Synthesis of quercetin 3-O-(2″-galloyl)-α-L-arabinopyranoside

Article abstract of DOI:10.1016/S0040-4039(02)02139-1

Quercetin 3-O-(2″-galloyl)-α-L-arabinopyranoside, a plant flavonol glycoside showing HIV-1 integrase inhibition, antibacterial, and antioxidant activities, was synthesized, with the glycosylation of the flavonol 3-OH being explored.

Full text of DOI:10.1016/S0040-4039(02)02139-1

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 9467–9470
Synthesis of quercetin 3-O-(2%%-galloyl)-a-L-arabinopyranoside
Ming Li,a Xiuwen Hana,* and Biao Yub,*
aState Key Laboratory of Catalyst, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, China
bState Key Laboratory of Bio-organic and Natural Products Chemistry, Shanghai Institute of Organic Chemistry,
Chinese Academy of Sciences, Shanghai 200032, China
Received 31 July 2002; revised 18 September 2002; accepted 26 September 2002
Abstract—Quercetin 3-O-(2%%-galloyl)-a-L-arabinopyranoside, a plant flavonol glycoside showing HIV-1 integrase inhibition,
antibacterial, and antioxidant activities, was synthesized, with the glycosylation of the flavonol 3-OH being explored. © 2002
Elsevier Science Ltd. All rights reserved.
with IC50=34 mM).7 Moreover, inhibition against HIV-
More than 1000 flavonol O-glycosides have so far been
isolated, mostly, from higher plants.1 The majority of
1 integrase (IC50=18.1 mg/mL) is also demonstrated,
which is the highest among ten of its congeners.8 Here
these structures (ꢀ80%) have a sugar linkage at the
3-OH,1 and over 20% possess one or more acyl groups
attached through the sugar residues, which further
enhances the structural diversity.1 While flavonol O-
glycosides play a variety of important roles in the
growth and development of plants,1,2 e.g. as inter-spe-
cies signaling molecules, they have also demonstrated a
wide range of properties which might be beneficial to
humans, such as antimicrobial, anti-cancer, and radical-
scavenging activities.1,2 In contrast to the wide occur-
rence and importance of flavonol glycosides, synthetic
studies toward this important group of natural prod-
ucts are surprisingly rare.3–5 Quercetin 3-O-(2%%-galloyl)-
we report the synthesis of flavonol glycoside 1.
Since the flavonol quercetin is commercially available,
the synthesis of glycoside 1 should simply involve pro-
tective manipulation of the hydroxyls and glycosylation
of the 3-OH of the quercetin derivatives. The 7-OH and
4%-OH of polyhydroxyflavones are known to have pref-
erential reactivity toward nucleophilic substitution, thus
using Jurd’s procedure,9 7-O-benzylquercetin 2 and
7,4%-di-O-benzylquercetin 11 were readily obtained via
controlled benzylation of quercetin pentacetate. Ketal
protection of the ortho-3%,4%-OH on 2 would leave the 3
and 5-OHs free, and a selective glycosylation on the
3-OH was then expected due to the chelation of the
5-OH with the 4-carbonyl group.3 We chose the
lipophilic cyclohexylidene group to block the catechol
with the intention of increasing the solubility of the
quercetin derivative in the glycosylation solvents
(Scheme 1). In the previous syntheses of flavonol O-gly-
a-L
-arabinopyranoside 1, representing
a
typical
structure of a flavonol O-glycoside, has been isolated
from Lasiobema japonica,6 Eucalyptus rostrata,7 and
Acer okamotoanum.8 This compound shows a potent
antibacterial activity (against E. coli B at 100 ppm
concentration)6 and an antioxidant activity (against
peroxidation of rabbit erythrocyte membrane ghost
Keywords: flavonol glycosides; glycosylation; phase transfer.
* Corresponding author.
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)02139-1

9468
M. Li et al. / Tetrahedron Letters 43 (2002) 9467–9470
Scheme 1. Reagents and conditions: (a) cyclohexanone, PhMe, p-TsOH, reflux, 61%; (b) for 4, TMSOTf or BF3·OEt2, CH2Cl2, 4
+
−
,
,
A MS; (c) for 5, Tf2O, DTBMP, CH2Cl2, 4 A MS; (d) Bu4N Br , CHCl3/aq. K2CO3 (0.18 M) (v:v, 1:1), 42°C, 91%; (e) NaOMe,
HOMe/CH2Cl2, rt; (f) DMP, CSA (10-camphorsulfonic acid), DMF, 91% (2 steps); (g) tribenzylgalloyl chloride, pyridine, DMAP,
rt, 94%; (h) PhSH, DBU, DMF, 40°C, 92%.
cosides, all the glycosylation steps employed glycosyl
bromides or chlorides under either Koenigs–Knorr con-
ditions (promoted by silver salts)3 or, later, under
phase-transfer-catalyzed (PTC) alkaline conditions.4
Although the glycosylation yields are comparable (nor-
mally lower than 50%) under these conditions, the PTC
conditions greatly improved the workup procedure.
Here, we intended first to glycosylate the phenolic
3-OH of 3 with arabinopyranosyl trichloroacetimidate
4 and sulfoxide 5,10 because glycosyl trichloroacetimi-
date and sulfoxide donors have recently been used
successfully in the glycosylation of phenols.11,12 How-
ever, attempts under various conditions failed, unavoid-
ably leading to complex products, conceivably due to
decomposition of the donors or the resulting products
under the acidic conditions. We then turned to basic
PTC conditions for glycosylation. Thus, treatment of
with the tribenzylgalloyl group. However, no selectivity
was achieved between the 5- and 2%%-OHs under various
acylation conditions (e.g., tribenzylgallic acid, DCC).
Therefore, the 5,2%%-di-acylated product 9 was prepared
(94%) under forced conditions. Selective removal of the
5-O-tribenzylgalloyl group was then attempted and
found to work well in the presence of PhSH and DBU
in DMF at 40°C, giving 10 in excellent yield (92%).14
Unfortunately, the final removal of the protecting
groups from 10 was unsuccessful under various condi-
tions;15 and the 3%,4%-cyclohexylidene could not be
removed without cleavage of the 3-O-glycosidic bond.
However, treatment of triol 11 with 2,3,4-tri-O-acetyl-
a- -arabinopyranosyl bromide 12 under the above opti-
L
mized PTC conditions, after blocking the remaining 5
and 3%-OH with benzyl groups,16 gave the expected
glycoside 13 in a satisfactory 60% yield (Scheme 2).
Subjection of 13 to deacetylation and acetonization
provided 14 with the 2-OH free (86%), which was
acylated with tribenzylgalloyl chloride to provide 15 in
86% yield. Final removal of the protective groups on 15
was straightforward. Thus, the sugar propylidene
residue was readily cleaved using 80% AcOH (86%),
then the remaining seven benzyl groups on the phenolic
hydroxyls were removed cleanly by hydrogenation over
10% Pd–C in a mixed solvent of ethanol and ethyl
acetate (1:1) under normal pressure to furnish the target
molecule 1 (94%). All the data recorded for 1 were
identical with those reported.6,17
3,5-diol 3 with 2,3,4-tri-O-benzoyl-a-L-arabinopyran-
osyl bromide 6 under conditions similar to those in the
literature (in the presence of 1.3 M aq. KOH) provided
the desired a-glycoside 7 in 36% yield. However, when
the reaction was carried out in dilute aq. KOH (0.18
M), the yield of 7 was increased to 81%. This yield was
further improved to 91% when K2CO3 (0.18 M) was
employed as a base.13 These results might be attributed
to the decrease of the side reactions of the glycosyl
bromide 6 under basic conditions, i.e. elimination to
give the glycal and cleavage of the acyl protecting
groups. No b-anomer was detected under these PTC
glycosylation conditions. After conversion of 7 into
5,2%%-diol 8, we expected to acylate the 2%%-OH selectively

M. Li et al. / Tetrahedron Letters 43 (2002) 9467–9470
9469
Scheme 2. Reagents and conditions: (a) Bu4N+Br−, CHCl3/aq. K2CO3 (0.12 M) (v:v, 1:1), 42°C; (b) BnBr, Ag2O, AgOTf (cat.),
DIPEA (cat.), DMF, 60% (2 steps); (c) NaOMe, HOMe/CH2Cl2, rt; (d) DMP, CSA, DMF, 86% (2 steps); (e) tribenzylgalloyl
chloride, pyridine, DMAP, 86%; (f) 80% AcOH, THF, 86%; (g) H2, 10% Pd–C, EtOH/EtOAc, 94%.
In summary, a structurally typical flavonol O-glycoside
with interesting bioactivities was readily synthesized
from 7,4%-di-O-benzylquercetin in seven steps and 36%
overall yield. The conventional PTC glycosylation con-
ditions have been improved by using much diluted aq.
K2CO3 (ꢀ0.1 M). The present successful conditions for
the selective removal of the phenolic acyl group (PhSH,
DBU, DMF) and the unsuccessful attempts at glycosyl-
ation of the flavonol 3-OH under acidic conditions are
worth further exploration.
Mabry, T. J.; Mabry, H. The Flavonoids; Chapman and
Hall: London, 1975.
3. (a) Vermes, B.; Farkas, L.; No´gra´di, M. In Topics in
Flavonoid Chemistry and Biochemistry; Farkas, L.;
Ga´bor, M.; Ka´llay, F., Eds.; Elsevier: Amsterdam, 1975;
pp. 162–170; (b) Farkas, L.; Vermes, B.; Miha´ly, N.;
Ka´lma´n, A. Phytochemistry 1976, 15, 1184; (c) Vermes,
B.; Chari, V. M.; Wagner, H. Helv. Chim. Acta 1981, 64,
189.
4. (a) Demetzos, C.; Skaltsounis, A.-L.; Tillequin, F.; Koch,
M. Carbohydr. Res. 1990, 207, 131; (b) Razanamahefa,
B.; Demetzos, C.; Skaltsounis, A.-L.; Andriantisiferana,
M.; Tillequin, F. Heterocycles 1994, 38, 357; (c) Demet-
zos, C.; Skaltsounis, A.-L.; Razanamahefa, B.; Tillequin,
F. J. Nat. Prod. 1994, 57, 1234.
Acknowledgements
This work was supported by the National Natural
Science Foundation of China (29925203) and the Com-
mittee of Science and Technology of Shanghai
(01JC14053).
5. Enzymatic introduction of acyl groups onto the sugar
moiety of flavonol glycosides has been reported, see: (a)
Danieli, B.; De Bellis, P.; Carrea, G.; Riva, S. Helv.
Chim. Acta 1990, 73, 1837; (b) Danieli, B.; Bertario, A.;
Carrea, G.; Redigolo, B.; Secundo, F.; Riva, S. Helv.
Chim. Acta 1993, 76, 2981.
References
6. Iwagawa, T.; Kawasaki, J.-I.; Hase, T.; Sako, S.; Okubo,
T.; Ishida, M.; Kim, M. Phytochemistry 1990, 29, 1013.
7. Okamura, H.; Mimura, A.; Yakou, Y.; Niwano, M.;
Takahara, Y. Phytochemistry 1993, 33, 557.
8. Kim, H. J.; Woo, E.-R.; Shin, C.-G.; Park, H. J. Nat.
Prod. 1998, 61, 145.
9. Jurd, L. J. Org. Chem. 1962, 27, 1294.
1. Harborne, J. B.; Baxter, H. The Handbook of Natural
Flavonoids; John Wiley & Sons: Chichester, 1999; Vol. 1.
2. (a) Harborne, J. B. The Flavonoids: Advances in Research
since 1986; Chapman and Hall: London, 1994; (b) Har-
borne, J. B. The Flavonoids: Advances in Research since
1980; Chapman and Hall: London, 1988; (c) Harborne, J.
B.; Mabry, T. J. The Flavonoids: Advances in Research;
Chapman and Hall: London, 1982; (d) Harborne, J. B.;
10. Imidate 4 and sulfoxide 5 were prepared from phenyl
3,4 - O - isopropylidene - 1 - thio - a -
L - arabinopyranoside
employing the following approaches:

9470
M. Li et al. / Tetrahedron Letters 43 (2002) 9467–9470
11. Mahling, J.-A.; Schmidt, R. R. Synthesis 1993, 325.
12. Thompson, C.; Ge, M.; Kahne, D. J. Am. Chem. Soc.
1999, 121, 1237.
13. K2CO3 was found superior to KOH in a solid–liquid
PTC glycosylation of phenols, see: Hongu, M.; Satio, K.;
Tsujihara, K. Synth. Commun. 1999, 29, 2775.
14. The deprotection of the aryl acetate using literature con-
ditions (PhSH, K2CO3, DMF, 40°C) gave 10 in 70%
yield. Here, we found that using DBU as a base instead
of K2CO3 increased the yield of 10 to 92%. See:
Chakraborti, A. K.; Nayak, M. K.; Sharma, L. J. Org.
Chem. 1999, 64, 8027.
15. Exhaustive attempts were taken following literature
methods for cleavage of ketals, see: Greene, T. W.; Wuts,
P. G. M. Protective Groups in Organic Synthesis, 3rd ed.;
John Wiley & Sons: Chichester, 1999.
NMR (300 MHz, CDCl3): l 8.01 (brs, 1 H), 7.74 (d, 1 H,
J=7.5), 7.59–7.27 (m, 20 H), 7.00 (d, 1 H, J=8.1), 6.53
(brs, 1 H), 6.46 (brs, 1 H), 5.23 (s, 4 H), 5.21 (s, 2 H),
5.08 (s, 2 H), 4.97 (d, 1 H, J=7.5), 4.19 (m, 2 H), 4.01
(m, 2 H), 3.66 (dd, 1 H, J=9.3, 3.3), 1.54 (s, 3 H),
1.37 (s, 3 H). ESI–MS: 857.5 [M+Na+]. Calcd for
C51H46O11·0.5H2O: C, 72.58, H, 5.61; Found: C, 72.84,
1
H, 5.75. 15: [h]2D4=−103.0 (c 1.62, CHCl3); H NMR (300
MHz, CDCl3): l 7.95 (brs, 1 H), 7.65–7.23 (m, 36 H),
7.01 (d, 1 H, J=8.5), 6.51 (brs, 1 H), 6.36 (brs, 1 H), 6.00
(d, 1 H, J=6.0), 5.55 (m, 1 H), 5.32 (s, 2 H), 5.24 (s, 2
H), 5.14 (s, 4 H), 5.11 (s, 2 H), 5.10 (s, 2 H), 5.03 (s, 2 H),
4.37 (m, 2 H), 4.05 (d, 1 H, J=11.3), 3.81 (d, 1 H,
J=11.0), 1.54 (s, 3 H), 1.37 (s, 3 H). ESI-MS: 1257.9
[M+H+]. Calcd. for C79H68O15: C, 75.46, H, 5.45; Found:
C, 75.17, H, 5.66. 1: [h]2D0=−96.4 (c 0.33, MeOH); IR:
1
3300, 1703, 1656, 1607, 1503 cm−1; H NMR (300 MHz,
16. The presence of phenolic hydroxyls usually makes chro-
matography on silica gel difficult due to heavy adhesion.
17. Selected data for the key compounds. 13: [h]1D9=−87.9 (c
DMSO-d6): l 12.55 (brs, 1 H), 9.82 (brs, 1 H), 9.28 (brs,
1 H), 9.21 (brs, 2 H), 8.94 (brs, 1 H), 7.71 (d, 1 H,
J=7.4), 7.49 (brs, 1 H), 7.03 (s, 2 H), 6.85 (d, 1 H,
J=8.2), 6.40 (brs, 1 H), 6.19 (brs, 1 H), 5.59 (d, 1 H,
J=6.1), 5.32 (t, 1 H, J=6.9), 5.07 (brs, 1 H), 4.88 (brs, 1
H), 3.71 (m, 3 H), 3.41 (m, 1 H); 13C NMR (75 MHz,
DMSO-d6): l 177.32, 165.24, 164.41, 156.44, 148.87,
145.68, 145.21, 138.64, 133.28, 120.55, 120.97, 119.74,
115.80, 115.60, 109.13, 104.10, 98.92, 93.71, 72.45, 69.98,
67.14, 65.23. ESI-MS: 609.1 [M+Na+].
1
1.03, CHCl3); H NMR (300 MHz, CDCl3): l 8.02 (brs,
1 H), 7.71 (d, 1 H, J=8.5), 7.59–7.32 (m, 20 H), 7.02 (d,
1 H, J=8.5), 6.53 (brs, 1 H), 6.44 (brs, 1 H), 5.80 (d, 1 H,
J=7.1), 5.53 (t, 1 H, J=8.6), 5.38–5.15 (m, 8 H), 5.08
(brs, 2 H), 3.87 (d, 1 H, J=12.9), 3.58 (d, 1 H, J=12.9),
2.17 (s, 3 H), 2.05 (s, 3 H), 1.87 (s, 3 H). ESI-MS: 943.5
[M+Na+]. Calcd for C54H48O14: C, 70.42, H, 5.25; Found:
C, 70.20, H, 5.53. 14: [h]1D9=−40.6 (c 1.38, CHCl3); 1H