both H-5' and H-6' was thus atm'imted
to H-2' while the singiet at 6.90 ppm which gives no scalar correlation was assigned to H-6.
The ROESY spectnan also showed that protons H-2 and H-3 gave correlations with only two aromatic
protons : one at 6.66 ppm which was already attrihaed to proton H-2', and the other at 6.54, which can be either
H-5' or H-6'. Since the latter is in a more favourable position to give such coupling, the signal at 6.54 ppm was
attn'buted to H-6", whereas the broad aromatic doublet resonating at 6.65 ppm was assigned to H-5'.
After the proton resonances had been assigned, all the corresponding carbons were attn'buted from the
short-range HSQC ~ .
The assignment of quaternary carbons was obtained from a long-range HMBC
experiment. On the basis of ~ s e amlyses, attrflmtions of the various proton and carbon chemical shiRs were
achieved, allowing to establish that the synthesised compound NJ2 was xanthylium salt 4.
From a mechanistic point of view, the formation of compound 4 may result from ~lisation of the
colourless dimer giving a xanthene structure as previonsl~y reported for 9-methyl-xalRhene17. The olRained
~
3 is tbe~ oxidised to xanthylhan salt as sbown in figure 3. The detection ofthe ~Anthene 3 compound
(Mr: 618) among the products formed from NJ2 precursor 2 by LC-MS analysis confnm~ thus the proposed
mechanism. Moreover the intermediate ~
derivative was obtained by reduction of the xanthylium NJ2 and
its structure elucidated by UV, MS and NMR spectroscopies.
5872
4-
HO,/-~O
2
• ,1 / / " " O H
03
HO f '~'O
4
Figure 3: Mechanism ofxanthylium 4 formation from the colourless dimer 2 via the xanthene 3 derivative.
Com~ar~ to the results obtained with the xanthylinm salt 4, an additional singlet was observed at 4.86
ppm in the H NMR spectrum and atm~outed to the proton H-9. This proton correlated, in HSQC e ~ t ,
with a carbon located at 34.45 ppm which was then attnl~uted to C-9. In the HMBC experiment spectrum,
correlations with C-7, C-8 and C-10 were observed, confirming the structure of the xanthene compound and
thus offering another argument to support the proposed structure 4 for the yellow compound NJ2.
The formation of such yellow xanthylium compounds in wine-like model solutions suggests their poss~le
contributions in colour evolution and browning observed during conservation and ageing of grape derived foods.
In addition, more polymerised compounds where xanthene and xanthylium nuclei are incorporated were also
detected in model solution system containing (+)-catechin and glyoxylic acid, aRer 24 hour incubation. This
indicates the implication of such derivatives in the polymeric pigments respons~le for the high absorption around
450 nm formed during wine ageing.
Our results also indicate that other reaction pathways conm'butdng to browning compete with
polycondensation reactions and offer new information and support to the conm'bution of xanthylium salts in
colour evolution and browning. They finally open perspectives for fiLrtherinvestigations of similar compounds. A
number of properties such as temperature stability and copigmentation, in addition to their poss~le use as food
colorants, remain of a high interest.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
Somers, T. C. Phytochemistry 1971, 10, 2175-2186.
Hnslam,E. Phytochemistry 1980, 19, 2577-2582.
Liao, H.; Cai, Y.; Haslara, E. J. Sci. FoodAgric. 1992, 59, 299-305.
Jurd, L. and Somers, T. C. Phytochemistry 1970, 9, 419-427.
Es-Safi, N.; Fulcrand H.; Cheynier V.; Moutounet, M. J. Agric. Food Chem. 1999, 47, 2096-2102.
Timberlake, C. F. and Bridle, P. Am. J. Enol. Vitic. 1976, 27, 97-105.
Baranowski, E. S. and Nagel, C. W. J. Food Sci. 1983, 48, 419-421.
Bakker, J., Picinelfi, A. and Bridle, P. Vitis 1993,32, 111-118.
Escribano-Bailon, T., Dangles, O., and Brouillard, R. Phytochemistry 1996, 41, 1583-1593.
Fraug3la-Arich~E. M., G u f r ~ M. T., Rivas-GonT.qlo,J. C. and Santos-Buelgeg C. J. Agric. Food Chem.
1997, 45, 2262-2266.
Rivas-Gonzalo, J. C., Bravo-Haro, S. and Santos-Buelga, C. J. Agric. Food Chem. 1995, 43, 1444-1449.
Jurd, L. Am. J. Enol. Vitic. 1969, 20, 191-195.
Hrazdina, G. and BorzelL A. J. Phytochemistry 1971, 10, 2211-2213.
Fulcrand, H., Cheynier, V., Oszmians~ J. and Moutounet, M. Phytochemistry 1997, 46, 223-227.
Os27ninrmki~J., Cheynier, V. and Moutounet, M. J. Agric. Food Chem. 1996, 44, 1712-1715.
Dangles, O. and Brouillard, R. NewJ. Chem. 1994, 18, 287-294.
White, J. and Foo, L. Y. Tetrahedron Lett. 1990, 31, 2789-2792.