Methylation of Dihydroquercetin Acetates: Synthesis of 5-O-Methyldihydroquercetin

Article abstract of DOI:10.1021/np034005w

The major products of methylation of dihydroquercetin (1, dhq) with diazomethane have been identified as 7-O-methyldhq (9), 7,3′ -di-O-methyldhq (10), 7,4′-di-O-methyldhq (11), and 7,3′,4′ -tri-O-methyldhq (2). With dhq 3,7,3′,4′-tetraacetate (6), dhq 3,5,3′,4′-tetraacetate (5), dhq 3,3′,4′-triacetate (7), dhq 7,3′,4′-triacetate (8), and dhq 3-acetate (4) the same reaction affords mainly 5-O-methyldhq 3,7,3′,4′-tetraacetate (15), 7-O-methyldhq 3,5,3′,4′-tetraacetate (13), 7-O-methyldhq 3,3′,4′-triacetate (14), 5-O-methyldhq 7,3′,4′ -triacetate (25), and 7-O-methyldhq 3-acetate (12), respectively. The methylation of 8 is accompanied by intermolecular acetyl migration from phenolic oxygen to the hydroxy group of the heterocyclic ring. 5-O-Methyldhq (28) is accessible by deacetylation of 25. 5,7-Di-O-methyldhq (29), four known methyl ethers, 13 new and three known methyl ether acetates of dhq, and six new isomers with 2,3-cis stereochemistry were spectroscopically identified.

Full text of DOI:10.1021/np034005w

1562
J . Nat. Prod. 2003, 66, 1562-1566
Meth yla tion of Dih yd r oqu er cetin Aceta tes: Syn th esis of
5-O-Meth yld ih yd r oqu er cetin
Eberhard Kiehlmann*,† and Peter W. Slade‡
Department of Chemistry, Simon Fraser University, Burnaby, British Columbia, Canada V5A 1S6
Received August 14, 2003
The major products of methylation of dihydroquercetin (1, dhq) with diazomethane have been identified
as 7-O-methyldhq (9), 7,3′-di-O-methyldhq (10), 7,4′-di-O-methyldhq (11), and 7,3′,4′-tri-O-methyldhq (2).
With dhq 3,7,3′,4′-tetraacetate (6), dhq 3,5,3′,4′-tetraacetate (5), dhq 3,3′,4′-triacetate (7), dhq 7,3′,4′-
triacetate (8), and dhq 3-acetate (4) the same reaction affords mainly 5-O-methyldhq 3,7,3′,4′-tetraacetate
(15), 7-O-methyldhq 3,5,3′,4′-tetraacetate (13), 7-O-methyldhq 3,3′,4′-triacetate (14), 5-O-methyldhq 7,3′,4′-
triacetate (25), and 7-O-methyldhq 3-acetate (12), respectively. The methylation of 8 is accompanied by
intermolecular acetyl migration from phenolic oxygen to the hydroxy group of the heterocyclic ring. 5-O-
Methyldhq (28) is accessible by deacetylation of 25. 5,7-Di-O-methyldhq (29), four known methyl ethers,
13 new and three known methyl ether acetates of dhq, and six new isomers with 2,3-cis stereochemistry
were spectroscopically identified.
Dihydroflavonols are found in many plants and have
been shown to be intermediates in the biosynthesis of other
flavonoid classes. The most common member of this family
is dihydroquercetin (1, dhq),1 which occurs in nature as free
phenol, as glycoside, and in the form of free and glycosyl-
ated phenol ethers or esters. Seven optically active methyl
ethers and four methyl ether 3-acetates of dhq have been
isolated from plants:2 dhqMe5,3,4 dhqMe7 (padmatin),
dhqMe3′,5,6 dhqMe4′,7-9 dhqMe73′, dhqMe74′,9 cis-
dhqMe74′,9 dhqMe7Ac3, dhqMe3′Ac3,10 dhqMe4′Ac3,11 and
dhqMe73′Ac3.12 All except one (cis-dhqMe74′) have the 2,3-
trans configuration, and only one of them (dhqMe5) has a
5-O-methyl group. Neoastilbin, a dhq 3-O-rhamnoside,13
and dhqAc3 are responsible for the sweetness of the herb
Tessaria dodoneifolia; dhqMe4′, dhqAc3, and dhqMe4′Ac3,
respectively, were rated by Kinghorn et al.14 to be 40, 80,
and 400 times sweeter than sucrose. Many flavonoids of
similar structures are known to be effective radical scav-
engers (antioxidants).15
ments with dhq (1), dhqAc3 (4), dhqAc353′4′ (5), dhqAc373′4′
(6), dhqAc33′4′ (7), and dhqAc73′4′ (8) are reported in this
paper.
Resu lts a n d Discu ssion
Of the four phenolic hydroxy groups of dhq, 7-OH is most
and 5-OH least acidic.19 Their reactivities toward CH2N2
decrease in the same order. Thus from dhq we obtained
mainly dhqMe7 (9), exclusively in the presence of sodium
borate, and smaller amounts of dhqMe73′ (10), dhqMe74′
(11), and dhqMe73′4′ (2) (Scheme 1a), while only dhqMe7Ac3
1
(12) was formed from dhqAc3 (4). The H NMR data of our
synthetic samples (Table 1) agree with those listed in the
literature and thereby confirm the structures initially
assigned to the natural products 9,20 1021, 11,7,9 and 12.20
However, our proton chemical shifts of cis-dhqMe74′ (11a ,
see Table 1), which was recently claimed to co-occur with
the trans-diastereomer in the stem bark of Lannea coro-
mandelia, differ from the reported values.9 We identified
11a as one of the four di-O-methyldihydroquercetins
obtained by methylation of an equimolar mixture of trans-
and cis-dhq22 and chromatographic separation from the
mono- and trimethylated coproducts. Although the four
isomers (trans-dhqMe73′, trans-dhqMe74′, cis-dhqMe73′,
and cis-dhqMe74′) were not separable from each other,
knowledge of the NMR spectra of the two 2,3-trans
structures permitted peak assignments for their respective
cis-diastereomers (Table 1). The trans-isomer predominated
in each of the PTLC fractions, indicating slow cis f trans
isomerization during and after the reaction.
The structures of the natural products have been deter-
mined by NMR and UV spectroscopy and were not verified
by independent synthesis because efficient procedures for
the regioselective alkylation of polyhydroxy-dihydrofla-
vonols at only one or two phenolic oxygens are not avail-
able. Alkyl halides and dimethyl sulfate are known to
convert dhq mainly to the tri- and tetra-O-methyl ethers
dhqMe73′4′ (2) and dhqMe573′4′ (3) in alkaline solution,
and several methylated oxidation and rearrangement
byproducts (quercetins, alphitonins, aurones, and chal-
cones) result from attack of the labile pyranone ring by
base.16 With excess diazomethane in methanol, these side
reactions are not observed, but 2 and 3 are formed in much
lower yields.16,17 We confirmed these findings and reasoned
that our recently described dhq tri- and tetraacetates18
should permit access to specific di- and monomethyl ethers
by reaction of their free phenolic OH groups with CH2N2
and subsequent deprotection. The results of our experi-
1
The H NMR spectra reveal several interesting features
of diagnostic value for the distinction between cis- and
trans-dihydroflavonols. The C-ring proton chemical shifts
of the five diastereomeric pairs listed at the top of Table 1
(1, 2, 9, 10, 11 and their cis isomers 1a , 2a , 9a , 10a , 11a )
are significantly affected by the 2,3-stereochemistry but
vary only slightly with the number of methoxy groups in
rings A and B. In each case, changing the configuration
from trans to cis shifts the H-2 resonance downfield by 0.40
ppm and H-3 upfield by 0.34-0.41 ppm; that is, cis-
structures differ from their trans-diastereomers not only
by their much smaller J 23 coupling constant (2-3 Hz vs
11-12 Hz) but also by a larger H-2/H-3 peak separation.
* To whom correspondence should be addressed. Tel: 604-291-4880.
Fax: 604-291-3765. E-mail: ekie@sfu.ca.
† Simon Fraser University.
‡ Present address: University College of the Fraser Valley, 33844 King
Rd., Abbotsford, B.C. V2S 7M8.
10.1021/np034005w CCC: $25.00
© 2003 American Chemical Society and American Society of Pharmacognosy
Published on Web 11/26/2003

Synthesis of 5-O-Methyldihydroquercetin
J ournal of Natural Products, 2003, Vol. 66, No. 12 1563
Sch em e 1. Methylation of dhq (1), dhqAc353′4′ (5), dhqAc373′4′ (6), and dhqAc33′4′ (7)
Ta ble 1. Proton Chemical Shifts of Dihydroquercetin Derivativesa
compound
#
H-2
H-3
J 23
H-6/8
H-2′
H-5′
H-6′
5-OR
7-OR
9.69
3′,4′-OR
3-OR
dhq
1
5.01
5.41
5.13
5.53
4.98
5.88
5.04
5.44
5.10
5.49
5.08
5.49
6.10
5.38
5.57
5.60
5.91
5.59
5.46
5.37
5.55
5.49
5.50
5.82
5.57
5.50
5.11
5.22
5.66
5.61
4.85
4.82
4.86
4.89
5.25
5.56
4.60
4.26
4.73
4.32
4.46
5.70
4.63
4.28
4.72
4.31
4.66
4.28
5.02
5.84
5.71
5.87
5.72
5.68
5.55
5.65
5.75
5.68
5.57
5.54
5.96
5.87
4.41
4.54
6.08
6.00
4.34
4.33
4.68
4.38
5.53
5.48
11.4
2.7
11.4
2.5
12.1
2.6
11.6
2.6
11.3
2.6
11.3
2.6
5.98/5.94
5.99/5.95
6.08/6.05
6.12/6.06
6.23/6.13
6.04/6.11
6.07/6.04
6.09/6.04
6.08/6.05
6.11/6.04
6.08/6.05
6.11/6.04
6.14/6.00
6.10/6.08
6.56/6.42
6.13/6.15
6.14/6.22
6.56/6.47
6.20/6.09
6.52/6.41
6.54/6.44
6.55/6.45
6.26/6.22
6.28/6.25
6.12/6.13
6.12/6.13
6.25/6.19
6.54/6.44
6.37/6.38
6.36/6.37
6.16/6.01
6.11/5.95
6.14/5.98
6.21/6.12
6.17/6.04
6.15/6.10
7.06
7.09
7.22
7.22
7.21
7.43
7.06
7.09
7.21
7.21
7.00
7.00
6.84
7.06
7.48
7.44
7.43
7.47
7.43
7.04
7.40
7.28
7.45
7.42
7.39
7.28
7.48
7.49
7.42
7.29
7.05
6.99
6.83
7.06
7.01
6.98
6.85
6.80
6.99
6.94
6.98
7.28
6.85
6.81
6.87
6.82
6.97
6.93
6.89
6.85
7.34
7.34
7.28
7.33
7.32
6.85
7.12
7.16
7.33
7.26
7.13
7.17
7.31
7.31
7.13
7.18
6.84
6.81
7.03
6.85
6.85
6.79
6.90
6.89
7.11
7.11
7.10
7.43
6.91
6.89
7.03
7.03
7.09
7.09
6.77
6.88
7.56
7.53
7.43
7.54
7.50
6.88
7.16
7.45
7.51
7.43
7.16
7.44
7.51
7.52
7.17
7.47
6.88
6.81
6.85
6.90
6.85
6.83
11.71
11.85
11.69
11.81
3.86
11.71
11.69
11.81
11.69
11.82
11.68
11.80
11.65
11.53
2.26
11.49
11.65
3.86
3.81
3.85
3.86
3.85
3.84
3.84
7.98/8.04
4.69
5.14
4.78
5.27
4.27
1.93
4.75
5.20
4.77
5.25
4.76
5.22
cis-dhq
dhqMe73′4′
cis-dhqMe73′4′
dhqMe573′4′
cis-dhqAc33′4′
dhqMe7
1a
2
2a
3
7a
9
9a
10
10a
11
11a
3.85
3.81
3.86
9.98
3.85
3.85
3.87
3.84
3.87
3.83
3.88
3.87
3.91
3.88
3.90
2.26
9.72
2.26
2.24
2.24
3.88
3.89
3.88
3.87
3.87
2.26
2.30
2.29
9.85
3.83
3.80
3.83
2.27/2.28
8.06
cis-dhqMe7
dhqMe73′
8.06
3.85/7.79
3.84/7.70
7.73/3.85
7.65/3.83
7.71/3.80
8.4
2.28
2.28
2.27
2.28
2.27/2.28
8.17
3.86/2.26
2.26/3.85
2.28
cis-dhqMe73′
dhqMe74′
cis-dhqMe74′
cis-dhqMe74′ (ref 9)
dhqMe7Ac3
3.2
12
13
14
14a
15
16
17
18
19
20
20a
21
22
24
25
26
27
28
28
11.8
12.3
11.9
2.6
12.1
11.9
11.9
12.1
12.1
12.1
2.8
11.9
11.9
11.7
12.4
12.1
12.1
11.9
11.8
12.1
11.9
11.8
3.2
1.95
1.95
1.99
1.93
1.98
1.96
1.95
1.96
1.97
1.97
1.91
1.98
1.97
4.41
4.54
1.98
1.99
4.26
dhqMe7Ac353′4′
dhqMe7Ac33′4′
cis-dhqMe7Ac33′4′
dhqMe5Ac373′4′
dhqMe5Ac33′4′
dhqMe5Ac37
dhqMe53′Ac374′
dhqMe54′Ac373′
dhqMe57Ac33′4′
cis-dhqMe57Ac33′4′
dhqMe73′Ac34′
dhqMe74′Ac33′
dhqMe57Ac3′4′
dhqMe5Ac73′4′
dhqMe3′Ac374′
dhqMe4′Ac373′
dhqMe5
2.26
11.51
11.51
3.86
3.86/2.25
2.25/3.85
2.28
2.27/2.28
3.91/2.26
2.26/3.94
8.16/8.11
3.88
11.37
11.37
3.82
3.80
3.81
3.85
3.80
3.79
dhqMe5b
dhqMe5 (ref 3)
dhqMe57
dhqMe5Ac3
cis-dhqMe5Ac3
29
30
30a
3.86
9.68
8.08/8.02
8.16
4.26
1.94
1.95
a
b
Solvent: acetone-d6. R ) H, Me, or Ac; J 2′6′ ) 1.8-2.1 Hz, J 5′6′ ) 8.0-9.0 Hz, J 68 ) 1.9-2.3 Hz. In acetone-d6/D2O (4:1).
Kasai et al.13 reported H-2 and H-3 chemical shift incre-
ments of comparable magnitude for the trans- and cis-
diastereomers of dhq 3-O-rhamnoside; the values for the
2R,3R and 2S,3S and for the 2R,3S and 2S,3R stereoiso-
mers, respectively, differed by considerably smaller amounts
(-0.10 to -0.12 ppm for H-2, +0.05 to -0.10 ppm for H-3).
The chemical shifts of the A- and B-ring protons of their
samples, recorded in acetone-d6 in the presence of D2O, are
within 0.10 ppm of our values for 1/1a and 9/9a , recorded
in acetone-d6 without D2O. This is also true for the spectral
parameters of dhqMe5 (28) in the presence versus absence
of D2O (Table 1); that is, the effect of water appears to be
insignificant. The 3-OH proton of the cis-isomers is ad-
ditionally deshielded (∆δ 0.4-0.5 ppm) by the anisotropy
cone of the carbonyl group, and its vicinal coupling to H-3
is relatively large (6 Hz vs 4 Hz for trans); decoupling by

1564 J ournal of Natural Products, 2003, Vol. 66, No. 12
Kiehlmann and Slade
Sch em e 2. Methylation of Dihydroquercetin 7,3′,4′-triacetate
(8)a
addition of trifluoroacetic acid collapsed the H-3 double
doublets to simple doublets. 3′-O-Methylation induces a
downfield shift of the H-2′ and H-6′ signals (∆δ 0.12-0.15
ppm) but has no effect on H-5′, while 4′-O-methylation
induces a downfield shift of the H-5′ and H-6′ (∆δ 0.12-
0.20 ppm) and a weak upfield shift of the H-2′ signal (∆δ
0.06-0.09). The five aromatic protons are virtually unaf-
fected by the C-ring geometry. We believe that our NMR
data, especially the consistency of the chemical shift
variations resulting from minor structural changes, prove
the structures assigned to the four di-O-methyldhqs 10,
10a , 11, and 11a . Unfortunately, our failure to isolate pure
samples precluded confirmation of the absolute stereo-
chemistry (assumed to be 2R,3R for 10 and 11 and 2S,3R
for 10a and 11a ) by CD spectroscopy.
Partially acetylated dihydroflavonols undergo not only
O-methylation but also ester cleavage (trans-esterification)
on treatment with CH2N2, as previously reported for simple
phenol esters.23,24 Thus dhq 3,5,3′,4′-tetraacetate (5) gave
not only the expected dhqMe7Ac353′4′ (13) but also some
dhqMe7Ac33′4′ (14), the product of 5-OAc methanolysis
(Scheme 1b). Similarly, the 5-O-methyl ether 15 formed
from the isomeric dhq 3,7,3′,4′-tetraacetate (6) was ac-
companied by five minor byproducts (Scheme 1c) resulting
from solvolysis of the phenolic ester functions (7-OAc f
7-OH, 16, and 3′/4′-OAc f 3′/4′-OH, 17) and simultaneous
or subsequent methylation (3′-OAc f 3′-OMe, 18, 4′-OAc
f 4′-OMe, 19, and 7-OAc f 7-OMe, 20).
a
Ar ) 3′,4′-(OAc)2C6H3.
Sch em e 3. Deacetylation of dhqMe5Ac373′4′ (15), dhq57Ac3′4′
(24), and dhqMe5Ac73′4′ (25)
As expected from these results for the dhq tetraacetates
5 and 6, dhq 3,3′,4′-triacetate (7) reacts with diazomethane
in MeOH slowly and exclusively at the 7-OH group; 80%
of the educt was recovered after 3 days at 4 °C. Changing
the solvent to benzene led to complete reaction and
formation of dhqMe7Ac33′4′ (14) in 88% yield as well as
small amounts of dhqMe73′Ac34′ (21), dhqMe74′Ac33′ (22),
and dhqMe57Ac33′4′ (20); that is, a few B-ring acetyl
groups were again replaced by methyl groups (Scheme 1d).
The presence of some cis-dhqMe7Ac33′4′ (14a ) and cis-
dhqMe57Ac33′4′ (20a ) was accounted for by methylation
of the (∼15%) cis-dhqAc33′4′ (7a ) formed by epimerization
of the trans-isomer during its preparation from 6.18 We
found no evidence for trans f cis isomerization under the
mild experimental conditions used for methylation.
troscopy. Pistelli et al.4 isolated it from the aerial parts of
Genista corsica and presented no experimental data to
support their structure assignment. The proton NMR
spectrum (in acetone-d6) of our synthetic sample shows low-
field aromatic proton signals at δ 7.05 (doublet, J 1.9 Hz,
H-2′), 6.84 (doublet, J 8.1 Hz, H-5′), and 6.88 ppm (double
doublet, J 1.9 and 8.1 Hz, H-6′), typical of a 3′,4′-dihy-
droxylated B-ring, the C-ring protons at δ 4.85 (doublet, J
11.9 Hz, H-2) and 4.34 ppm (double doublet, J 11.9 and
2.6 Hz, H-3), and the OH protons at δ 9.85 (7-OH), 8.16/
8.11 (3′,4′-OH), and 4.24-4.34 ppm (doublet, J 2.6 Hz,
3-OH); see Table 1, which also shows the data recorded in
acetone-d6/D2O (4:1) and the previously reported3 chemical
shifts. The hydroxylic protons were cleanly separated by
addition of cadmium nitrate.26 The H-3 signal, which is
easily overlooked due to its overlap with 3-OH, experiences
an upfield shift relative to dhq (from 4.60 to 4.34 ppm)
because the deshielding effect of the 5-OH‚‚‚OdC< hydro-
gen bond is nullified on methylation of the 5-OH group.
Several additional examples from the NMR table, e.g., 2
versus 3, 9 versus 29, 14 versus 20, 26 versus 18, and 27
versus 19, demonstrate the generality of this effect. Cross-
peaks in the NOESY spectrum at 8.16/6.84, 8.11/7.05, and
6.16/3.82 ppm verify the chemical shifts assigned to OH-
4′/H-5′, OH-3′/H-2′, and H-6/MeO-5, respectively, as well
as O-5 as the methylation site (the H-8 signal at 6.01 ppm
is not enhanced by saturation of the OMe protons). The
13C signals of C-6 (93.9 ppm) and C-8 (96.3 ppm) were
identified by correlation with the proton signals at δ 6.16
and 6.01 ppm, respectively.
The knowledge gained from the above model reactions,
combined with our experience on the deacetylation of dhq
acetates,18 suggested dhq 7,3′,4′-triacetate (8) as the most
suitable substrate for a short synthesis of dhqMe5. Unex-
pectedly, its methylation in acetonitrile led to the formation
of dhqMe57Ac3′4′ (24) and dhqMe57Ac33′4′ (20) as major
products, together with smaller amounts of dhqMe5Ac373′4′
(15) and dhqMe5Ac73′4′ (25). This outcome can be ratio-
nalized by an intermolecular acetyl migration25 from the
oxygen at C-7 of the initially formed 25 to the more
nucleophilic oxygen at C-3, with concomitant methylation
at O-7 (Scheme 2). In benzene, acetyl migration was found
to be less significant and 25 was formed as the main
product, together with five byproducts (14, 26, 27, 18, and
19) resulting from acetyl migration to O-3 from O-7 of 8
and from O-3′ and O-4′ of 8 and 25, respectively, followed
by methylation (Scheme 2). Finally, the aromatic ester
functions were cleaved by sodium sulfite in aqueous
methanol,18 providing dhqMe5 (28) from dhqMe5Ac73′4′
(25), dhqMe57 (29) from dhqMe57Ac3′4′ (24), and
dhqMe5Ac3 (30) and its cis-diastereomer (30a ) from
dhqMe5Ac373′4′ (15) (Scheme 3).
In summary, in the presence of strong base Me2SO4
converts dhq and its pentaacetate mainly to dhqMe73′4′
and dhqMe573′4′. CH2N2 is a more suitable methylating
agent for the preparation of mono- and di-O-methyldihy-
droquercetins and their acetates. It permits a limited
control of regiochemistry by appropriate choice of educt and
Foo3 extracted dhqMe5 (28) from the heartwood of Acacia
melanoxylon and characterized it by UV and NMR spec-

Synthesis of 5-O-Methyldihydroquercetin
J ournal of Natural Products, 2003, Vol. 66, No. 12 1565
(91% yield), which was purified by PTLC (blue fluorescent
band at Rf 0.57). 15: H NMR, see Table 1; NOESY correla-
solvent. The 7-OH is most and the 5-OH least reactive,
while 3-OH is usually inert. In methanol the H-bonded
5-OH is also inert, permitting exclusive 7-O-methylation
when 3′,4′-OH are protected. 5-O-Methylation succeeds in
aprotic solvents (acetonitrile or benzene) and is followed
by trans-esterification if an electron-rich oxygen, such as
3-OH, is available to accept the migrating acetyl group.
DhqMe5 has been synthesized from dhq by taking advan-
tage of these reactivity differences.
1
tions: H-6 T MeO-5/AcO-7, H-8 T H-2/AcO-7, H-3 T H-2′/
AcO-3, H-2′ T H-2/AcO-3/AcO-3′, and H-5′ T H-6′/AcO-4′; 13
C
NMR (Me2CO-d6) δ (ppm) 184.7 (C-4), 169.4 (3-OAc), 168.8 (7-
OAc), 168.6 (3′,4′-OAc), 163.6 (C-7), 162.8 (C-8a), 158.2 (C-5),
144.0/143.3 (C-3′,4′), 135.6 (C-1′), 126.3 (C-6′), 124.6/123.9 (C-
2′/5′), 108.3 (C-4a), 104.0 (C-6), 100.5 (C-8), 80.6 (C-2), 74.5
(C-3), 56.7 (OMe), 21.0 (7-OAc), 20.5 (3′,4′-OAc), and 20.3 (3-
OAc). The following minor byproducts were isolated from three
1
weak PTLC bands and identified by H NMR (Table 1): 5-O-
Exp er im en ta l Section
methylquercetin tetraacetate25ab (Rf 0.30) 1H NMR (Me2CO-
d6): δ 7.89 (dd, J 2.1/8.6 Hz, H-6′), 7.85 (d, H-2′), 7.46 (d, H-5′),
7.07 (d, J 2.0 Hz, H-8), 6.82 (d, H-6), 3.92 (s, OMe), 2.31, 2.28-
2.26 (12H, OAc); dhqMe7Ac33′4′ (14), dhqMe5Ac33′4′ (16),
dhqMe5Ac37 (17) (Rf 0.36); dhqMe57Ac33′4′ (20),
dhqMe53′Ac374′ (18), and dhqMe54′Ac373′ (19) (Rf 0.61).
7-O-Met h yld ih yd r oq u er cet in 3,3′,4′-Tr ia cet a t e (14,
d h qMe7Ac33′4′). DhqAc33′4′, prepared by refluxing a metha-
nolic solution of dhqAc373′4′ (6) for 47 h, followed by PTLC
(Rf 0.55) of the crude product,18 was found (by 1H NMR
analysis) to be a mixture of the trans-(7) and the cis-isomer
(7a ), molar ratio 6:1. A solution of 40 mg (0.093 mmol) of this
mixture in 2 mL of MeOH was methylated with excess ethereal
CH2N2 (3 days at 4 °C). The crude product (30.4 mg) consisting
of 20% 14 and 80% 7/7a was then dissolved in 1 mL of benzene
and 1 mL of ether and methylated again (3 days at 4 °C). The
yellow solid (27.6 mg) left behind after solvent evaporation
contained no more unreacted starting material. It was recrys-
tallized from MeOH to give dhqMe7Ac33′4′ (14), white needles,
mp 180-1 °C. Anal. Calcd: C 59.70, H 4.71 (calcd for
Gen er a l Exp er im en ta l P r oced u r es. Dihydroquercetin (1,
dhq) was extracted from Douglas fir bark.22,27 The partial dhq
acetates dhqAc353′4′ (5), dhqAc373′4′ (6), dhqAc33′4′ (7),
dhqAc73′4′ (8), and dhqAc3 (4) were prepared by acetylation
of dhq as described previously.18 Dimethyl sulfate (Aldrich)
was purified by extraction with ice water and 0.5 M NaHCO3
and dried over anhydrous Na2SO4. Melting points (uncor-
rected) were recorded on a Fisher-J ohns melting point ap-
paratus, and NMR spectra on a Bruker spectrometer (400
MHz). Chemical shifts are reported relative to the acetone-d6
peaks centered at δ 2.04 (1H NMR) and 29.80 ppm (13C NMR).
Evaporations were carried out on a water bath (70-80 °C) or
at reduced pressure on a rotary evaporator. Analytical thin-
layer chromatography (TLC) was performed on Merck Kiesel-
gel 60F-254 (0.2 mm) DC-Plastikfolien, and preparative thin-
layer chromatography (PTLC) on Merck Kieselgel 60HF-254/
366 (0.75 mm) with benzene/acetone (4:1 v/v) as eluant, unless
specified otherwise. Elemental analyses were run on a Carlo-
Erba elemental analyzer model 1106.
C
22H20O10: C 59.46, H 4.54). Three fractions were separated
by PTLC of the concentrated mother liquor: Rf 0.84 (1.9 mg),
equimolar mixture of dhqMe73′Ac34′ (21) and dhqMe74′Ac33′
(22); Rf 0.77 (11.7 mg), 14 and 10% cis-dhqMe7Ac33′4′ (14a );
Rf 0.51 (2.0 mg, blue fluorescent), dhqMe57Ac33′4′ (20), and
Meth yla tion of d h q (1/1a ) w ith CH2N2. An ethereal
solution of diazomethane generated from 105 mg of N-
nitrosomethylurea (1.02 mmol) and 0.5 mL of 40% KOH28 was
added to a solution of an equimolar mixture (65 mg, 0.21 mmol)
of trans- and cis-dhq (1 and 1a )22 in 3 mL of Et2O and 1 mL of
benzene. The yellow reaction mixture was stored at 4 °C for 9
days. Solvent evaporation and air-drying at 80 °C left a yellow
1
5% cis-dhqMe57Ac33′4′ (20a ); H NMR, see Table 1.
Meth yla tion of Dih yd r oqu er cetin 7,3′,4′-Tr ia ceta te (8,
d h qAc73′4′). A solution of 45.9 mg (0.107 mmol) of dhqAc73′4′-
(8)18 in 2 mL of MeCN was methylated with excess ethereal
CH2N2 (4 days at 4°C). The two major bands obtained by PTLC
of the crude product (46.2 mg) were collected and analyzed by
1H NMR: Rf 0.49 (10.2 mg) dhqMe57Ac33′4′ (20)/
dhqMe5Ac733′4′ (15) (molar ratio 4:1) and Rf 0.36 (9.4 mg)
dhqMe57Ac3′4′ (24)/dhqMe5Ac73′4′ (25) (7:1); 1H NMR, see
Table 1. Methylation of 90.0 mg (0.209 mmol) of dhqAc73′4′-
(8) in 8 mL of benzene (1 h/4 °C) gave 99.0 mg of a 3:1 mixture
of unreacted educt and dhqMe5Ac73′4′ (25), which was me-
thylated again with excess ethereal CH2N2 (14 h/4 °C). PTLC
(elution with toluene/EtOAc, 7:3) of the crude product (86.8
mg) afforded 26.0 mg of 25 (Rf 0.18); 13C NMR (Me2CO-d6) δ
(ppm) 191.2 (C-4), 168.9 (7-OAc), 168.6 (3′,4′-OAc), 164.0 (C-
7), 162.5 (C-8a), 158.3 (C-5), 143.7/143.3 (C-3′/4′), 136.8 (C-
1′), 126.8 (C-6′), 124.2/123.7 (C-2′/5′), 107.6 (C-4a), 104.0 (C-
6), 100.1 (C-8), 83.1 (C-2), 73.8 (C-3), 56.7 (OMe), 21.0 (7-OAc),
and 20.5 (3′,4′-OAc). The following minor byproducts were
detected in the 1H NMR spectra of the more mobile PTLC
fractions: dhqMe3′Ac374′ (26) and dhqMe4′Ac373′ (27) at Rf
0.63, dhqMe7Ac33′4′ (14) at Rf 0.57, dhqMe53′Ac374′ (18) and
dhqMe54′Ac373′ (19) at Rf 0.35.
1
solid residue (70 mg), which was found by PTLC and H NMR
analysis of each fraction to contain dhq (1/1a , Rf 0.22, trans/
cis 2:3), dhqMe729 (9/9a , Rf 0.34, trans/cis 5:3), dhqMe73′ (10/
10a , Rf 0.52, trans/cis 2:1), dhqMe74′ (11/11a , Rf 0.52, trans/
cis 2:1), and dhqMe73′4′ (2/2a , Rf 0.65, trans/cis 4:1) in the
approximate molar ratio 2:10:3:3:1 (overall yield 78%). 1H
NMR, Table 1; 13C NMR (Me2CO-d6), cis-dhqMe7 (9a ): δ 196.7
(C-4), 169.0/165.3 (C-5/7), 163.8 (C-8a), 146.1/145.6 (C-3′/4′),
128.5 (C-1′), 120.1 (C-6′), 102.2 (C-4a), 95.5/94.5 (C-6/8), 82.4
(C-2), and 72.7 (C-3). When a solution of 32.2 mg of pure trans-
dhq (1, 0.106 mmol) and 15.3 mg of NaBO2‚4H2O (0.111 mmol)
in 2 mL of MeOH was treated with ethereal CH2N2 (3 days at
4 °C) by the same procedure, diluted with water, neutralized
with 0.1 mL of 1 M HCl, extracted with EtOAc, and dried
(MgSO4), the evaporation residue (26.6 mg, 82%) was found
to contain only trans-dhqMe7 (9) and dhq (ratio 1:7).
P a d m a tin 3-Aceta te (12, d h qMe7Ac3). Application of the
methylation procedure described above (CH2N2 from 1.03
mmol of N-nitrosomethylurea) to a solution of 57 mg of dhqAc3
(4, 0.165 mmol) in 1.5 mL of MeOH (27 h at 4 °C) gave 17 mg
of 12 (30% yield). 1H NMR in CDCl3 was consistent with
literature;20a most of the educt was recovered.
5-O-Met h yld ih yd r oq u er cet in
3-Acet a t e
(30,
d h qMe5Ac3). A solution of 41.3 mg (0.328 mmol) of sodium
sulfite in 2 mL of water was added dropwise to a stirred
solution of 51.0 mg (0.105 mmol) of dhqMe5Ac373′4′ (15) in 2
mL of methanol. The clear yellow reaction mixture was stirred
for 22 h at 24 °C. DhqMe5Ac3 (30) crystallized (21.7 mg of
white solid after filtration, washing, and vacuum-drying over
P2O5) on evaporation of the MeOH, mp 214-6 °C (from
acetone). Anal. Calcd: C 60.23, H 4.54 (calcd for C18H16O8, C
60.00, H 4.48). 13C NMR (Me2CO-d6): δ (ppm) 184.3 (C-4),
169.5 (OAc), 165.8 (C-5), 164.9/163.8 (C-7/8a), 146.8/145.9 (C-
3′,4′), 128.9 (C-1′), 120.5 (C-6′), 115.9/115.3 (C-2′/5′), 96.6 (C-
8), 94.4 (C-6), 81.5 (C-2), 74.5 (C-3), 56.1 (OMe), and 20.4 (OAc).
The chemical shifts of H-6 (6.17 ppm) and H-8 (6.04 ppm) were
distinguished by NOESY (H-6/OMe cross-peak at δ 6.17/3.80),
7-O-Met h yld ih yd r oq u er cet in
Tet r a a cet a t e
(13,
d h qMe7Ac353′4′).12,20 Methylation of a solution of 40 mg
(0.085 mmol) of dhqAc353′4′(5)18 in 3 mL of benzene/MeOH
(1:1) with excess ethereal CH2N2 (3 days at 4 °C) gave 22.6
mg of crude product (55% yield), which was separated by
PTLC. Three bands were eluted: Rf 0.68 (10.7 mg)
dhqMe7Ac353′4′ (13), 1H NMR (Me2CO-d6), Table 1; Rf 0.49
(2.4 mg) unreacted starting material (5); Rf 0.74 (2.3 mg)
1
dhqMe7Ac33′4′ (14, H NMR, Table 1).
5-O-Met h yld ih yd r oq u er cet in
Tet r a a cet a t e
(15,
d h qMe5Ac373′4′). Methylation of a solution of 196 mg (0.415
mmol) of dhqAc373′4′ (6)18 in 4 mL of benzene with excess
ethereal CH2N2 (4 days at 4 °C) gave 184 mg of crude product

1566 J ournal of Natural Products, 2003, Vol. 66, No. 12
Kiehlmann and Slade
(3) Foo, L. Y. Phytochemistry 1987, 26, 813-817.
(4) Pistelli, L.; Giachi, I. J . Nat. Prod. 2000, 63, 504-506.
(5) Pavanasasivam, G.; Sultanbawa, M. U. S. J . Chem. Soc., Perkin
Trans. 1 1975, 612-613.
(6) Marco, J . A.; Sanz-Cervera, J . F.; Yunte, A.; Oriola, M. C. Phytochem-
istry 1994, 36, 725-729.
(7) Ruangrungsi, N.; Tappayuthpijarn, P.; Tantivatana, P.; Borris, R. P.;
Cordell, G. A. J . Nat. Prod. 1981, 44, 541-545.
(8) Ferreira, V. F.; Parente, J . P.; Pinto, M. C. F. R.; Pinto, A. V.; Porter,
B.; Silva, M. M.; Moutinho, J . L. Biochem. Syst. Ecol. 1986, 14, 199-
202.
(9) Islam, M. T.; Tahara, S. Phytochemistry 2000, 54, 901-907.
(10) Zdero, C.; Bohlmann, F.; King, R. M.; Robinson, H. Phytochemistry
1986, 25, 2841-2855.
(11) Bohlmann, F.; Czerson, H.; Scho¨neweiss, S. Chem. Ber. 1977, 110,
1330-1334.
(12) Marambio, O.; Silva, M. Bol. Soc. Chil. Quim. 1989, 34, 105-113.
(13) Kasai, R.; Hirono, S.; Chou, W.-H.; Tanaka, O.; Chen, F.-H. Chem.
Pharm. Bull. 1988, 36, 4167-4170.
(14) Nanayakkara, N. P. D.; Hussain, R. A.; Pezzuto, J . M.; Soejarto, D.
D.; Kinghorn, A. D. J . Med. Chem. 1988, 31, 1250-1253.
(15) Tu¨ckmantel, W.; Kozikowski, A. P.; Romanczyk, L. J ., J r. J . Am.
Chem. Soc. 1999, 121, 12073-12081.
(16) Hergert, H. L.; Coad, P.; Logan, A. V. J . Org. Chem. 1956, 21, 304-
310.
(17) Porter, L. J .; Foo, L. Y. Phytochemistry 1982, 21, 2947-2952.
(18) Kiehlmann, E. Org. Prep. Proceed. Int. 1999, 31, 87-97.
(19) Slabbert, N. P. Tetrahedron 1977, 33, 821-824.
(20) (a) Grande, M.; Piera, F.; Cuenca, A.; Torres, P.; Bellido, I. S. Planta
Med. 1985, 414-419. (b) Wollenweber, E.; Mayer, K.; Roitman, J . N.
Phytochemistry 1991, 30, 2445-2446.
(21) (a) Timmermann, B. N.; Hoffmann, J . J .; Shivanand, D. J .; Bates, R.
B.; Siahaan, T. J . Phytochemistry 1986, 25, 723-727. (b) Rauter, A.
P.; Branco, I.; Tostao, Z.; Pais, M. S.; Gonzalez, A. G.; Bermejo, J . B.
Phytochemistry 1989, 28, 2173-2175.
those of C-6 (94.4) and C-8 (96.6) by C-H COSY. An additional
fraction of crude product (10.3 mg), obtained by two consecu-
tive extractions (before and after acidification with HCl) of the
aqueous filtrate with EtOAc, drying (MgSO4), and solvent
evaporation, was found to contain 30, cis-dhqMe5Ac3 (30a ),
and dhqMe5 (28) in the approximate molar ratio 6:4:1 (com-
1
bined yield of 72% 30 and 10% 30a ); H NMR, Table 1.
5-O-Met h yld ih yd r oq u er cet in (28, d h q Me5) a n d 5,7-
Di-O-m eth yld ih yd r oqu er cetin (29, d h qMe57). A solution
of 39 mg (0.31 mmol) of sodium sulfite in 2.5 mL of water was
added dropwise to a stirred solution of 25.7 mg (0.058 mmol)
of dhqMe5Ac73′4′ (25) in 2.5 mL of methanol. The clear yellow
reaction mixture was stirred for 16 h at 23 °C, then acidified
with 0.3 mL of 1 M HCl. Methanol evaporation, extraction with
EtOAc, drying (MgSO4), and solvent evaporation gave 11.3 mg
of dhqMe5 (28), colorless crystals (61% yield), mp 244-5° (from
acetone/water) (lit.3 251-3 °C). Anal. Calcd: C 60.54, H 4.57
(calcd for C16H14O7, C 60.38, H 4.43). 13C NMR (Me2CO-d6/
D2O): δ (ppm) 191.5 (C-4), 166.3/165.1 (C-5/7), 163.3 (C-8a),
146.2/145.4 (C-3′/4′), 129.3 (C-1′), 120.4 (C-6′), 115.7/115.6
(C-2′/5′), 102.7 (C-4a), 96.3 (C-8), 93.9 (C-6), 83.6 (C-2), 73.3
(C-3), and 56.1 (OMe). Deacetylation of a 7:1 mixture of
dhqMe57Ac3′4′ (24) and dhqMe5Ac73′4′ (25) (9.4 mg, Rf 0.36)
(see above: methylation of dhqAc73′4′ in MeCN) by the same
procedure gave 28 as minor product and dhqMe57 (29); 1H
NMR, see Table 1. On addition of trifluoroacetic acid, the δ
4.26 doublet (J 2.6 Hz, 3-OH) shifted off-range and the δ 4.38
double doublet (J 11.9 and 2.6 Hz, H-3) collapsed to a doublet.
Ack n ow led gm en t. We thank the Dean of Science for
financial support and Ms. Tracey for recording the NMR
spectra.
(22) Kiehlmann, E.; Li, E. P. M. J . Nat. Prod. 1995, 58, 450-455.
(23) Nierenstein, M. J . Am. Chem. Soc. 1930, 52, 4012-4013.
(24) Bredereck, H.; Sieber, R.; Kamphenkel, L. Chem. Ber. 1956, 89, 1169-
1176.
Refer en ces a n d Notes
(25) (a) Kubota, O.; Perkin, A. G. J . Chem. Soc. 1925, 127, 1889-1896.
(b) J urd, L.; Rolle, L. A. J . Am. Chem. Soc. 1958, 80, 5527-5531. (c)
Simpson, T. H.; Wright, R. S. J . Org. Chem. 1961, 26, 4686-4689.
(26) Suzuki, Y.; Anazawa, I. Bunseki Kagaku 1982, 31, 498-503.
(27) Hergert, H. L. In The Chemistry of Flavonoid Compounds; Geissman,
T. A., Ed.; MacMillan: New York, 1962; Chapter 17, pp 575-579.
(28) (a) Arndt, F. Organic Syntheses; Wiley: New York, 1943; Collect. Vol.
2, pp 165-167. (b) Arndt, F.; Amstutz, E. D.; Myers, R. R. Organic
Syntheses; Wiley: New York, 1943; Collect. Vol. 2, pp 461-464.
(29) O¨ ksu¨z, S.; Topc¸u, G. Phytochemistry 1992, 31, 195-197.
(1) In the abbreviated names, such as dhqMe7Ac3 for 7-O-methyldihy-
droquercetin 3-acetate (12), the numbers conform to standard fla-
vonoid nomenclature (see Scheme 1, structure of dhq) and designate
the positions of the oxygen atoms of dihydroquercetin (dhq), to which
the methyl (Me) and acetyl (Ac) substituents are attached.
(2) (a) Bohm, B. A. In The Flavonoids; Harborne, J . B., Mabry, T. J .,
Mabry, H., Eds.; Chapman and Hall: London, 1975; Chapter 11, pp
591-592. (b) Bohm, B. A. In The Flavonoids: Advances in Research
since 1980; Harborne, J . B., Ed.; Chapman and Hall: London, 1988;
Chapter 9, pp 378 and 383. (c) Bohm, B. A. In The Flavonoids:
Advances in Research since 1986; Harborne, J . B., Ed.; Chapman and
Hall: London, 1994; Chapter 9, p 424.
NP034005W