Radical scavenging activity of dicaffeoyloxycyclohexanes: Contribution of an intramolecular interaction of two caffeoyl residues

Article abstract of DOI:10.1016/j.bmc.2005.04.049

Six regio- and stereoisomers of dicaffeoyloxycyclohexanes and 2,4-di-O-caffeoyl-1,6-anhydro-β-d-glucose were synthesized as model compounds of dicaffeoylquinic acids, and their radical scavenging activity was evaluated by DPPH (2,2-diphenyl-1-picrylhydraz

Full text of DOI:10.1016/j.bmc.2005.04.049

Bioorganic & Medicinal Chemistry 13 (2005) 4191–4199
Radical scavenging activity of
dicaffeoyloxycyclohexanes: Contribution of an
intramolecular interaction of two caffeoyl residues
Shizuka Saito, Satoshi Kurakane, Miyuki Seki, Eiji Takai,
Takanori Kasai and Jun Kawabata*
Laboratory of Food Biochemistry, Division of Applied Bioscience, Graduate School of Agriculture, Hokkaido University,
Kita-ku, Sapporo 060-8589, Japan
Received 2 March 2005; revised 13 April 2005; accepted 13 April 2005
Available online 12 May 2005
Abstract—Six regio- and stereoisomers of dicaffeoyloxycyclohexanes and 2,4-di-O-caffeoyl-1,6-anhydro-b-D-glucose were synthe-
sized as model compounds of dicaffeoylquinic acids, and their radical scavenging activity was evaluated by DPPH (2,2-diphenyl-
1-picrylhydrazyl) and ABTS (2,20-azinobis(3-ethylbenzthiazoline-6-sulfonic acid) diammonium salt) radical scavenging tests. Both
DPPH and ABTS radical scavenging reactions of these compounds consisted of two different steps. In the first step, catechol moi-
eties of the caffeoyl residues were rapidly converted to o-quinone structures and no significant difference in the reactivity was
observed among the tested compounds. In the second step, however, the rate of the reaction increased as the intramolecular distance
of the two caffeoyl residues decreased. A novel intramolecular coupling product, which could scavenge additional radicals, was iso-
lated from the reaction mixture of trans-1,2-dicaffeoyloxycyclohexane and DPPH radical. The result suggests that the second step of
the radical scavenging reaction is arising from an intramolecular interaction between the two caffeoquinone residues to regenerate
catechol structures, and that the closer their distance is, the more rapidly they react. The radical scavenging activity of natural di-
caffeoylquinic acids in a biological aqueous system might also depend on the positions of caffeoyl ester groups.
ꢀ 2005 Elsevier Ltd. All rights reserved.
1. Introduction
In the present study, we synthesized six regio- and ster-
eoisomers of dicaffeoyloxycyclohexanes and 2,4-di-O-
caffeoyl-1,6-anhydro-b-D-glucose as model compounds.
Their radical scavenging activity was compared with
that of cyclohexyl caffeate by DPPH (2,2-diphenyl-1-picr-
ylhydrazyl) and ABTS (2,20-azinobis(3-ethylbenzthiazo-
line-6-sulfonic acid) diammonium salt) radical
scavenging tests to examine whether the distance of
intramolecular caffeoyl residues would affect the total
radical scavenging activity. In addition, we report the
isolation of an intramolecular coupling product formed
by the oxidation of trans-1,2-dicaffeoyloxycyclohexane
with DPPH radical, and propose the radical scavenging
mechanism of dicaffeoyloxycyclohexanes via an intra-
molecular coupling reaction of the caffeoyl residues.
Caffeoylquinic acids are widely distributed in the plant
kingdom and known as potent antioxidants.1–6 In a pre-
vious paper, we reported the antioxidant activity of
mono-, di-, and tricaffeoylquinic acids isolated from
burdock.1 Their activity increased in proportion to the
number of caffeoyl residues. It has also been reported
that three dicaffeoylquinic acids (1,3-, 3,4-, and 4,5-di-
O-caffeoylquinic acid) showed similar antioxidant activ-
ity using the b-carotene/linoleate method.2 However,
since caffeic acid is known to be quite prone to oxidative
dimerization,7–12 intramolecular coupling reactions be-
tween two adjacent caffeoyl residues may occur during
the oxidation reactions of dicaffeoylquinic acids and
could influence their total antioxidant potency. Thus,
we have been interested in the relationships between
the esterified position of caffeoyl residues in dicaffeoyl-
quinic acids and their radical scavenging activity.
2. Results and discussion
Six regio- and stereoisomers of dicaffeoyloxycyclohex-
anes (3–8), cyclohexyl caffeate (9), and 2,4-di-O-caf-
feoyl-1,6-anhydro-b-D-glucose (10) were synthesized.
*
Corresponding author. Tel./fax: +81 11 706 2496; e-mail: junk@
chem.agr.hokudai.ac.jp
0968-0896/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2005.04.049

4192
S. Saito et al. / Bioorg. Med. Chem. 13 (2005) 4191–4199
Figure 2. Time course of reduction of DPPH radical at OD 517 nm by
C13aa (}), T12 (d), C12 (s), T13 (m), C13 (n), T14 (j), C14 (h) and
CC (·). Concentrations of tested compounds are 4 lM except for CC
(8 lM).
Figure 1. Structures of dicaffeoyloxycyclohexanes, 2,4-di-O-caffeoyl-
1,6-anhydro-b-D-glucose, cyclohexyl caffeate and trolox.
Compound 10 was prepared as a model compound of
1,3-di-O-caffeoylquinic acid (cynarin), which bears two
caffeoyl residues in 1,3-diaxial orientation.13 The struc-
tures of dicaffeoyloxycyclohexanes were confirmed by
the comparison of the spectral data with those in the lit-
erature.14 Compound 10 was synthesized from 1,6-anhy-
dro-b-D-glucose. From the HMBC spectrum of 10, H-2
(dH 4.66) and H-4 (dH 4.76) of the glucose correlated
with each carbonyl carbon in acyl residues. Therefore,
the caffeoyl residues were determined to be esterified
with 2 and 4 position of 1,6-anhydro-b-D-glucose. The
structures of these compounds are shown in Figure 1.
The time-dependent decay of the absorbance of the
DPPH radical by an addition of the test compounds in
40% buffer/EtOH is presented in Figure 2. The DPPH
radical scavenging reactions of all test compounds con-
sisted of two successive but different steps. The absor-
bance rapidly decayed over 0–0.5 min in the first step
followed by a much slower decrease of the absorbance
in the second step (5–30 min). There was no significant
difference among the tested compounds in the reactivity
of the first step. In the second step, however, the rate of
the reaction decreased in the order of C13aa > T12,
C12 > C13 > T13, C14, T14. Total radical consumption
of C13aa, T12, C12, and C13 was apparently larger than
that of 2 equiv of CC. On the other hand, T13, C14, and
T14 only scavenged DPPH radical in the same rate as
2 equiv of CC. Interestingly, the second step of the rad-
ical scavenging reaction accelerates in the presence of
water (Fig. 3).
Figure 3. Time course of reduction of DPPH radical at OD 517 nm by
T12 in aqueous ethanol. 60% water in ethanol (s), 40% water in
ethanol (d), 20% water in ethanol (n), 10% water in ethanol (m) and
ethanol (h).
compounds also consisted of two steps. The tendency
of the reactivity of the isomers was quite similar to that
obtained in the DPPH radical scavenging reaction. The
decrease in the absorbance of ABTS radical by the di-
caffeoyl derivatives in the first step was same as 2 equiv
of trolox, indicating the quantitative formation of bis-o-
quinone in this step.
ABTS radical scavenging activity was tested in the aque-
ous solution, and the result is shown in Figure 4. Trolox,
which is known to scavenge two radicals, was used as a
positive control (Fig. 1). As was seen with DPPH radi-
cal, the ABTS radical scavenging reaction of the tested
The formation of bis-o-quinone (11) from T12 was con-
firmed by H NMR analysis of the DPPH radical and
T12 mixture in acetonitrile-d3. The coupling constant
of H-50 (dH 6.42, d, J = 10.3 Hz) and H-60 (dH 7.39, d,
J = 10.3 Hz) of 11 was larger than that of H-50
1

S. Saito et al. / Bioorg. Med. Chem. 13 (2005) 4191–4199
4193
Figure 5. Structures of compounds 12, 13, and 14.
(11) is 436, 12 is indicated to be a methanol adduct of
T12. Appearance of a singlet signal at dH 3.48 in the
1H NMR spectrum supports the presence of an addi-
tional methoxy group. Two sets of doublet signals due
to trans double bonds at dH 5.91 and 7.87
(J = 15.4 Hz) and 6.35 and 7.20 (J = 16.1 Hz) indicate
that the side chains of the two caffeoyl residues remained
unchanged. The HMBC spectrum showed correlations
between C-20 and H-500, and C-30 and H-500, suggesting
that the two caffeoyl residues are linked at C-20 and C-
Figure 4. Time course of reduction of ABTS radical at OD 734 nm by
C13aa (}), T12 (d), C12 (s), T13 (m), C13 (n), T14 (j), C14 (h), CC
(·) and trolox (+). Concentration of tested compounds are 4 lM
except for CC and trolox (8 lM).
(dH 6.73, d, J = 8.2 Hz) and H-60 (dH 6.88, dd, J = 8.2,
2.0 Hz) of T12, as well as the low-field shift of H-60,
and the high-field shift of H-20 and H-50 in 11 compared
to T12 indicated the formation of the o-quinone moiety
in 11.
1
500 in two aromatic rings. On the basis of H and 13C
NMR, HMQC, and HMBC spectral data, the structure
of 12 was deduced to be as shown in Figure 5. Stereo-
chemistry of 12 was tentatively determined by the
NOESY experiment.
These results indicate that the reaction rate of the sec-
ond step increases as the intramolecular distance of
the two caffeoyl residues decreases. Thus, the second
step seems to arise from an intramolecular interaction
of the two caffeoyl residues. There is no difference in
the distance of the intramolecular caffeoyl residues in
C12 and T12, since two caffeoyl residues are oriented
in the gauche conformation in both isomers. On the
other hand, C13, T13, C14, and T14 seem to exist as
their extended conformers of diequatorial (C13, T14)
and axial-equatorial (T13, C14) orientations. However,
when the cyclohexane ring flips in C13, the intramolec-
ular caffeoyl residues become close enough to interact
with each other. In contrast, the caffeoyl residues of
T13, C14, and T14 are too far to interact with each other
even if the ring flipping occurs, which accounts well for
the similar reactivity of these isomers as 2 equiv of CC.
The highest radical scavenging activity of C13aa would
reasonably be explained by the fixed conformation of
the two caffeoyl residues into 1,3-diaxial manner in
which they could interact with each other easily.
The FD-MS of 13 exhibited the [M]+ peak at m/z 482,
indicating that the molecular weight of 13 was 14 mass
units larger than that of 12. The H NMR spectrum of
1
13 was very similar to that of 12, except for the presence
of signals of an ethoxy group at dH 1.11 (3H, t,
J = 7.0 Hz), 3.62 (1H, dq, J = 9.0, 7.0 (q) Hz) and
3.87(1H, dq, J = 9.0, 7.0 (q) Hz) in place of the methoxy
group at dH 3.48 in 12. Hence, the structure of 13 was
concluded to be the ethoxy analog of 12.
The DPPH radical scavenging activity of T12 and 12 in
40% H2O/MeOH was compared with DL-a-tocopherol,
which scavenges two molecules of DPPH radical. The
decrease in the absorbance of DPPH radical by T12 in
the first step corresponded to that by 2 equiv of DL-a-
tocopherol. In the second step, radical scavenging reac-
tion reached plateau at 2 h and the decrease in the
absorbance coincided with 2.5-fold of that by DL-a-
tocopherol. The result suggests that T12 rapidly scav-
enged four radicals in the first step, and then slowly
consumed additional five radicals in the second step. On
the other hand, 12 consumed DPPH radicals correspond-
ing to 1 equiv of DL-a-tocopherol. It seems that 12 is de-
rived from an addition of methanol to T12 bis-o-quinone
(11) as discussed later, and hence, 4 equiv of DPPH rad-
icals should be consumed during its production from T12
in the first step. Compound 12 scavenged approximately
two radicals, which suggests that 12 partly contributes
To confirm that intramolecular couplings of the caffeoyl
residues actually occur, we attempted to isolate the oxi-
dation products from the reaction mixture. Compounds
12 and 13 were isolated from the mixture of T12 and
DPPH radical in 40% H2O/MeOH and 40% H2O/EtOH,
respectively.
Compound 12 showed the [M]+ peak at m/z 468 in FD-
MS. Since the molecular weight of T12 bis-o-quinone

4194
S. Saito et al. / Bioorg. Med. Chem. 13 (2005) 4191–4199
to the second step of the radical scavenging reaction of
T12.
one and cysteine at C-2.15,16 On the other hand, it was
speculated that a water molecule attacks at C-6 of caffeo-
quinone.17 In this study, only the C-6 alcohol adduct
was isolated, and neither C-2 nor C-5 adduct was de-
tected. However, the reason why the position of nucleo-
philic attack differs by nucleophiles remains to be
clarified. The stereoselective introduction of the C-600
methoxy group is probably a result from the conforma-
tion of T12 in the reaction solution and from the relative
stability of the resultant functionalized macrocyclic ring.
Previously, we showed that protocatechuquinone methyl
ester undergoes a nucleophilic attack by alcohol mole-
cules and thiol compounds, followed by a regeneration
of the catechol structure.18,19 However, in the case of
T12, an intramolecular Michael addition preferentially
occurs due to the presence of a reactive quinone in the
vicinity, without reproducing the catechol structure.
Then subsequent aromatization leads to a regeneration
of the catechol structure of the A-ring to form 12a, which
is equilibrated with more stable 12 in the reaction solu-
tion. The regeneration of the catechol structure of the
B-ring of 12a is unlikely to occur by steric disfavor of
the generation of the highly functionalized12 metacyclo-
phan ring. When an excess amount of radical exists in
the reaction mixture, 12a would scavenge two additional
radicals to yield 12b. Then, a methanol molecule attacks
C-30 quinone carbonyl of 12b to form the corresponding
3-hemiacetal by a similar mechanism described for pro-
tocatechuquinone 3-hemiacetal.20 The resultant alkoxide
ion in the A-ring attacks adjacent C-400 carbonyl of the B-
ring to form more stable cyclic hemiacetal 14. Although
the formation of 14 from 12 was confirmed by NMR in
methanol-d4/acetone-d6 = 3:2, compound 12 would also
be converted to 14 in H2O/MeOH, since 12 consumed
To elucidate the radical scavenging reaction after the
formation of 12, the reaction mixture of 12 and DPPH
1
radical was directly analyzed by NMR. The H NMR
spectrum of the mixture of 12 and DPPH radical in
methanol-d4/acetone-d6 (3:2) showed the disappearance
of signals due to 12, and the appearance of signals of
a closely related compound to 12. In situ 2D NMR anal-
yses of the reaction mixture suggested that two propeno-
ate parts and B-ring in 12 remained unchanged.
However, the coupling constant of the doublet signals
of H-50 and H-60 (J = 10.3 Hz) was larger than that of
H-50 and H-60 of 12 (J = 8.4 Hz), indicating that the cate-
chol structure of A-ring in 12 was oxidized to the qui-
none form. This was supported by the appearance of
an additional carbonyl carbon at dC 193.1. In addition,
high-field shift of C-30 (dC 100.3) of A-ring compared to
C-30 (dC 146.4) in 12, suggests the formation of acetal
via a nucleophilic addition of methanol-d4 at C-30 qui-
none carbonyl. Therefore, the structure of 14 was deter-
mined to be as shown in Figure 5.
A plausible radical scavenging mechanism of T12 in
water–methanol is outlined in Figure 6. In the first step,
the caffeoyl residues rapidly scavenge four radicals and
are converted to the corresponding o-quinone structures
(11). The resultant o-quinone (11) undergoes a nucleo-
philic attack by a methanol molecule at C-600 of the
B-ring to yield an enolate intermediate, which spont-
aneously attacks C-20 of A-ring quinone. It is well
documented that caffeoquinone is susceptible to
nucleophilic attack by thiol compounds such as glutathi-
Figure 6. Plausible radical scavenging mechanism of T12 in water–methanol.

S. Saito et al. / Bioorg. Med. Chem. 13 (2005) 4191–4199
4195
approximately two radicals in both 40% H2O/MeOH
and 40% acetone/MeOH (methanol/acetone = 3:2).
quired using a UV-1600 spectrophotometer (Shimadzu).
Preparative and analytical thin-layer chromatography
was performed on silica gel plates Merck 60 F254
(0.5 mm and 0.25 mm thickness), respectively. Column
chromatography was performed with silica gel, Wakogel
C-300 (Wako Pure Chemical Industries) and Diaion HP
20 (Mitsubishi Chemical Co.).
It is suspected that a water molecule may also attack
T12 bis-o-quinone (11) in water/MeOH or water/EtOH
solution, but the formation of a water adduct was not
confirmed in the present study. The reason that alcohol
adducts form preferentially might be explained by the
higher nucleophilicity of alcohol than water. However,
the presence of water may play a crucial role in the sec-
ond step, since the DPPH radical scavenging reaction
accelerates by the increase of water content in the sol-
vent. In addition, the fact that 12 was not detected in
the reaction mixture of T12 and DPPH radical in meth-
anol also supports the importance of water. The increas-
ing amount of water results in higher polar nature of the
solution, which might be advantageous for closer stack-
ing of hydrophobic caffeoyl or caffeoquinone groups.
trans-1,2-Cyclohexanediol, 1,3- and 1,4-cyclohexanediols
(cis and trans mixture) and 1,6-anhydro-b-D-glucose
were purchased from Tokyo Kasei Kogyo Co. cis-1,2-
Cyclohexanediol and 2,20-azinobis(3-ethylbenzthiazo-
line-6-sulfonic acid) diammonium salt (ABTS) were
obtained from Aldrich Chemical Co. and 6-hydroxy-
2,5,7,8-tetramethylchroman-2-carboxylic acid (trolox)
from Sigma Chemical Co. Caffeic acid, 2,2-diphenyl-1-
picrylhydrazyl (DPPH) radical, potassium peroxodisul-
fate (potassium persulfate), and other reagents were
purchased from Wako Pure Chemical Industries. All
solvents used were of reagent grade.
Although the two caffeoyl residues need to be close to
each other during the intramolecular coupling reactions,
T13, C14, T14, and CC slowly consumed radicals after
the first step. The result suggests that intermolecular
dimerizations might also occur, which could contribute
to the radical scavenging activity. However, in a dilute
solution, intramolecular couplings of the caffeoyl resi-
dues would effectively increase the rate of the second
step of the radical scavenging reactions.
4.2. Colorimetric radical scavenging tests
4.2.1. DPPH radical scavenging activity. To a solution of
a test compound in ethanol (2 mL) and 100 mM acetate
buffer (pH 5.5, 2 mL) was added 1 mL of DPPH radical
(500 lM, in ethanol) in a test tube. In the different
water–ethanol ratio experiments, the amount of water
and ethanol was adjusted so as to match the required
final ratio. Ethanol was used in the place of antioxidant
solution as a control. The solution was immediately
mixed vigorously for 10 s by a Vortex mixer and trans-
ferred to a cuvette. The absorbance reading was taken
0.5, 1, 2, 5, 10, and 30 min after initial mixing. The rad-
ical scavenging activity of T12 (3) and 12 was compared
with DL-a-tocopherol in 40% H2O/MeOH as described
previously. The radical scavenging equivalence was ex-
pressed as the values relative to that of DL-a-tocopherol
as 2.0.21 All experiments were performed in triplicate.
3. Conclusion
Radical scavenging reactions of dicaffeoyloxycyclohex-
anes in an aqueous alcohol consisted of two different
steps. In the first step, catechol moieties of the caffeoyl
residues were rapidly converted to o-quinone structures.
The second step was triggered by the addition of an
alcohol molecule to the quinone followed by an intra-
molecular coupling of the two caffeoquinone residues
to regenerate catechol structures, which then react with
additional radicals, especially in the aqueous solution.
The total radical scavenging potency was thus affected
by the intramolecular distance of the two caffeoyl resi-
dues in the molecule. The antiradical activity of natural
dicaffeoylquinic acids might also be dependent on posi-
tions of caffeoyl ester groups in the biological aqueous
system.
4.2.2. ABTS radical scavenging activity. ABTS radical
scavenging activity was measured using a modified Re
et al.22 method. ABTS radical cation (ABTS+Å) was pro-
duced by reacting a 7 mM aqueous solution of ABTS
with 2.45 mM potassium persulfate (final concentra-
tion). The reaction mixture was allowed to stand in
the dark at room temperature for 12–16 h prior to use.
ABTS+ solution was diluted with distilled water to an
absorbance of 0.70 ( 0.02) at 734 nm. To a diluted
ABTS+ solution (2.97 mL) was added 0.03 mL of a test
compound in ethanol. The solution was immediately
mixed vigorously for 10 s by a Vortex mixer and trans-
ferred to a cuvette. The absorbance reading was taken
0.5, 1, 2, 5, 10, and 30 min after initial mixing. Ethanol
was used in place of antioxidant solution as a control.
An ethanol solution of trolox was measured as a positive
control. All experiments were performed in triplicate.
4. Experimental
4.1. General
NMR spectra were recorded on a Bruker AMX-500
spectrometer (1H, 500 MHz; 13C, 125 MHz); chemical
shifts are expressed relative to the residual signals of
chloroform-d (dH 7.24, dC 77.0), methanol-d4 (dH 3.30,
dC 49.0), dimethyl sulfoxide (DMSO)-d6 (dH 2.49), ace-
tone-d6 (dH 2.04, dC 29.8) and acetonitrile-d3 (dH 1.93).
Field desorption mass spectra (FD-MS) and fast atom
bombardment mass spectra (FAB-MS) were obtained
with JEOL JMS-SX102A and JEOL JMS-AX-500
instruments, respectively. Optical absorbance was ac-
4.3. Chemistry
4.3.1. 3,4-Di-O-(methoxycarbonyl)caffeic acid (1).14 To a
solution of caffeic acid (9.0 g, 50 mmol) in 1 M NaOH
(110 mL) was added methyl chloroformate (23 mL,

4196
S. Saito et al. / Bioorg. Med. Chem. 13 (2005) 4191–4199
300 mmol) at 0 ꢁC. The reaction mixture was stirred for
1 h. The precipitate was collected by filtration and
washed with water. The crude product was recrystallized
from ethanol to afford 1 as a white powder (13.7 g,
H-3eq and 6eq), 1.78–1.79 (2H, m, H-4ax and 5ax),
2.08–2.11 (2H, m, H-3ax and 6ax), 4.95 (2H, m, H-1
and 2), 6.18 (2H, d, J = 15.9 Hz, H-80), 6.73 (2H, d,
J = 8.2 Hz, H-50), 6.88 (2H, dd, J = 8.2, 2.0 Hz, H-60),
6.98 (2H, d, J = 2.0 Hz, H-20), 7.49 (2H, d,
J = 15.9 Hz, H-70).
1
93%); FD-MS m/z: 296 [M]+; H NMR (methanol-d4):
d 3.80 (6H, s, OCH3), 6.42 (1H, d, J = 16.0 Hz, H-8),
7.28 (1H, d, J = 8.4 Hz, H-5), 7.47 (1H, dd, J = 8.4,
2.0 Hz, H-6), 7.52 (1H, d, J = 2.0 Hz, H-2), 7.56 (1H,
d, J = 16.0 Hz, H-7).
4.3.6. cis-1,2-Dicaffeoyloxycyclohexane (C12, 4).14 cis-
1,2-Cyclohexanediol was converted to cis-1,2-dicaffeoyl-
oxycyclohexane (4) by the same method as 3. 4: pale
1
4.3.2. 3,4-Di-O-acetylcaffeic acid (2). To a solution of
caffeic acid (9.0 g, 50 mmol) in 1 M NaOH (110 mL)
was added acetic anhydride (19 mL, 200 mmol) at
0 ꢁC. The reaction mixture was stirred for 1 h. The pre-
cipitate was collected by filtration and washed with
water. The crude product was recrystallized from etha-
nol to afford 2 as a white powder (12.3 g, 93%); FD-
yellow powder; FD-MS m/z: 440 [M]+; H NMR (ace-
tone-d6): d 1.48–1.50 (2H, m), 1.69–1.76 (4H, m), 1.91–
1.95 (2H, m), 5.16 (2H, m, H-1 and 2), 6.28 (2H, d,
J = 15.9 Hz, H-80), 6.84 (2H, d, J = 8.2 Hz, H-50), 7.01
(2H, dd, J = 8.2, 2.0 Hz, H-60), 7.15 (2H, d,
J = 2.0 Hz, H-20), 7.55 (2H, d, J = 15.9 Hz, H-70).
1
MS m/z: 264 [M]+; H NMR (acetone-d6): d 2.28 (6H,
4.3.7. trans- and cis-1,3-Bis[3,4-di-O-(methoxycarbonyl)-
caffeoyloxy]cyclohexanes (5a and 6a).14 A mixture of cis-
and trans-1,3-cyclohexanediols (ca. 1:1) was reacted
with 1a by the same method as described for 3a. A ste-
reoisomeric mixture of 1,3-bis[3,4-di-O-(methoxycar-
bonyl)caffeoyloxy]cyclohexanes (5a and 6a) was
subjected to preparative TLC (chloroform/methanol/
acetic acid = 100:1:0.1) to afford trans-form (5a)
(Rf = 0.37, 20%) and cis-form (6a) (Rf = 0.29, 23%).
5a: yellow oil; FD-MS m/z: 672 [M]+; 1H NMR (chloro-
form-d): d 1.25–1.99 (8H, m, H-2, 4, 5 and 6), 3.92 (12H,
s, OCH3), 5.28 (2H, br s, H-1 and 3), 6.41 (2H, d, J =
16.0 Hz, H-80), 7.32 (2H, d, J = 8.4 Hz, H-50), 7.44
(2H, dd, J = 8.4, 2.0 Hz, H-60), 7.47 (2H, d, J =
2.0 Hz, H-20), 7.62 (2H, d, J = 16.0 Hz, H-70) 6a: yellow
s, COCH3), 6.52 (1H, d, J = 16.0 Hz, H-8), 7.30 (1H,
d, J = 8.1 Hz, H-5), 7.59 (1H, d, J = 2.0 Hz, H-2), 7.60
(1H, dd, J = 8.1, 2.0 Hz, H-6), 7.65 (1H, d,
J = 16.0 Hz, H-7).
4.3.3. 3,4-Di-O-(methoxycarbonyl)caffeoyl chloride (1a)
and 3,4-di-O-acetylcaffeoyl chloride (2a).14 To a solution
of 1 (2.96 g, 10 mmol) in toluene (20 mL) was added
oxalyl chloride (5.2 mL, 60 mmol, 6 equiv). After stir-
ring for 1.5 h at 60 ꢁC, toluene and unreacted oxalyl
chloride were removed under the reduced pressure.
Compound 2a was prepared using the same method de-
scribed for 1a. The crude products 1a and 2a were used
immediately without purification.
1
oil; FD-MS m/z: 672 [M]+; H NMR (chloroform-d): d
4.3.4. trans-1,2-Bis[3,4-di-O-(methoxycarbonyl)caffeoyl-
oxy]cyclohexane (3a).14
1.25–2.19 (8H, m, H-2, 4, 5 and 6), 3.91 (12H, s,
OCH3), 4.95 (2H, br s, H-1 and 3), 6.37 (2H, d,
J = 16.0 Hz, H-80), 7.30 (2H, d, J = 8.6 Hz, H-50), 7.39
(2H, dd, J = 8.6, 2.0 Hz, H-60), 7.45 (2H, d, J =
2.0 Hz, H-20), 7.60 (2H, d, J = 16.0 Hz, H-70).
trans-1,2-Cyclohexanediol
(1.7 g, 15 mmol) was reacted with 1a (12.6 g, 40 mmol,
2.7 equiv) without solvent in an oil bath at 150 ꢁC for
20 min. The residue was subjected to silica gel column
chromatography (hexane/ethyl acetate = 2:1) to afford
3a (4.8 g, 47%) as a yellow oil; FD-MS m/z: 672 [M]+;
1H NMR (chloroform-d): d 1.35–1.39 (2H, m, H-4eq
and 5eq), 1.41–1.47 (2H, m, H-3eq and 6eq), 1.74–1.75
(2H, m, H-4ax and 5ax), 2.09–2.11 (2H, m, H-3ax and
6ax), 3.85 (6H, s, OCH3), 3.86 (6H, s, OCH3), 4.97
(2H, m, H-1 and 2), 6.29 (2H, d, J = 16.0 Hz, H-80),
7.24 (2H, d, J = 8.5 Hz, H-50), 7.34 (2H, dd, J = 8.5,
2.0 Hz, H-60), 7.38 (2H, d, J = 2.0 Hz, H-20), 7.53 (2H,
d, J = 16.0 Hz, H-70).
4.3.8. trans- and cis-1,3-Dicaffeoyloxycyclohexanes (T13,
5 and C13, 6).14 Compounds 5 and 6 were prepared by
selective hydrolysis of 5a and 6a, respectively, by the
same method as described for 3 except for using 5% NaH-
CO3 in place of 2% Na2CO3. The crude product was sub-
jected to preparative TLC (hexane/ethyl acetate = 1:2) to
afford 5 (Rf = 0.21, 76%) and 6 (Rf = 0.38, 51%). 5: yellow
1
oil; FD-MS m/z: 440 [M]+; H NMR (methanol-d4): d
1.65–2.00 (8H, m, H-2, 4, 5 and 6), 5.21 (2H, br s, H-1
and 3), 6.26 (2H, d, J = 16.0 Hz, H-80), 6.77 (2H, d,
J = 8.4 Hz, H-50), 6.95 (2H, dd, J = 8.4, 2.0 Hz, H-60),
7.04 (2H, d, J = 2.0 Hz, H-20), 7.54 (2H, d, J = 16.0 Hz,
4.3.5. trans-1,2-Dicaffeoyloxycyclohexane (T12, 3).14 To
a solution of 3a (4.8 g, 7.1 mmol) in tetrahydrofuran
(50 mL) and methanol (200 mL) was added 2% Na2CO3
(100 mL) and stirred at room temperature for 12 h.
After acidification with 10 M HCl to pH 1–2, the mix-
ture was diluted with water and extracted with diethyl
ether. The organic layer was dried over anhydrous mag-
nesium sulfate and evaporated under the reduced pres-
sure. The residue was subjected to silica gel column
chromatography (hexane/ethyl acetate = 1:2). The crude
solid from the eluate was recrystallized from ethyl ace-
tate/hexane to afford 3 (2.8 g, 90%) as a white powder;
FD-MS m/z: 440 [M]+; 1H NMR (methanol-d4): d
1.41–1.45 (2H, m, H-4eq and 5eq), 1.49–1.54 (2H, m,
1
H-70) 6: yellow oil; FD-MS m/z: 440 [M]+; H NMR
(methanol-d4): d 1.30–2.30 (8H, m, H-2, 4, 5 and 6),
4.90 (2H, br s, H-1 and 3), 6.22 (2H, d, J = 15.9 Hz, H-
80), 6.73 (2H, d, J = 8.2 Hz, H-50), 6.89 (2H, dd,
J = 8.2, 2.0 Hz, H-60), 7.02 (2H, d, J = 2.0 Hz, H-20),
7.53 (2H, d, J = 15.9 Hz, H-70).
4.3.9. trans- and cis-1,4-Bis[3,4-di-O-(methoxycarbonyl)-
caffeoyloxy]cyclohexanes (7a and 8a).14 A mixture of
trans- and cis- 1,4-cyclohexanediols (ca. 1:1) was reacted
with 1a by the same method as described for 3a. A ste-
reoisomeric mixture of 1,4-bis[3,4-di-O-(methoxycar-

S. Saito et al. / Bioorg. Med. Chem. 13 (2005) 4191–4199
4197
bonyl)caffeoyloxy]cyclohexanes (7a and 8a) was sub-
jected to preparative TLC (chloroform/methanol/acetic
acid = 100:1:0.1) to afford trans-form (7a) (Rf = 0.69,
17%) and cis-form (8a) (Rf = 0.57, 16%). 7a: white pow-
4.3.13. 2,4-Di-O-caffeoyl-1,6-anhydro-b-D-glucose (C13aa,
10). Compound 10 was prepared by hydrolysis of 10a
with 5% NaHCO3, using the same method as described
for 3. 10: white powder; FAB-HR-MS (negative) m/z:
1
1
der; FD-MS m/z: 672 [M]+; H NMR (chloroform-d): d
485.1111 [MꢀH]ꢀ; calcd for C24H21O11, 485.1083; H
1.65 (4H, m), 2.07 (4H, m), 3.90 (12H, s, OCH3), 4.95
(2H, br s, H-1 and 4), 6.38 (2H, d, J = 16.0 Hz, H-80),
7.31 (2H, d, J = 8.4 Hz, H-50), 7.41 (2H, dd, J = 8.4,
2.0 Hz, H-60), 7.45 (2H, d, J = 2.0 Hz, H-20), 7.59 (2H,
d, J = 16.0 Hz, H-70) 8a: colorless crystal; FD-MS m/z:
672 [M]+; 1H NMR (chloroform-d): d 1.80 (4H, m),
1.91 (4H, m), 3.90 (12H, s, OCH3), 5.00 (2H, br s, H-1
and 4), 6.40 (2H, d, J = 16.0 Hz, H-80), 7.31 (2H, d,
J = 8.5 Hz, H-50), 7.42 (2H, dd, J = 8.5, 2.0 Hz, H-60),
7.46 (2H, d, J = 2.0 Hz, H-20), 7.62 (2H, d,
J = 16.0 Hz, H-70).
NMR (acetone-d6): d 3.73 (1H, dd, J = 7.1, 5.9 Hz, H-
6), 3.88 (1H, m, H-3), 4.25 (1H, d, J = 7.1 Hz, H-6),
4.66 (2H, m, H-2 and 5), 4.76 (1H, m, H-4), 4.89 (1H,
d, J = 4.7 Hz, 3-OH), 5.43 (1H, br s, H-1), 6.32 and
6.34 (each 1H, d, J = 16.0 Hz, H-80 and 800), 6.82 and
6.83 (each 1H, d, J = 8.4 Hz, H-50 and 500), 7.01 and
7.02 (each 1H, dd, J = 8.4, 2.0 Hz, H-60 and 600), 7.18
and 7.19 (each 1H, d, J = 2.0 Hz, H-20 and 200), 7.60
and 7.62 (each 1H, d, J = 16.0 Hz, H-70 and 700); 13C
NMR (acetone-d6): d 66.5 (C-6), 70.0 (C-3), 73.9 (C-5),
74.7 (C-4), 75.4 (C-2), 101.0 (C-1), 115.0 (C-80 and 800),
115.2 (C-20 and 200), 116.4 (C-50 and 500), 122.9 (C-60
and 600), 127.4 (C-10 and 100), 146.3 (C-30 and 300), 146.6
(C-70 and 700), 149.0 (C-40 and 400), 166.7, 166.9 (C-90
and 900); HMBC correlation peaks: H-1/C-2, C-3, C-5,
C-6, H-2 and H-5/C-1, C-3, C-4, C-90, C-900, 3-OH/C-
2, C-4, H-4/C-3, C-90, C-900, H-6/C-1, C-4, H-20 (200)/C-
40 (400), C-60 (600), C-70 (700), H-50 (500)/C-10 (100), C-30
(300), C-40 (400), H-60 (600)/C-20 (200), C-40 (400), C-50 (500),
C-70 (700), H-70 (700)/C-20 (200), C-60 (600), C-80 (800), C-90
(900), H-80 (800)/C-10 (100), C-90 (900).
4.3.10. trans- and cis-1,4-Dicaffeoyloxycyclohexanes
(T14, 7 and C14, 8).14 Compounds 7 and 8 were pre-
pared by selective hydrolysis of 7a and 8a, respectively,
by the same method as described for 3. The crude prod-
uct was subjected to preparative TLC (hexane/ethyl ace-
tate = 1:2) to afford 7 (Rf = 0.31, 38%) and 8 (Rf = 0.25,
1
39%). 7: yellow oil; FD-MS m/z: 440 [M]+; H NMR
(DMSO-d6): d 1.57 (4H, m), 1.97 (4H, m), 4.82 (2H,
br s, H-1 and 4), 6.24 (2H, d, J = 15.8 Hz, H-80), 6.74
(2H, d, J = 8.1 Hz, H-50), 7.00 (2H, dd, J = 8.1,
2.0 Hz, H-60), 7.03 (2H, d, J = 2.0 Hz, H-20), 7.46 (2H,
d, J = 15.8 Hz, H-70) 8: yellow oil; FD-MS m/z: 440
[M]+; 1H NMR (acetone-d6) ppm (J in Hz): d 1.82
(4H, m), 1.90 (4H, m), 4.95 (2H, br s, H-1 and 4), 6.31
(2H, d, J = 15.9 Hz, H-80), 6.86 (2H, d, J = 8.2 Hz, H-
50), 7.06 (2H, dd, J = 8.2, 2.0 Hz, H-60), 7.17 (2H, d,
J = 2.0 Hz, H-20), 7.56 (2H, d, J = 15.9 Hz, H-70).
4.4. Isolation of oxidation products of trans-1,2-dicaffe-
oyloxycyclohexane (3)
4.4.1. Isolation of compound 12. To a solution of T12 (3,
55 mg, 125 lmol) in 40% H2O/MeOH (3 L) was added
DPPH radical (394 mg, 1.0 mmol, 8 equiv) in methanol
(40 mL) and stirred for 20 min at room temperature. So-
dium dithionite (400 mg, 2.3 mmol) was added to the
mixture to reduce unstable o-quinones to their catechol
forms. The reaction mixture was concentrated until
methanol was completely removed. The residue was sub-
jected to Diaion HP 20 column chromatography and
washed with water, and then eluted with methanol.
The methanol fraction was evaporated under the re-
duced pressure. The residue was subjected to silica gel
column chromatography with chloroform to remove
DPPH radical, and then with ethyl acetate. The ethyl
acetate fraction was concentrated, and the crude prod-
uct was further purified by preparative TLC (chloro-
form/ethyl acetate/formic acid = 50:50:1) to afford
compound 12 (Rf = 0.63, 4.3 mg, 7.4%) as a yellow crys-
tal; mp > 300 ꢁC; FD-HR-MS m/z: 468.1404 [M]+; calcd
4.3.11. Cyclohexyl caffeate (CC, 9). Cyclohexanol was
converted to cyclohexyl caffeate (9) by the same method
as 3. 9: yellow solid; mp 151–155 ꢁC; FD-HR-MS m/z:
262.1211 [M]+; calcd for C15H18O4, 262.1205; 1H
NMR (acetone-d6): d 1.27–1.32 (1H, m, H-4ax), 1.35–
1.49 (4H, m, H-2ax, 3eq, 5eq and 6ax), 1.52–1.56 (1H,
m, H-4eq), 1.73–1.75 (2H, m, H-3ax and 5ax), 1.85 (2H,
m, H-2eq and 6eq), 4.78 (1H, m, H-1ax), 6.25 (1H, d,
J = 15.9 Hz, H-80), 6.85 (1H, d, J = 8.2 Hz, H-50), 7.02
(1H, dd, J = 8.2, 2.0 Hz, H-60), 7.15 (1H, d,
J = 2.0 Hz, H-20), 7.51 (1H, d, J = 15.9 Hz, H-70).
4.3.12. 2,4-Bis-O-(3,4-di-O-acetylcaffeoyl)-1,6-anhydro-b-
D-glucose (10a). To a solution of 1,6-anhydro-b-D-glu-
cose (2.6 g, 16 mmol) in pyridine (40 mL) was added
dropwise a solution of 2a (11.3 g, 40 mmol, 2.5 equiv)
in toluene (100 mL). After stirring at room temperature
for 8 h, the reaction mixture was evaporated under the
reduced pressure. The crude product was subjected to
silica gel column chromatography (chloroform/metha-
nol = 100:3) to afford 10a (3.0 g, 57%) as a yellow oil;
1
for C25H24O9, 468.1420; H NMR (acetone-d6): d 1.45
(2H, m, H-4ax and 5ax), 1.57 (2H, m, H-3ax and 6ax),
1.82 (2H, m, H-4eq and 5eq), 2.00 (2H, m, H-3eq and
6eq), 3.48 (3H, s, H-1000), 4.63 (1H, ddd, J = 11.6, 9.5,
4.7 Hz, H-2), 4.70 (1H, d, J = 4.1 Hz, H-500), 5.06 (1H,
dd, J = 4.1, 1.5 Hz, H-600), 5.08 (1H, ddd, J = 11.6, 9.5,
4.7 Hz, H-1), 5.91 (1H, d, J = 15.4 Hz, H-80), 6.35
(1H, d, J = 16.1 Hz, H-800), 6.51 (1H, dd, J = 1.5,
0.7 Hz, H-200), 6.74 (1H, d, J = 8.4 Hz, H-50), 6.87 (1H,
s, 400-OH), 6.94 (1H, d, J = 8.4 Hz, H-60), 7.20 (1H, dd,
J = 16.1, 0.7 Hz, H-700), 7.87 (1H, d, J = 15.4 Hz, H-
70), 8.54 (1H, s, 40-OH); 13C NMR (acetone-d6): d 24.5
(C-4), 24.6 (C-5), 30.7 (C-6), 31.2 (C-3), 51.5 (C-500),
57.1 (C-1000), 74.0 (C-600), 75.3 (C-1), 78.9 (C-2), 106.2
1
FD-MS m/z: 655 [M+H]+; H NMR (chloroform-d): d
2.27 (12H, s, COCH3), 3.81 (1H, dd, J = 7.1, 6.0 Hz,
H-6), 3.92 (1H, m, H-3), 4.23 (1H, d, J = 7.1 Hz, H-6),
4.69 (2H, m, H-2 and 5), 4.80 (1H, m, H-4), 5.56 (1H,
m, H-1), 6.44 (2H, d, J = 16.0 Hz, H-80 and 800), 7.17
(2H, m, H-50 and 500), 7.32–7.35 (4H, m, H-20, 200, 60
and 600), 7.67 (2H, d, J = 16.0 Hz, H-70 and 700).

4198
S. Saito et al. / Bioorg. Med. Chem. 13 (2005) 4191–4199
(C-400), 118.2 (C-50), 121.4 (C-60), 122.5 (C-80), 125.2 (C-
10), 126.5 (C-20), 127.3 (C-800), 132.1 (C-200), 140.7 (C-700),
143.7 (C-40), 144.2 (C-70), 146.4 (C-30), 154.0 (C-100),
165.7 (C-900), 166.8 (C-90), 190.7 (C-300); HMBC correla-
tion peaks: H-1/C-2, C-90, H-2/C-1, C-900, H-50/C-10, C-
30, C-400, C-60, H-60/C-20, C-30, C-40, C-50, H-70/C-10, C-
60, C-9 , H-80/C-10, C-90, H-200/C-100, C-400, C-600, C-700,
C-800, 400-OH/C-300, C-400, C-500, H-500/C-20, C-30, C-100,
C-300, C-400, C-600, H-600/C-100, C-200, C-400, C-500, C-700, C-
1000, H-700/C-100, C-200, C-600, C-800, C-900, H-800/C-100, C-
200, C-700, H-1000/C-600.
Compound 14. 1H NMR (methanol-d4/acetone-d6 = 3:2):
d 1.46, 1.59, 1.85, 2.05 (each 2H, m, H-3, 4, 5 and 6), 3.42
(3H, s, H-1000), 4.28 (1H, d, J = 3.2 Hz, H-500), 4.68 (1H,
m, H-2), 5.04 (1H, d, J = 3.2 Hz, H-600), 5.15 (1H, m,
H-1), 5.94 (1H, d, J = 10.3 Hz, H-50), 6.22 (1H, d,
J = 15.8 Hz, H-80), 6.35 (1H, d, J = 16.0 Hz, H-800), 6.49
(1H, s, H-200), 7.08 (1H, d, J = 10.3 Hz, H-60), 7.27 (1H,
d, J = 16.0 Hz, H-700), 7.71 (1H, d, J = 15.8 Hz, H-70);
13C NMR (methanol-d4/acetone-d6 = 3:2): d 24.8, 24.9,
30.9, 31.4 (C-3, 4, 5 and 6), 51.9 (C-500), 57.3 (C-1000),
73.8 (C-600), 76.3 (C-1), 79.5 (C-2), 100.3 (C-30), 103.6
(C-400), 125.5 (C-10), 125.7 (C-80), 125.9 (C-50), 010 26.6 (C-
800), 132.4 (C-200), 141.6 (C-60), 141.7 (C-7 ), 141.8
(C-70), 148.9 (C-20), 153.2 (C-100), 166.6 (C-900), 167.9
(C-90), 191.7 (C-300), 193.1 (C-40); HMBC correlation
peaks: H-50/C-30, H-60/C-20, C-40, H-70/C-90, H-200/C-400,
C-600, H-500/C-100, H-700/C-600, C-900, H-800/C-100, H-1000/C-600.
4.4.2. Isolation of compound 13. Compound 13 was iso-
lated from the reaction mixture of T12 (3) and DPPH
radical in 40% H2O/EtOH by the same method as de-
scribed for 12. 13: 2.4 mg, 4.0%; pale yellow crystal;
Rf = 0.68 (chloroform/ethyl acetate/formic acid =
50:50:1); mp > 300 ꢁC; FD-HR-MS m/z: 482.1590
[M]+; calcd for C26H26O9, 482.1577; 1H NMR
(acetone-d6): d 1.11 (3H, t, J = 7.0 Hz, 1000-CH3), 1.46
(2H, m, H-4ax and 5ax), 1.57 (2H, m, H-3ax and 6ax),
1.82 (2H, m, H-4eq and 5eq), 2.00 (2H, m, H-3eq and
6eq), 3.62 (1H, dq, J = 9.0, 7.0 (q) Hz, H-1000), 3.87
(1H, dq, J = 9.0, 7.0 (q) Hz, H-1000), 4.63 (1H, ddd,
J = 11.6, 9.5, 4.7 Hz, H-2), 4.69 (1H, d, J = 4.1 Hz, H-
500), 5.08 (1H, ddd, J = 11.6, 9.5, 4.7 Hz, H-1), 5.13
(1H, dd, J = 4.1, 1.2 Hz, H-600), 5.90 (1H, d,
J = 15.4 Hz, H-80), 6.32 (1H, d, J = 16.1 Hz, H-800),
6.50 (1H, d, J = 1.2 Hz, H-200), 6.74 (1H, d, J = 8.4 Hz,
H-50), 6.86 (1H, s, 400-OH), 6.93 (1H, d, J = 8.4 Hz, H-
60), 7.18 (1H, d, J = 16.1 Hz, H-700), 7.87 (1H, d,
J = 15.4 Hz, H-70), 8.53 (1H, s, 40-OH); 13C NMR (ace-
tone-d6): d 15.5 (C-1100), 24.5 (C-4), 24.6 (C-5), 30.7 (C-
6), 31.2 (C-3), 51.5 (C-500), 65.2 (C-1000), 72.5 (C-600), 75.3
(C-1), 78.9 (C-2), 106.3 (C-400), 118.1 (C-50), 121.3 (C-60),
122.4 (C-80), 125.1 (C-10), 126.6 (C-20), 127.1 (C-800),
132.1 (C-200), 140.7 (C-700), 143.7 (C-40), 144.2 (C-70),
146.4 (C-30), 154.1 (C-100), 165.6 (C-900), 166.8 (C-90),
190.8 (C-300).
Acknowledgments
We are grateful to Mr. Kenji Watanabe and Dr. Eri
Fukushi, of the GC–MS and NMR Laboratory of our
faculty, for measuring mass spectra.
References and notes
1. Maruta, Y.; Kawabata, J.; Niki, R. J. Agric. Food Chem.
1995, 43, 2592.
2. Chuda, Y.; Ono, H.; Ohnishi-Kameyama, M.; Nagata, T.;
Tsushida, T. J. Agric. Food Chem. 1996, 44, 2037.
3. Clifford, M. N. J. Sci. Food Agric. 1999, 79, 362.
4. Iwai, K.; Kishimoto, N.; Kakino, Y.; Mochida, K.; Fujita,
T. J. Agric. Food Chem. 2004, 52, 4893.
5. Nakatani, N.; Kayano, S.; Kikuzaki, H.; Sumino, K.;
Katagiri, K.; Mitani, T. J. Agric. Food Chem. 2000, 48,
5512.
6. Yoshimoto, M.; Yahara, S.; Okuno, S.; Islam, M. S.;
Ishiguro, K.; Yamakawa, O. Biosci. Biotechnol. Biochem.
2002, 66, 2336.
7. Fulcrand, H.; Cheminat, A.; Brouillard, R.; Cheynier, V.
Phytochemistry 1994, 35, 499.
4.5. NMR analyses
4.5.1. Reaction of trans-1,2-dicaffeoyloxycyclohexane (3)
and DPPH radical. To DPPH radical (4.9 mg,
12.5 lmol, 2.5 equiv) was added a solution of trans-
1,2-dicaffeoyloxycyclohexane (3, 2.2 mg, 5 lmol) in ace-
tonitrile-d3 (0.4 mL). The mixture was immediately
transferred to an NMR tube and analyzed by 1H NMR.
8. Cilliers, J. J. L.; Singleton, V. L. J. Agric. Food Chem.
1991, 39, 1298.
9. Tazaki, H.; Taguchi, D.; Hayashida, T.; Nabeta, K.
Biosci. Biotechnol. Biochem. 2001, 65, 2613.
10. Tazaki, H.; Kawabata, J.; Fujita, T. Biosci. Biotechnol.
Biochem. 2003, 67, 1185.
11. Arakawa, R.; Yamaguchi, M.; Hotta, H.; Osakai, T.;
Kimoto, T. J. Am. Soc. Mass Spectrom. 2004, 15, 1228.
12. Takahashi, H.; Matsumoto, K.; Ueda, M.; Miyake, Y.;
Fukuyama, Y. Heterocycles 2002, 56, 245.
4.5.2. trans-1,2-Dicaffeoyloxycyclohexane bis-o-quinone
1
(11). H NMR (acetonitrile-d3): d 1.41–1.45 (2H, m),
1.52–1.54 (2H, m), 1.75–1.77 (2H, m), 2.08–2.10 (2H,
m), 4.97 (2H, m, H-1 and 2), 6.42 (2H, d, J = 10.3 Hz,
H-50), 6.49 (2H, s, H-20), 6.56 (2H, d, J = 16.0 Hz, H-
80), 7.39 (2H, d, J = 10.3 Hz, H-60), 7.40 (2H, d,
J = 16.0 Hz, H-70).
13. Horman, I.; Badoud, R.; Ammann, W. J. Agric. Food
Chem. 1984, 32, 538.
14. King, P. J.; Ma, G.; Miao, W.; Jia, Q.; McDougall, B. R.;
Reinecke, M. G.; Cornell, C.; Kuan, J.; Kim, T. R.;
Robinson, W. E., Jr. J. Med. Chem. 1999, 42, 497.
15. Panzella, L.; Napolitano, A.; dÕIschia, M. Bioorg. Med.
Chem. 2003, 11, 4797.
16. Cheynier, V. F.; Van Hulst, M. W. J. J. Agric. Food Chem.
1988, 36, 10.
17. Moridani, M. Y.; Scobie, H.; Jamshidzadeh, A.; Salehi, P.;
OÕBrien, P. J. Drug Metab. Dispos. 2001, 29, 1432.
18. Saito, S.; Okamoto, Y.; Kawabata, J. Biosci. Biotechnol.
Biochem. 2004, 68, 1221.
4.5.3. Reaction of compound 12 and DPPH radical. To a
solution of DPPH radical (7.0 mg, 17.8 lmol, 3.0 equiv)
in acetone-d6 (0.1 mL) was added compound 12 (2.8 mg,
5.9 lmol) in methanol-d4/acetone-d6 (3:1, 0.4 mL). The
mixture was immediately transferred to an NMR tube
and analyzed by NMR.

S. Saito et al. / Bioorg. Med. Chem. 13 (2005) 4191–4199
4199
19. Saito, S.; Kawabata, J. J. Agric. Food Chem. 2004, 52,
8163.
21. Kawabata, J.; Okamoto, Y.; Kodama, A.; Makimoto, T.;
Kasai, T. J. Agric. Food Chem. 2002, 50, 5468.
20. Saito, S.; Okamoto, Y.; Kawabata, J.; Kasai, T. Biosci.
Biotechnol. Biochem. 2003, 67, 1578.
22. Re, R.; Pellegrini, N.; Proteggente, A.; Pannala, A.; Yang,
M.; Rice-Evans, C. Free Radic. Biol. Med. 1999, 26, 1231.