The mitochondrial monoamine oxidase - Aldehyde dehydrogenase pathway: A potential site of action of daidzin

Article abstract of DOI:10.1021/jm990614i

Recent studies showed that daidzin suppresses ethanol intake in ethanol-preferring laboratory animals. In vitro, it potently and selectively inhibits the mitochondrial aldehyde dehydrogenase (ALDH-2). Further, it inhibits the conversion of monoamines such as serotonin (5-HT) and dopamine (DA) into their respective acid metabolites, 5-hydroxyindole-3-acetic acid (5-HIAA) and 3,4-dihydroxyphenylacetic acid (DOPAC) in isolated hamster or rat liver mitochondria. Studies on the suppression of ethanol intake and inhibition of 5-HIAA (or DOPAC) formation by six structural analogues of daidzin suggested a potential link between these two activities. This, together with the finding that daidzin does not affect the rates of mitochondria-catalyzed oxidative deamination of these monoamines, raised the possibility that the ethanol intake-suppressive (antidipsotropic) action of daidzin is not mediated by the monoamines but rather by their reactive biogenic aldehyde intermediates such as 5-hydroxyindole-3-acetaldehyde (5-HIAL) and/or 3,4-dihydroxyphenylacetaldehyde (DOPAL) which accumulate in the presence of daidzin. To further evaluate this possibility, we synthesized more structural analogues of daidzin and tested and compared their antidipsotropic activities in Syrian golden hamsters with their effects on monoamine metabolism in isolated hamster liver mitochondria using 5-HT as the substrate. Effects of daidzin and its structural analogues on the activities of monoamine oxidase (MAO) and ALDH-2, the key enzymes involved in 5-HT metabolism in the mitochondria, were also examined. Results from these studies reveal a positive correlation between the antidipsotropic activities of these analogues and their abilities to increase 5-HIAL accumulation during 5-HT metabolism in isolated hamster liver mitochondria. Daidzin analogues that potently inhibit ALDH-2 but have no or little effect on MAO are most antidipsotropic, whereas those that also potently inhibit MAO exhibit little, if any, antidipsotropic activity. These results, although inconclusive, are consistent with the hypothesis that daidzin may act via the mitochondrial MAO/ALDH pathway and that a biogenic aldehyde such as 5-HIAL may be important in mediating its antidipsotropic action.

Full text of DOI:10.1021/jm990614i

J . Med. Chem. 2000, 43, 4169-4179
4169
Th e Mitoch on d r ia l Mon oa m in e Oxid a se-Ald eh yd e Deh yd r ogen a se P a th w a y: A
P oten tia l Site of Action of Da id zin
Nade`ge Rooke, Dian-J un Li, J unqing Li, and Wing Ming Keung*
Center for Biochemical and Biophysical Sciences and Medicine, Harvard Medical School, 250 Longwood Avenue,
Boston, Massachusetts 02115
Received March 13, 2000
Recent studies showed that daidzin suppresses ethanol intake in ethanol-preferring laboratory
animals. In vitro, it potently and selectively inhibits the mitochondrial aldehyde dehydrogenase
(ALDH-2). Further, it inhibits the conversion of monoamines such as serotonin (5-HT) and
dopamine (DA) into their respective acid metabolites, 5-hydroxyindole-3-acetic acid (5-HIAA)
and 3,4-dihydroxyphenylacetic acid (DOPAC) in isolated hamster or rat liver mitochondria.
Studies on the suppression of ethanol intake and inhibition of 5-HIAA (or DOPAC) formation
by six structural analogues of daidzin suggested a potential link between these two activities.
This, together with the finding that daidzin does not affect the rates of mitochondria-catalyzed
oxidative deamination of these monoamines, raised the possibility that the ethanol intake-
suppressive (antidipsotropic) action of daidzin is not mediated by the monoamines but rather
by their reactive biogenic aldehyde intermediates such as 5-hydroxyindole-3-acetaldehyde (5-
HIAL) and/or 3,4-dihydroxyphenylacetaldehyde (DOPAL) which accumulate in the presence
of daidzin. To further evaluate this possibility, we synthesized more structural analogues of
daidzin and tested and compared their antidipsotropic activities in Syrian golden hamsters
with their effects on monoamine metabolism in isolated hamster liver mitochondria using 5-HT
as the substrate. Effects of daidzin and its structural analogues on the activities of monoamine
oxidase (MAO) and ALDH-2, the key enzymes involved in 5-HT metabolism in the mitochondria,
were also examined. Results from these studies reveal a positive correlation between the
antidipsotropic activities of these analogues and their abilities to increase 5-HIAL accumulation
during 5-HT metabolism in isolated hamster liver mitochondria. Daidzin analogues that potently
inhibit ALDH-2 but have no or little effect on MAO are most antidipsotropic, whereas those
that also potently inhibit MAO exhibit little, if any, antidipsotropic activity. These results,
although inconclusive, are consistent with the hypothesis that daidzin may act via the
mitochondrial MAO/ALDH pathway and that a biogenic aldehyde such as 5-HIAL may be
important in mediating its antidipsotropic action.
In tr od u ction
earlier study, we showed that daidzin potently and
selectively inhibits human, rat, and hamster liver mito-
chondrial aldehyde dehydrogenase isozyme (ALDH-2),
the major ALDH isozyme that catalyzes the oxidation
of ethanol-derived acetaldehyde, and suggested that it
may act as an alcohol sensitizing agent.1,5 However, in
an in vivo study, we demonstrated that daidzin at a dose
that suppresses ethanol intake does not affect the
overall acetaldehyde metabolism in golden hamsters.6,7
These apparently conflicting results were later at-
tributed to the presence of a cytosolic ALDH isozyme
in hamster liver. This ALDH isozyme catalyzes acetal-
dehyde oxidation efficiently and with high capacity, and
it is not inhibited by daidzin.8 On the basis of these
results, we concluded that the mechanism by which
daidzin suppresses hamster ethanol intake is different
from that proposed for the classic ALDH inhibitors
such as disulfiram and calcium carbimide,9 and we
suggested that it may act by modulating the activity of
an as-yet-undefined physiological pathway catalyzed by
ALDH-2.
For more than a millennium, herbalists practicing
traditional Chinese medicine have used Radix puerariae
(RP, root of kudzu)- or Flos puerariae (FP, flower of
kudzu)-based herbal medicines for the treatment of
“alcohol addiction.” Recently, we showed that a crude
extract of RP suppresses ethanol intake in ethanol-
preferring Syrian golden hamsters (Mesocricetus aura-
tus). Further, we isolated and identified daidzin (7-O-
â-glucosyl-4′-hydroxyisoflavone) as its major active
principle and demonstrated that pure daidzin synthe-
sized in our laboratory also suppresses hamster ethanol
intake.1 Since then, we and others have confirmed the
antidipsotropic (ethanol intake-suppressive) activity of
RP extracts and daidzin in additional laboratory rodents
such as Wistar rats, Fawn Hooded rats, and the geneti-
cally bred alcohol-preferring P rats under various
experimental conditions including two-lever choice, two-
bottle free-choice, limited access, and ethanol-deprived
paradigms.2-4
The mechanism by which daidzin suppresses ethanol
intake in laboratory rodents is still unknown. In an
Recently, we showed that daidzin inhibits the conver-
sion of monoamines such as serotonin (5-HT) and
dopamine (DA) to their respective acid metabolites
5-hydroxyindole-3-acetic acid (5-HIAA) and 3,4-dihy-
* To whom correspondence should be addressed. Tel: (617) 432-
4001. Fax: (617) 566-3137. E-mail: wingming_keung@hms.harvard.edu.
10.1021/jm990614i CCC: $19.00 © 2000 American Chemical Society
Published on Web 10/06/2000

4170 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Rooke et al.
substituted derivatives of the 4′,7-dihydroxyisoflavone,
daidzein, for this study.
The target 7-O-substituted benzopyranone compounds
were synthesized by selective 7-O-functionalization of
daidzein. The pKa’s of the 4′- and 7-hydroxyl groups of
daidzein are different enough (11.2 and 7.6, respectively)
to allow selective ionization of the 7-OH group under
appropriate reaction conditions, thus permitting regio-
selective functionalization.11 The derivatives were pre-
pared via Williamson synthesis, by reacting the 7-O-
monoaroxide anion of daidzein with an appropriate alkyl
halide RX (X ) Br, I). The 7-O-monoaroxide anion could
be obtained by reacting daidzein with 1 equiv of aqueous
KOH in acetone (when X ) Br) or with 1 equiv of
powdered KOH in DMSO (when X ) I).12,13
Structures of daidzin and its analogues included in
this study are shown in Figure 2. Compounds 1-13 are
the 7-O-substituted derivatives of daidzein synthesized
in this laboratory, whereas compounds 14-18 are
7-hydroxyisoflavones and flavones purchased from com-
mercial sources.
F igu r e 1. 5-HT metabolism in hamster liver mitochondria:
a proposed site of action for daidzin.
droxyphenylacetic acid (DOPAC) in isolated hamster
liver mitochondria, which is catalyzed by monoamine
oxidase (MAO) and ALDH-2.10 Further, we synthesized
six structural analogues of daidzin and studied and
compared their effects on 5-HT f 5-HIAA and DA f
DOPAC conversion in isolated liver mitochondria and
on ethanol intake in ethanol-preferring golden ham-
sters. Results of these studies suggested a potential link
between these two activities: the stronger the inhibition
of 5-HIAA (or DOPAC) formation, the greater the
suppression of ethanol intake. Moreover, inhibition of
5-HIAA formation by daidzin was accompanied by a
concomitant increase in 5-HIAL accumulation in the
assay medium. This, together with the fact that daidzin
does not affect the rate of 5-HT depletion in the same
assay, raised the possibility that the antidipsotropic
action of daidzin is not mediated by the monoamines
but rather by their reactive metabolic intermediates
such as 5-HIAL and DOPAL which accumulated in the
presence of daidzin (Figure 1). To further evaluate this
possibility, we here synthesized more structural ana-
logues of daidzin and tested and compared their anti-
dipsotropic activities in alcohol-drinking golden ham-
sters with their effects on mitochondria-catalyzed
monoamine metabolism using 5-HT as the substrate.
Effects of daidzin and its structural analogues on the
activities of MAO and ALDH-2, the key enzymes
involved in monoamine metabolism in the mitochondria,
were also examined and compared with their antidip-
sotropic activities.
Effect of Da id zin a n d Its Str u ctu r a l An a logu es
on Ha m ster Eth a n ol In ta k e. The antidipsotropic
activities of daidzin and its structural analogues were
determined using ethanol-preferring Syrian golden
hamsters, and the results are shown in Table 1. Daidzin
and eight of its structural analogues suppress hamster
ethanol intake when administered intraperitoneally at
a dose of 0.07 mmol/hamster/day. Among them, daidzin
(1) and the 7-O-ω-carboxyalkyl derivatives of daidzein
(3-6) are most potent: they suppress daily ethanol
intake by more than 60%. The aglycone daidzein (2) and
its 7-[O-2-(1,3-dioxanyl)ethyl] (10) and 7-O-ω-bromo-
hexyl (9) derivatives are moderately active: they sup-
press ethanol intake by about 22%, 32%, and 29%,
respectively. At the same dose tested, the 7-O-ω-bro-
mobutyl derivative of daidzein (8) is only marginally
active, whereas the 7-O-ω-bromopropyl (7), 7-O-allyl
(11), 7-O-(2,3-dihydroxypropyl) (12), and 7-O-ethyl (13),
5-hydroxy (14), and 5-hydroxy-4′-methyl (17) derivatives
are not active at all. Puerarin (18), the 8-C-glucosyl
derivative of daidzein, which has been shown to sup-
press ethanol intake in rats,3 does not suppress ethanol
intake in Syrian golden hamsters. None of the flavone
analogues tested (15, 16) are antidipsotropic.
Resu lts a n d Discu ssion
Relative Bioavailability of Daidzin an d Its Str u c-
tu r a l An a logu es. The relative potencies of daidzin and
its structural analogues in suppressing hamster ethanol
intake depend not only on their intrinsic pharmacologi-
cal activities but also on their relative bioavailabilities.
To determine whether differences in bioavailabilities of
these analogues contributed significantly to the ob-
served differences in antidipsotropic activities, we evalu-
ated and compared their basic pharmacokinetic prop-
erties under the same conditions employed in ethanol
drinking experiments. The results, expressed as relative
rate and extent of bioavailability, tmax and AUC (area
under the curve), respectively, are listed in Table 2.
Syn th esis of 7-O-Su bstitu ted Der ivatives of Daid-
zein . In an earlier study,5 we surveyed 41 commercially
available compounds structurally related to daidzin for
ALDH-2 inhibition and found that only a few of the
isoflavones tested inhibit ALDH-2 potently (Ki < 1 µM)
and selectively (Ki(ALDH-1)/Ki(ALDH-2) > 50). The
characteristics of the isoflavones tested show that the
substituents at the 4′, 5, and 7 positions are important
for ALDH-2 inhibition. Isoflavones with a free 4′-
hydroxyl group inhibit ALDH-2 more potently than
those with a blocked 4′-hydroxyl group. Substituting the
hydrogen with a hydroxyl group at position 5 decreases
potency by at least 1 order of magnitude. Furthermore,
blocking the 7 position hydroxyl group enhance ALDH-2
inhibition. In fact, the 7-O-glucosyl isoflavones are the
most potent yet selective inhibitors of ALDH-2 among
the naturally occurring isoflavones tested. On the basis
of these results, we decided to synthesize a series of 7-O-
The tmax of daidzin and its structural analogues vary
from 0.5 to 2 h. In the ethanol-drinking experiments,
accumulated ethanol intake over a period of 24 h was
measured. Therefore, it is unlikely that this relatively
small variation in tmax will have any significant effect

Potential Site of Action of Daidzin
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4171
F igu r e 2. Structures of daidzin and its analogues.
Ta ble 1. Suppression of Hamster Ethanol Intake by Daidzin
Ta ble 2. AUC and tmax of Daidzin and Its Structural
and Its Structural Analoguesa
Analoguesa
compd
% suppression
compd
% suppression
tmax
h
AUC
tmax
h
AUC
compd
nmol‚h‚mL-1
compd
nmol‚h‚mL-1
1
2
3
4
5
6
7
8
9
62 ( 4
22 ( 5
69 ( 12
69 ( 8
84 ( 5
86 ( 7
0
10
11
12
13
14
15
16
17
18
32 ( 10
0
0
0
0
0
0
0
0
1
2
3
4
5
6
7
8
9
1
1
120 ( 5
130 ( 16
77 ( 18.1
68 ( 14
73 ( 10
56 ( 14
73 ( 12.5
68 ( 17
62 ( 10
10
11
12
13
14
15
16
17
18
0.5
1
0.5
1
1
1
0.5
1
1
99 ( 22
29 ( 1.6
141 ( 11
38 ( 1.5
134 ( 25
110 ( 44
111 ( 22
99 ( 5.5
129 ( 7
0.5
0.5
0.5
0.5
2
14 ( 5
29 ( 5
2
2
a
Ethanol intake suppression was measured as described in the
Experimental Section. Dose ) 0.07 mmol/hamster/day, ip. Values
are mean ( SD of 4-6 hamsters.
a
For details of the method, refer to the Experimental Section.
Values are mean ( SD of 3 hamsters.
on the antidipsotropic activities determined for these
compounds.
The AUC values of daidzin and its antidipsotropic
analogues also are very similar. The AUC values
estimated for the active analogues vary by less than
3-fold, ranging from 56 nmol‚h‚mL for 6 to 130 nmol‚
h‚mL for 2. In fact, except for 1 and 2, which have
relatively high AUC values (120 and 130 nmol‚h‚mL,

4172 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Rooke et al.
Ta ble 3. Inhibition of 5-HIAA Formation in Isolated Hamster
respectively), the AUC values estimated for all other
active analogues are, within experimental error, virtu-
ally identical. These results indicate that differences in
relative bioavailability contribute little, if any, to the
differences in the antidipsotropic activities observed for
the active analogues.
Except for analogues 11 and 13, which have relatively
low AUC values (29 and 38 nmol‚h‚mL, respectively),
those estimated for the other inactive analogues (7, 12,
14-18) range from 73 to 141 nmol‚h‚mL, similar to or
slightly higher than those for the active analogues. This
result suggests that these analogues do not suppress
hamster ethanol intake because they are intrinsically
inactive, not because of poor bioavailability.
Effect of Da id zin a n d Its Str u ctu r a l An a logu es
on Mitoch on d r ia -Ca ta lyzed 5-HT f 5-HIAA Con -
ver sion . Monoamine metabolism and its inhibition by
daidzin and its structural analogues were studied in
isolated hamster liver mitochondria using 5-HT as the
substrate. Oxidative deamination of biogenic monoam-
ines is catalyzed by the membrane-bound MAO on the
outer surface of mitochondria. This reaction generates
reactive aldehyde intermediates which are either oxi-
dized to their corresponding acid metabolites by
ALDH-2 present in the matrix of the mitochondria14,15
or reduced to their corresponding alcohols by NADH-
dependent alcohol dehydrogenase (ADH) and/or NADPH-
dependent aldehyde reductase (AR).16,17 In liver and
brain, 5-HT is primarily converted to its acid product
5-HIAA, indicating that the mitochondrion is probably
an important subcellular compartment in which 5-HT
metabolism occurs in vivo. The mitochondria prepared
for this study contained both MAO and ALDH-2 but not
ADH or AR.10 Therefore, they provide a simple yet
physiologically relevant system in which the effects of
daidzin and its structural analogues on 5-HT metabo-
lism can be examined.
In an assay medium containing 10 µM 5-HT and 0.08
mg/mL of the isolated mitochondria, about 5% of the
substrate is converted to 5-HIAA after incubation at 37
°C for 30 min. Under these conditions, daidzin potently
inhibits the formation of 5-HIAA. Inhibition is concen-
tration-dependent with an estimated IC50 of about 2 µM
(Table 3), similar to that reported in a previous study.10
Among the daidzin analogues tested, all but analogue
18 inhibit 5-HT f 5-HIAA conversion. Analogues 5-7,
13, and 17 are the most potent with IC50 values <0.3
µM; followed by 3, 4, 8-12, 14 (0.3 µM < IC50 < 0.9
µM); 1, 2, 15 (0.9 µM < IC50 < 3 µM); and 16 (3 µM <
IC50 < 9 µM) (Table 3).
Figure 3A, a scatter plot of the antidipsotropic activi-
ties of daidzin and its structural analogues versus their
abilities in inhibiting mitochondria-catalyzed 5-HIAA
formation, shows that the two activities do not correlate
with each other. It appears that daidzin analogues
which suppress hamster ethanol intake (1-6, 8-10)
also inhibit 5-HIAA formation. However, not all daidzin
analogues that inhibit 5-HIAA formation are antidip-
sotropic. In fact, among the most potent inhibitors of
5-HT f 5-HIAA conversion, six (7, 11-14, 17) do not
suppress hamster ethanol intake (Figure 3A).
Liver Mitochondria by Daidzin and Its Structural Analoguesa
% inhibition
compd
0.3 µM
0.9 µM
3 µM
9 µM
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
18.7 ( 2.6
9.5 ( 1.6
48.4 ( 1.3
48.6 ( 3.2
74.4 ( 0.9
57.0 ( 5.9
51.1 ( 2.9
19.2 ( 2.1
47.2 ( 3.0
41.0 ( 6.0
37.9 ( 12.0
33.3 ( 0.3
64.8 ( 11.0
31.3 ( 3.2
20.7 ( 3.1
4.6 ( 2.0
74.1 ( 6.8
0
43.7 ( 3.5
29.2 ( 5.8
72.7 ( 0.1
71.2 ( 0.8
85.1 ( 3.2
72.7 ( 2.0
67.7 ( 5.7
55.3 ( 10.9
71.2 ( 2.5
57.6 ( 4.7
68.8 ( 7.6
63.8 ( 1.9
86.9 ( 5.4
65.0 ( 5.6
28.1 ( 3.0
7.9 ( 2.5
90.7 ( 6.2
0
64.4 ( 2.2
58.9 ( 0.9
95.1 ( 6.9
87.4 ( 6.7
89.6 ( 3.8
82.5 ( 6.6
90.8 ( 4.7
84.7 ( 6.5
88.5 ( 4.4
75.3 ( 9.3
89.3 ( 0.6
85.0 ( 0.1
95.6 ( 24.4
77.6 ( 5.2
64.7 ( 4.1
25.4 ( 4.1
96.8 ( 4.9
0
76.3 ( 1.9
80.6 ( 2.5
98.9 ( 1.6
94.6 ( 8.1
95.0 ( 3.5
96.7 ( 4.7
100
98.3 ( 2.5
95.7 ( 4.3
88.5 ( 5.4
100
95.1 ( 0.6
100
96.8 ( 5.3
83.5 ( 6.8
53.2 ( 1.6
100
0
a
See the Experimental Section for assay conditions. Values are
mean ( SD of 3 separate determinations.
5-HIAA in the isolated mitochondria (Figure 1). There-
fore, in principle, daidzin and its structural analogues
can inhibit 5-HIAA formation by blocking either MAO
and/or ALDH-2. To identify the target enzyme(s) of
daidzin and its structural analogues, we studied their
effects on MAO and ALDH-2 independently using the
membrane and lysate of a density-gradient-purified
mitochondria preparation as the respective enzyme
sources. The results, expressed in IC50 values, are listed
in Table 4.
As we reported previously,5 daidzin is a potent inhibi-
tor of ALDH-2. The IC50 value, determined at a sub-
strate concentration equal to Km, is 0.04 µM. Further,
at concentrations up to 10 µM, daidzin does not inhibit
MAO activity. This result is consistent with our previous
finding that daidzin at concentrations that significantly
inhibit 5-HIAA formation does not affect the rate of
5-HT depletion catalyzed by isolated hamster liver
mitochondria.10
The 7-O-ω-carboxyalkyl derivatives of daidzein (3-
6) are by far the most potent ALDH-2 inhibitors
synthesized thus far. The IC50 values determined under
the same conditions range from 0.009 to 0.003 µM, 4.4
to 13.3 times lower than that of daidzin. Analogues 2
and 7-13 also inhibit ALDH-2 but are relatively less
potent with IC50 values ranging from 0.04 to 9 µM.
Further, unlike daidzin, these analogues also inhibit
MAO with 7 (IC50 ) 0.15 µM), 13 (0.3 µM), and 11 (0.45
µM) being the most potent, followed by 12 (1.7 µM), 9
(2 µM), 4 (2.1 µM), 3 (3 µM), 8 (4 µM), 10 (7 µM), 5 (10
µM), 6 (13 µM), and 2 (14 µM).
The isoflavone aglycones 14 and 17 and the flavone
aglycones 15 and 16 do not inhibit ALDH-2. However,
they all inhibit MAO with 17 being the most potent (IC50
) 0.4 µM) followed by 14 (0.9 µM), 15 (1.6 µM), and 16
(17 µM). Analogue 18, at concentrations up to 10 µM,
does not affect ALDH-2 or MAO.
Analysis of a scatter plot of ethanol intake suppres-
sion versus ALDH-2 inhibition (Figure 3B) shows that
these two activities do not correlate with each other.
Nevertheless, the data appear to show that (i) ALDH-2
inhibition is necessary for antidipsotropic activity be-
Effect of Da id zin a n d Its Str u ctu r a l An a logu es
on MAO a n d ALDH-2 Activity. MAO and ALDH-2
act in tandem in the catalytic conversion of 5-HT f

Potential Site of Action of Daidzin
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4173
F igu r e 3. Scatter of ethanol intake suppression by daidzin and its structural analogues against their inhibition on (A) 5-HIAA
formation in liver mitochondria-catalyzed 5-HT metabolism, (B) ALDH-2 activity, (C) MAO activity, and (D) IC50(MAO)/IC50
-
(ALDH-2). Ethanol intake suppression were measured using ethanol-preferring golden hamsters at a dose of 0.07 mmol/hamster/
day, ip. Effects of daidzin and analogues on ALDH-2, MAO, and 5-HIAA formation were determined at a concentration of 0.9 µM.
IC50(MAO)/IC50(ALDH-2) ratios were calculated from data shown in Table 3. Analogues that do not inhibit both MAO and ALDH-2
are not included in plot D because their IC50(MAO)/IC50(ALDH-2) values are undefined.
Ta ble 4. Inhibition of Hamster Liver Mitochondrial MAO and
contributed negatively to antidipsotropic activity. As
shown in Figure 3B,C, daidzin analogues that inhibit
ALDH-2 Activities by Daidzin and Its Structural Analoguesa
IC50, µM
IC50, µM
ALDH-2 potently but which have little or no effect on
MAO (1, 3-6) are most antidipsotropic, whereas those
which also inhibit MAO potently do not suppress
ethanol intake (7, 11-13). These results appear to show
that the antidipsotropic activities of daidzin and its
active analogues may stem from their abilities to
increase the mitochondrial MAO:ALDH-2 activity ratio.
compd
MAO
nib
14
ALDH-2
compd
MAO
ALDH-2
1
2
3
4
5
6
7
8
9
0.04
9
10
11
12
13
14
15
16
17
18
7
0.4
0.8
0.1
0.04
ni
0.45
1.7
0.3
0.9
1.6
3
2.1
0.009
0.009
0.004
0.003
0.26
0.27
0.3
10
13
0.15
ni
To evaluate this possibility, we constructed a scatter
plot of suppression of ethanol intake by daidzin and its
structural analogues versus their abilities to increase
MAO:ALDH-2 activity ratio, IC50(MAO)/IC50(ALDH-2)
(Figure 3D). This plot clearly shows a positive correla-
tion between the two and suggests that the antidipso-
tropic isoflavones may indeed suppress hamster ethanol
intake by increasing the mitochondrial MAO:ALDH
activity ratio.
17
0.4
ni
ni
ni
ni
4
2
a
MAO and ALDH-2 activity were assayed as described in the
Experimental Section, using 10 µM 5-HT and 0.6 mM formalde-
hyde as the substrates, respectively. IC50 values were estimated
graphically with inhibition data determined at more than five
b
inhibitor concentrations. ni, no inhibition up to 10 µM.
cause all structural analogues that do not inhibit
ALDH-2 (14-18) are not antidipsotropic and (ii)
ALDH-2 inhibition alone is not sufficient for ethanol
intake suppression because not all daidzin analogues
that inhibit ALDH-2 are antidipsotropic. In fact, four
of the most potent ALDH-2 inhibitors (7, 11-13) exhibit
no effect on hamster ethanol intake.
Among the 18 compounds tested, 16 inhibit MAO. The
scatter plot of MAO inhibition versus ethanol intake
suppression by these compounds (Figure 3C) clearly
shows no correlation between the two. In fact, the MAO
inhibitory activity of some daidzin analogues may have
In this context, it is of interest to point out that in a
previous study,10 we showed that hamster liver mito-
chondria exhibit a lower MAO:ALDH-2 activity ratio
(0.18) than that of the rat (1.6). As a consequence, the
concentration of the metabolic intermediate 5-HIAL
found in isolated hamster liver mitochondria during
5-HT metabolism is also much lower than that in the
rat (0.2 µM vs 2.3 µM). We also pointed out that golden
hamsters are by nature inclined to prefer and consume
large quantities of ethanol,18 whereas the randomly bred
Wistar rats used in our studies avoid ethanol.19 Epide-

4174 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Rooke et al.
miological studies also have associated a low MAO:
ALDH-2 activity ratio with high ethanol consumption:
(i) low platelet MAO activity correlates with type II
alcoholism20 and (ii) Asians who have inherited a low
activity (or inactive) mutant form of ALDH-2 seldom
have a problem with alcohol abuse.21 These findings,
although inconclusive, are consistent with the hypoth-
esis that the antidipsotropic action of daidzin may be
mediated by a biogenic aldehyde intermediate of the
mitochondrial MAO/ALDH pathway such as 5-HIAL. To
further evaluate this hypothesis, we studied the effects
of daidzin and its structural analogues on the ac-
cumulation of 5-HIAL during 5-HT metabolism in
hamster liver mitochondria.
Effects of Da id zin a n d Its Str u ctu r a l An a logu es
on 5-HIAL Accu m u la tion d u r in g 5-HT Meta bolism
in Isola ted Ha m ster Liver Mitoch on d r ia . All but
analogue 18 affect 5-HIAL accumulation during 5-HT
metabolism. On the basis of their effects, these com-
pounds can be classified into four groups. Group I, which
includes only daidzin, increases 5-HIAL accumulation
during mitochondria-catalyzed 5-HT metabolism (Fig-
ure 4A). Within the concentration range studied, the
effect is concentration-dependent and is accompanied
by a concomitant decease in 5-HIAA formation. At 9 µM,
the highest daidzin concentration studied, 5-HIAL ac-
cumulation in the assay medium increased by nearly
200% (Table 5). These results are consistent with the
finding that daidzin is a potent inhibitor of ALDH-2 but
has no effect on MAO (Table 4).
Group II analogues, which include compounds 2-6
and 8-10, affect 5-HIAL accumulation in a biphasic
manner: they increase 5-HIAL accumulation at low
concentrations but decrease it at high concentrations.
Figure 4B shows a typical biphasic concentration-
dependent effect of this group of analogues on 5-HIAL
accumulation. These results are consistent with the fact
that this group of analogues are all more potent ALDH-2
inhibitors than they are MAO inhibitors (Table 4).
Group III comprises analogues 7 and 11-17, which
are strong MAO but relatively weak ALDH-2 inhibitors
(Table 4). As a consequence, they inhibit 5-HT f
5-HIAL conversion more potently and, hence, decrease
both 5-HIAL accumulation and 5-HIAA formation in the
mitochondria (Figure 4C, Table 4). Group IV, repre-
sented solely by analogue 18, affects neither 5-HIAL
accumulation nor 5-HIAA formation during 5-HT me-
tabolism in isolated hamster liver mitochondria (Table
5), consistent with the finding that analogue 18 has no
effect on either ALDH-2 or MAO (Table 4).
Comparing the abilities of daidzin and its structural
analogues in increasing 5-HIAL accumulation and sup-
pressing hamster ethanol intake reveals a positive
correlation (Figure 5). This result, together with that
shown in Figure 3, suggests that the mitochondrial
MAO/ALDH pathway is a potential target pathway of
daidzin and that a biogenic aldehyde intermediate, such
as 5-HIAL, derived from the action of MAO may be
involved in the regulation of hamster ethanol intake and
its suppression by daidzin.
F igu r e 4. Concentration effect of (A) daidzin, (B) analogue
10, and (C) analogue 13 on hamster liver mitochondria-
catalyzed serotonin metabolsim: formation of 5-HIAL and
5-HIAA.
pathway exists and operates not only in the liver but
also in other organs and tissues as well. Therefore, our
results provide no information on where daidzin act in
vivo. Daidzin can either act on the mitochondrial MAO/
ALDH pathway in the liver and transmit its antidip-
sotropic signal into the brain via a second messenger
or act directly on a similar pathway that operates in
the central nervous system.
Con clu d in g Rem a r k s
Early interest in biogenic aldehydes in relation to
alcohol research stems largely from the belief that
acetaldehyde, the reactive intermediate of ethanol
metabolism, interferes with their oxidative metabo-
lism.22 It was postulated that levels of biogenic alde-
hydes increase during ethanol metabolism because of
competitive inhibition of ALDH-2 by acetaldehyde.
Biogenic aldehydes, accumulated under such conditions,
can be diverted to a reductive pathway leading to the
formation of their alcohol metabolites16,17 and/or un-
dergo nonenzymatic condensation reactions forming
While the correlation coefficients obtained from the
scattered analyses (Figures 3 and 5) are good, we are
aware of the fact that correlations do not prove a
mechanism. Moreover, the mitochondrial MAO/ALDH

Potential Site of Action of Daidzin
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4175
Ta ble 5. Effect of Daidzin and Its Structural Analogues on
intake. Our results suggest that daidzin, and its active
antidipsotropic analogues, could act by mimicking the
effect of acetaldehyde on biogenic aldehyde metabolism
and provided a strong case for further pursuit of the
role biogenic aldehydes in the regulation of alcohol use
and abuse.
5-HIAL Accumulation in Isolated Hamster Liver Mitochondriaa
5-HIAL accumulation in isolated liver mitochondria,
% of control
compd
0.3 µM
0.9 µM
3 µM
9 µM
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
146.5 ( 21.0 203.5 ( 22.8 268.3 ( 53.0 299.0 ( 59.0
103.5 ( 2.0
229.0 ( 6.4
110.5 ( 3.5
103.5 ( 9.2
74.0 ( 4.2
182.5 ( 40.0 162.0 ( 35.0 130.5 ( 14.8
Exp er im en ta l Section
310.0 ( 21.2 265.0 ( 39.2 227.6 ( 12.6 192.0 ( 25.3
General chemicals were either purchased from Aldrich
Chemical Co. (Milwaukee, WI) or Lancaster Synthesis Inc.
(Windham, NH). All organic solvents used were of AR grade
and were supplied by J . P. Baker (Phillipsburg, NJ ) or Fisher
Scientific Co. (Pittsburgh, PA). Daidzin (1) was purchased from
L. C. Laboratories (Woburn, MA). Genistein (14), chrysin (15),
7,8-dihydroxyflavone (16), and biochanin A (17) were pur-
chased from Indofine Chemical Co. (Somerville, NJ ). Daidzein
(2) was prepared either in this laboratory or by Tyger Scientific
Inc., Princeton, NJ . Puerarin (18) and 5-hydroxytryptophol (5-
HTOL) were purchased from Sigma Chemical Co. (St. Louis,
MO). Serotonin (5-HT) and its metabolite 5-HIAA were
products of Research Biochemical International (Natick, MA).
The metabolic intermediate 5-HIAL was produced in this
laboratory by MAO-catalyzed oxidative deamination of 5-HT
using rat liver mitochondrial membrane as a source of MAO.23
All other reagents used were best grade available.
234.0 ( 5.7
233.0 ( 19.8 189.0 ( 4.2
129.0 ( 4.2
214.0 ( 69.5 292.0 ( 66.5 294.0 ( 55.2 205.5 ( 44.5
75.8 ( 5.0
111.5 ( 5.6
143.6 ( 8.6
57.8 ( 9.8
115.7 ( 9.5
112.5 ( 9.9
34.7 ( 6.0
87.7 ( 26.0
108.3 ( 13.7
141.1 ( 22.0
56.1 ( 0.2
63.0 ( 4.9
34.3 ( 9.8
43.4 ( 6.2
33.6 ( 4.6
79.1 ( 2.5
0
20.5 ( 4.0
42.9 ( 9.8
74.0 ( 26.0
86.1 ( 27.9
26.7 ( 1.6
30.5 ( 1.7
17.8 ( 3.3
23.5 ( 3.3
17.2 ( 2.8
59.9 ( 3.6
0
165.0 ( 24.4 175.8 ( 9.1
106.3 ( 5.2
111.9 ( 7.5
88.6 ( 12.1
97.5 ( 13.2
73.5 ( 12.3
92.6 ( 3.9
26.3 ( 1.8
87.9 ( 6.9
91.0 ( 0
97.5 ( 10.7
62.9 ( 6.9
66.7 ( 8.6
55.6 ( 3.8
92.3 ( 5.7
12.9 ( 1.3
94.9 ( 16.5 109.1 ( 2.5
130.3 ( 23.7
a
For details of the assay, refer to the Experimental Section.
Values are mean ( SD of 3 separate determinations.
An im a l Exp er im en ts. 1. Eth a n ol-Dr in k in g Exp er i-
m en ts. Mature male Syrian golden hamsters (80-120 g) were
obtained from Harlan Sprague-Dawley Inc. (Indianapolis, IN).
Animals were housed in a room maintained at 23 °C on a 12/
12-h light/dark cycle with ad libitum access to Purina rodent
laboratory chow (5001), water, and a 15% ethanol solution
prepared in tap water. After 1 week, hamsters were trans-
ferred to and housed in individual stainless steel metabolic
cages (26 × 18 × 17.5 cm3). Two 50-mL drinking bottles, one
containing tap water and the other a 15% ethanol solution,
were provided. The drinking bottles were placed in holders
equipped with tilted platforms which collect spillage in tubes
placed outside of the cages. The positions of the two drinking
bottles on each cage were alternated daily to prevent develop-
ment of positional preference. Fluid intake was measured
between 3:00 and 4:00 p.m. each day. Hamsters that drank
significant (>8 mL/day) and consistent (daily variance less
than (10%) amounts of ethanol solution were selected for drug
testing. To establish baseline fluid intake, 1 mL saline with
0.5% DMSO was administered (ip) to each hamster between
5:00 and 5:30 p.m. each day for 6 consecutive days. They were
then divided into groups of six animals each and given daily
doses (ip) of 0.07 mmol of daidzin or one of its structural
analogues for 6 consecutive days. All test compounds were
administered in 1 mL saline containing 0.5% DMSO. Food,
water and a 15% ethanol solution were available continuously.
The response of each hamster to daidzin or its analogues is
expressed as: percent (%) ethanol intake suppression (%
suppression) ) (Vo - Ve) × 100/Vo, where Vo and Ve are the
average daily intakes of 15% ethanol solution during the
baseline and treatment periods, respectively.
2. Tim e Cou r se of P la sm a Da id zin (or An a logu e)
Con cen tr a tion a n d Ar ea Un d er th e Cu r ve (AUC). Golden
hamsters used for this study were selected according to the
same criteria as those for ethanol drinking experiments.
Selected animals were also housed under the same conditions
with ad libitum access to Purina rodent laboratory chow
(5001), water, and a 15% ethanol solution throughout the
study. Three hamsters were used for each compound tested.
Before the administration of a test compound, the hamster was
anesthetized with diethyl ether and a 0-time blood sample (50
µL) was drawn from the orbital venous plexus into a tube
containing 5 µL of 9% K3EDTA. Animals were then given (ip)
0.07 mmol of a test compound in 1 mL saline containing 0.5%
DMSO. Blood samples (50 µL) were taken 0.5, 1, 2, 4, 8, 16 h
after drug administration. Plasma samples, obtained by
centrifugation, were extracted with 10 volumes of methanol/
85% perchloric acid (8:2). Protein precipitates were removed
F igu r e 5. Scatter of suppression of ethanol intake versus
increase in 5-HIAL accumulation by daidzin and its structural
analogues. Data on ethanol intake suppression were taken
from Table 1. Values of increase in 5-HIAL accumulation are
the sum of 5-HIAL accumulation (% increase) caused by a test
compound at 1-4 different concentrations (Table 5). Using
5-HIAL accumulation measured at a single drug concentration
could lead to wrong conclusion because the effect of group II
analogues on 5-HIAL accumulation is biphasic: increase
accumulation at low concentration but decrease it at high
concentration. Further, the sums of % increase in HIAL
accumulation measured at a series of drug concentrations
better approximate the real exposure (AUC) of a hamster to
5-HIAL in in vivo experiments (ethanol-drinking experiment).
Group III analogues, which inhibit 5-HIAL accumulation at
all concentrations studied, neither decrease nor increase
hamster ethanol intake (Tables 1 and 5) and are, therefore,
not included in this plot. For the same reason, the decrease in
5-HIAL accumulation caused by some group II analogues at
high drug concentrations is not included in the calculation of
the sum.
adducts with biogenic amines, proteins, and phospho-
lipids.22,23 While the physiological implication of shifting
from an oxidative to a reductive metabolic pathway is
completely unknown at this time, the condensation
products have been shown to affect ethanol-drinking
behavior in laboratory animals.24 Moreover, biogenic
aldehydes themselves could be physiologically active25
and may play a more direct role in regulating ethanol

4176 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Rooke et al.
by centrifugation. Supernatants were concentrated 10 times
(Speedvac, Savant) and were analyzed for the test compound
on a Waters HPLC system according to conditions described
in the Synthesis section. Plasma drug concentration-time
curves were constructed from these data and the AUCs
between the 0th and 16th hours were estimated by the
trapezoid method. Calibration curves were prepared using
solutions of standard compounds (0.5-10 µM) in methanol.
Stock solutions were stored at -20 °C for 1 month, and
working standard solutions were freshly prepared. Standards
were analyzed as control during the analysis of plasma
samples. The recoveries of test compounds from blood were
determined in control blood samples spiked with a single
concentration of one compound each time. The intra-assay
coefficient of variation determined at 1 and 10 µM for all tested
compounds was e17.3% and 10.6%, respectively. The calibra-
tion curves were linear in all cases.
contained only MAO but not ALDH-2 activity, was used for
MAO assay. ALDH-2 activity was assayed in 0.1 M NaPPi,
pH 9.5, containing 0.15 M KCl, 1.2 mM NAD+, 0.6 mM
formaldehyde, and specified concentrations of daidzin or its
structural analogues. Activity was determined by following the
increase in absorbance at 340 nm with a Varian Cary 1
spectrophotometer at 25 °C.23 MAO activity was assayed in
TKK buffer containing 10 µM 5-HT, 0.4 mM sodium bisulfite,
specified concentrations of daidzin or its structural analogues,
and MAO. Enzyme reaction was initiated by the addition of
enzyme and was allowed to proceed at 37 °C for 30 min. The
reaction was terminated by centrifugation at 4 °C in a Sorvall
Microspin at top speed for 15 min. The reaction product
5-HIAL, present in the supernatant as its stable bisulfite
complex, was liberated by diluting the supernatant 10-100-
fold in 50 mM NaPPi, pH 8.8 and analyzed by HPLC. Since
5-HIAL is relatively unstable at alkaline pH, 5-HIAL was
liberated not more than 4 h before HPLC analysis. The overall
recovery of 5-HIAL and 5-HIAA in assay samples spiked with
standard analytes were 0.78 and 0.86, and the intra-assay
coefficient of variation of the analytical methods determined
with samples spiked with 2 µM of the respective analytes are
11.2% and 7.5%. Effect of daidzin and its analogues on ALDH-2
and MAO activities is expressed as: percent (%) inhibition )
(Ao - Ae) × 100/Ao, where Ao and Ae are enzyme activities
measured in the absence and presence of a test compound,
respectively.
Meta bolic a n d En zym e Assa ys. 1. Ser oton in Meta bo-
lism in Isola ted Liver Mitoch on d r ia . Ethanol-naive ham-
sters were sacrificed in a CO2 chamber. Livers were removed
immediately and isolated mitochondria were prepared as
described previously.23 The integrity of these mitochondrial
preparations, evaluated by measuring their latent glutamate
dehydrogenase activity before and after the metabolic study,
were over 97% and 93%, respectively. Mitochondria-catalyzed
5-HT metabolism was carried out in a 0.5-mL standard assay
medium containing 10 mM Tris-HCl (pH 7.4), 0.3 M mannitol,
2.5 mM MgCl2, 10 mM K2HPO4, 10 mM KCl, 0.1% DMSO,
0.01% ethanol, 10 µM 5-HT, and specified concentrations of
test compounds and freshly prepared isolated mitochondria.
Most test compounds are not readily soluble in aqueous
solution. Therefore, stock solutions of the test compounds were
prepared in 90% DMSO and 10% ethanol. The final concentra-
tions of DMSO and ethanol in all assays, including controls,
were 0.1% and 0.01%, respectively, and in which all tested
compounds were soluble at the highest concentration tested,
i.e., 9 µM. Reactions were initiated by the addition of mito-
chondria and allowed to proceed in a 37 °C shaking water bath
for 30 min. Reactions were terminated by the addition of 0.05
mL each of ice-cold 1 M HClO4 and 10 mM EDTA. Samples
were kept on ice for 30 min and then centrifuged at 12 000
rpm in a microcentrifuge (Microspin 24S, Sorvall) for 15 min.
The metabolic intermediate 5-HIAL and product 5-HIAA in
the supernatant were analyzed by HPLC. In the presence of
HClO4 (final concentration ) 0.1 M), both 5-HIAL and 5-HIAA
are stable for at least 12 h at 4 °C. The calibration curves
obtained with standard 5-HIAA and 5-HIAL solutions (range
from 0.05-10 µM) are linear. At Eapp ) 650 mV and I ) 20
nA, on column detection limit for both 5-HIAA and 5-HIAL
was 0.2 ng with a signal-to-noise ratio of 3. The overall
recovery of 5-HIAL and 5-HIAA in the mitochondrial assay
samples spiked with standard compounds was 0.65 and 0.84,
and the intra-assay coefficient of variation of the analytical
methods determined with samples spiked with 2 µM of the
respective analytes was 12.5% and 8.6%. Effect of daidzin and
its analogues on 5-HIAA formation during the assay is
expressed as: percent (%) inhibition ) (5-HIAAo- 5-HIAAe)
× 100/5-HIAAo, where 5-HIAAo and 5-HIAAe are 5-HIAA
formation measured in the absence and presence of a test
compound, respectively. Effect of daidzin and its analogues on
5-HIAL accumulation during the assay is expressed as: per-
cent (%) of control ) (5-HIALe/5-HIALo) × 100, where 5-HIALe
and 5-HIALo are concentrations of 5-HIAL found in assay
media with and without a test compound, respectively.
3. HP LC An a lysis. The HPLC system for 5-HIAL and
5-HIAA analyses consisted of a BAS Sample Sentinel au-
tosampler with refrigerated sample compartment (set at 4 °C
for all analyses), a PM80 solvent delivery system, and a LC-
26 on-line degasser. The detector is a LC-4C amperometric
controller with a CC-5 cross-flow thin-layer (0.005 in.) elec-
trochemical cell composed of glassy carbon and silver/silver
chloride reference electrode (Bioanlytical Systems, Inc., West
Lafayette, IN). For routine analysis, the potential and sensi-
tivity were set at 650 mV and 20 nA full scale, respectively.
Column temperature was maintained with a Waters temper-
ature control module (Waters, Milford, MA). Metabolites
5-HIAL and 5-HIAA were analyzed on a Beckman ultrasphere
ODS, 5-µm, 4.6- × 250-mm column. The column was developed
at 37 °C at 1 mL/min in a mobile phase containing 3%
methanol (v/v), 1% acetonitrile (v/v), 0.2 mM EDTA, and 0.1%
TFA (v/v). The retention times for 5-HIAL and 5-HIAA were
14.2, and 21.4 min, respectively. Data were collected and
analyzed with a Waters 740 data module.
Syn th esis. Gen er a l Meth od s. 1H and 13C NMR spectra
were obtained on a Bruker AMX500 BQ spectrometer at 500
MHz and Bruker AM-500 spectrometer at 126 MHz (NuMega
Resonance Labs Inc., San Diego, CA) respectively, using
DMSO as solvent and as internal standard (2.50 and 39.51
1
ppm for H and 13C, respectively), unless otherwise indicated.
Mass spectra were obtained from NuMega Resonance Labs Inc.
on a Perkin-Elmer PE-SCIEX API 100 mass spectrometer by
infusion. Samples were ionized by electrospray and spectra
were recorded both in positive and negative modes (mass
spectral data are reported in this order). Melting points were
determined with a Hoover capillary melting point apparatus.
Purity of all daidzin analogues synthesized in this laboratory
was verified on a Waters HPLC system consisting of a 717
plus autosampler, 600s controller, 626 solvent delivering
system, and 440 absorbance detector. Samples were analyzed
on a Beckman ultrasphere ODS 5-µm, 4.6- × 250-mm column
under two elution conditions: (a) a 30-min linear gradient:
12-80% acetonitrile in 0.1% TFA and (b) an isocratic condition
in a mobile phase containing 0.1% TFA and a specified
concentration of acetonitrile. Flow rates were set at 1 mL/min.
7-H yd r oxy-3-(4-h yd r oxyp h en yl)-4H -1-b en zop yr a n -4-
on e (Da id zein , 2). A solution of 5.000 g (45.41 mmol) of
resorcinol and 6.910 g (45.41 mmol) of p-hydroxyphenylacetic
acid in 111.7 mL (0.908 mol) of freshly distilled BF3 etherate
was stirred at 70 °C (oil bath) for 24 h in a 500 mL three-neck
round-bottom flask fitted with a dropping funnel and a reflux
condenser. The resulting yellow suspension was cooled to room
2. MAO a n d ALDH-2 Assa ys. The mitochondrial pellet
obtained from 5 g of hamster liver was resuspended in 10 mL
of 10 mM sodium phosphate buffer (pH 7.4), kept on ice, and
sonicated for 3 × 15 s at 90 W of power with a Branson Sonifier
cell disruptor. This suspension was centrifuged at 105000g for
70 min in a Beckman L8 ultracentrifuge and the supernatant,
which contained ALDH-2 activity, was used for ALDH-2 assay.
The pellet, which contained mainly mitochondrial membrane,
was washed 3 times in 30 mL TKK buffer (10 mM Tris, 10
mM KCl, and 10 mM KPi, pH 7.4). The final pellet, which

Potential Site of Action of Daidzin
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4177
temperature, and 70 mL of dry DMF was added slowly via
the dropping funnel. The reaction mixture was then heated
to 50 °C, and a solution of 11 mL of mesyl chloride in 18 mL
of dry DMF was added slowly. The resulting orange mixture
was stirred at 65-70 °C and the reaction was monitored by
HPLC. After about 4.5 h, the reaction mixture was quickly
poured into 200 mL of ice water. The pH of the solution was
adjusted slowly to 7 with 30% NaOH at 0 °C and the product
was allowed to precipitate overnight at 5 °C. Precipitates were
collected by filtration, washed 4 times each with 20 mL of H2O,
and dried over P2O5 to give 10.3 g (91% yield) of crude product.
Recrystallization of this product from EtOH/CHCl3/ether gave
a total of 7.695 g of pure 2, a 67% yield: mp 310-315 °C dec;
1H NMR (DMSO-d6) δ 8.27 (s, 1 H, H-2), 7.95 (d, J ) 8.5 Hz,
1 H, H-5), 7.37 (d, J ) 9.0 Hz, 2 H, H-2′), 6.93 (d, J ) 8.5 Hz,
1 H, H-6), 6.84 (s, 1 H, H-8), 6.79 (d, J ) 9.0 Hz, 2 H, H-3′);
13C NMR (DMSO-d6) δ 174.72 (C-4), 162.51 (C-7), 157.45,
157.20, 152.79 (C-2), 130.10 (C-2′, 6′), 127.30 (C-5), 123.52 (C-
3), 122.58 (C-1′), 116.66 (C-4a), 115.13 (C-3′, 5′), 114.98 (C-6),
102.12 (C-8); MS m/z 255 (M + H+), 277 (M + Na+), 293 (M +
K+), 253 (M - H+), 289 (M + Cl-), 367 (M + CF3COO-); HPLC
tR 14.9 min (gradient), 14.2 min (24% acetonitrile, isocratic).
J ) 8.4 Hz, 2 H, H-2′), 7.02 (s, 1 H, H-8), 7.00 (d, J ) 8.9 Hz,
1 H, H-6), 6.82 (d, J ) 8.4 Hz, 2 H, H-3′), 4.02 (t, J ) 7.0 Hz,
2 H, CH2O), 1.88 (t, J ) 7.1 Hz, 2 H, CH2CdO), 1.70 (quintet,
J ) 7.3 Hz, 2 H), 1.46 (quintet, J ) 7.0 Hz, 2 H), 1.37 (quintet,
J ) 7.0 Hz, 2 H), 1.29 (quintet, J ) 7.0 Hz, 2 H); 13C NMR
(DMSO-d6) δ 174.72 (C-4, COOH), 163.03 (C-7), 157.33 (C-8a,
4′), 152.84 (C-2), 129.85 (C-2′, 6′), 126.84 (C-5), 123.75 (C-3),
121.59 (C-1′), 117.42 (C-4a), 115.19 (C-3′, 5′), 114.87 (C-6),
100.83 (C-8), 68.51 (C-O), 38.29, 29.08, 28.53, 26.27, 25.38;
MS m/z 383 (M + H+), 421 (M + K+), 253 (M - (CH2)6CO2H+),
381 (M - H+), 417 (M + Cl-), 495 (M + CF3COO-); HPLC tR
22.1 min, (gradient), 41.4 min (32% acetonitrile, isocratic).
7-O-(9-Ca r boxyn on yl)-3-(4-h yd r oxyp h en yl)-4H-1-ben -
zop yr a n -4-on e (7-O-ω-Ca r boxyn on yld a id zein , 5). To a
suspension of 1.362 g (5.36 mmol) of 2 and 22 mL of acetone
was added 5.35 mL of 2 N aq KOH (10.72 mmol). The resulting
mixture was stirred vigorously at room temperature until 2
was completely dissolved. To this solution, 1.345 g of 10-
bromodecanoic acid (5.36 mmol) was added and the reaction
mixture was stirred under gentle reflux for 3 days. The
carboxylic acid potassium salt formed was precipitated at 5-6
°C for 12 h, collected on a fritted funnel, and washed with small
portions of cold acetone to give 675 mg of pure 5: mp 248-
7-O-(5-Ca r boxyp en tyl)-3-(4-h yd r oxyp h en yl)-4H-1-ben -
zop yr a n -4-on e (7-O-ω-Ca r boxyp en tyld a id zein , 3). To a
suspension of 5.000 g (19.68 mmol) of 2 and 75 mL of acetone
was added 19 mL of 2 N aq KOH (39.37 mmol). The resulting
mixture was stirred vigorously at room temperature until 2
was totally dissolved. To this solution, 6-bromohexanoic acid
(3.837 g, 19.68 mmol) was added and the reaction mixture was
stirred under gentle reflux for 3 days. After 12 h at 4 °C, the
carboxylic acid potassium salt precipitate was collected on a
fritted funnel and washed with small portions of cold acetone.
The product was resuspended in 75 mL of H2O, and the pH of
the solution was adjusted to 2 with 1 N HCl. The white fluffy
precipitates were collected on a fritted funnel, rinsed with cold
H2O until the pH of the filtrate was close to neutral, and dried
over P2O5 to give 1.410 g of the desired product 3. Unreacted
2 was recovered by concentrating the acetone/aqueous KOH
filtrate on rotary evaporator. The remaining aqueous solution
was acidified to pH 3 with 1 N HCl, kept at 5 °C for 12 h, and
the precipitates were collected and washed with H2O to recover
3.963 g of 2. Compound 3: mp 223-225 °C; 1H NMR (DMSO-
d6) δ 9.58 (broad s, 1 H, 4′-OH), 8.35 (s, 1 H, H-2), 8.00 (d, J
) 8.9 Hz, 1 H, H-5), 7.39 (d, J ) 8.6 Hz, 2 H, H-2′), 7.12 (d, J
) 2.3 Hz, 1 H, H-8), 7.06 (dd, J ) 8.9, 2.3 Hz, 1 H, H-6), 6.80
(d, J ) 8.6 Hz, 2 H, H-3′), 4.11 (t, J ) 7.0 Hz, 2 H, CH2O),
2.24 (t, J ) 7.3 Hz, 2 H, CH2CdO), 1.76 (quintet, J ) 6.9 Hz,
2 H), 1.57 (quintet, J ) 7.4 Hz, 2 H), 1.43 (quintet, J ) 7.0
Hz, 2 H); 13C NMR (DMSO-d6) δ 174.70 (C-4), 174.42 (CO2H),
163.03 (C-7), 157.40 (C-8a), 157.23 (C-4′), 153.09 (C-2), 130.06
(C-2′, 6′), 126.90 (C-5), 123.67 (C-3), 122.38 (C-1′), 117.50 (C-
4a), 114.96 (C-3′, 5′, 6 overlap), 100.95 (C-8), 68.35 (C-O),
33.59, 28.15, 25.05, 24.22; MS m/z 369 (M + H+), 391 (M +
Na+), 407 (M + K+), 253 (M-[(CH2)5CO2H]+), 367 (M - H+),
403 (M + Cl-), 481 (M + CF3COO-); HPLC tR 20.3 min
(gradient), 19.5 min (32% acetonitrile, isocratic).
1
251 °C; H NMR (DMSO-d6, 58 °C) δ 8.28 (s, 1 H, H-2), 8.01
(d, J ) 8.9 Hz, 1 H, H-5), 7.38 (d, J ) 8.3 Hz, 2 H, H-2′), 7.07
(s, 1 H, H-8), 7.05 (d, J ) 8.9 Hz, 1 H, H-6), 6.82 (d, J ) 8.3
Hz, 2 H, H-3′), 4.12 (t, J ) 7.0 Hz, 2 H, CH2O), 2.24 (t, J ) 7.1
Hz, 2 H, CH2CdO), 1.76 (quintet, J ) 7.3 Hz, 2 H), 1.53 (broad
quintet, 2 H), 1.43 (broad quintet, 2 H), 1.30-1.2 (m, 8 H);
HPLC tR 27.7 min (gradient), 35.7 min (44% acetonitrile,
isocratic).
7-O-(10-Ca r boxyd ecyl)-3-(4-h yd r oxyp h en yl)-4H-1-ben -
zop yr a n -4-on e (7-O-ω-Ca r boxyd ecyld a id zein , 6). To a
suspension of 1.500 g of 2 (5.90 mmol) and 35 mL of acetone
was added 8.8 mL of 2 N aq KOH (17.71 mmol). The resulting
mixture was stirred vigorously at room temperature until 2
was completely dissolved. To this solution, 6-bromoundecanoic
acid (3.132 g, 11.81 mmol) was added and the reaction mixture
was stirred under gentle reflux for 3 days. Acetone was
removed by flash evaporation. The remaining aqueous solution
was neutralized by the addition of 17 mL of 1 N HCl and the
resulting suspension was kept at 5-6 °C for 24 h. The
precipitates were collected on a fritted funnel, washed with
small portions of H2O and 2 × 25 mL of petroleum ether, and
dried over P2O5 to yield 3.45 g of crude product. The dry
product was then washed with 3 × 75 mL of hot CHCl3 on a
fritted funnel and filtrates were combined and evaporated to
dryness (2.174 g). The resulting solid was washed with several
portions of diethyl ether until bromoundecanoic acid was no
longer detectable (by NMR) in the wash. This procedure
yielded 1 g of 6: mp 68-71 °C; 1H NMR (DMSO-d6) δ 9.53
(broad s, 1 H, 4′-OH), 8.36 (s, 1 H, H-2), 8.01 (d, J ) 8.7 Hz,
1 H, H-5), 7.39 (d, J ) 8.1 Hz, 2 H, H-2′), 7.13 (s, 1 H, H-8),
7.06 (d, J ) 8.7 Hz, 1 H, H-6), 6.81 (d, J ) 8.1 Hz, 2 H, H-3′),
4.11 (t, J ) 7.0 Hz, 2 H, CH2O), 2.15 (t, J ) 7.3 Hz, 2 H, CH2Cd
O), 1.75 (quintet, J ) 7.3 Hz, 2 H), 1.48 (quintet, J ) 7.3 Hz,
2 H), 1.42 (quintet, J ) 7.3 Hz, 2 H), 1.32-1.23 (m, 10 H); 13
C
7-O-(6-Ca r boxyh exyl)-3-(4-h yd r oxyp h en yl)-4H-1-ben -
zop yr a n -4-on e (7-O-ω-Ca r boxyh exyld a id zein , 4). To a
suspension of 2.540 g (10.00 mmol) of 2 and 40 mL of acetone
was added 10 mL of 2 N aq KOH (20.00 mmol). The resulting
mixture was stirred vigorously at room temperature until 2
was completely dissolved. To this solution, 2.09 g of 7-bromo-
heptanoic acid (10.00 mmol) was added and the reaction
mixture was stirred under gentle reflux for 3 days. After 12 h
at 4 °C, the carboxylic acid potassium salt formed was collected
on a fritted funnel and washed with small portions of cold
acetone. The precipitate was resuspended in 100 mL of H2O
and the pH of the suspension was adjusted to 2 with 15% HCl.
The white fluffy precipitate collected on a fritted funnel was
dissolved in 100 mL of MeOH with heating. The resulting
solution was slowly cooled to room temperature and kept at 4
°C overnight. The white precipitate was collected on a fritted
funnel to give 775 mg of 4: mp 193-198 °C; 1H NMR (DMSO-
d6) δ 8.24 (s, 1 H, H-2), 7.93 (d, J ) 8.9 Hz, 1 H, H-5), 7.34 (d,
NMR (DMSO-d6) δ 174.70 (C-4, CO2H), 163.06 (C-7), 157.42
(C-8a), 157.23 (C-4′), 153.11 (C-2), 130.06 (C-2′, 6′), 126.91 (C-
5), 123.67 (C-3), 122.38 (C-1′), 117.48 (C-4a), 114.96 (C-3′, 5′,
6 overlap), 100.95 (C-8), 68.46 (C-O), 33.65, 28.91, 28.83, 28.71
(2 C), 28.54, 28.39, 25.41, 24.48; HPLC tR 29.7 min (gradient),
43.7 min (46% acetonitrile, isocratic).
P r epar ation of ω-Br om oalkyldaidzein 7-O-Eth er s fr om
2. To a suspension of 2 in acetone was added 2 N aqueous
KOH (1 equiv) and a dibromoalkane (10 equiv). The mixtures
were refluxed for 18-24 h. After the reaction had reached
completion, as indicated by TLC, the mixture was cooled to
room temperature and acetone was evaporated on a rotary
evaporator. The residue was diluted with H2O and the pH of
the solution was adjusted to 4 with 1 N HCl. An aqual volume
of petroleum ether was then added to precipitate the crude
products. The suspension was stirred with a spatula for 10
min. The precipitates were collected on a fritted funnel, washed

4178 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Rooke et al.
successively with H2O and petroleum ether (3 times each), and
dried over P2O5. The crude solid was washed with hot CHCl3
(as many times as TLC proved necessary), leaving the in-
soluble unreacted 2. The combined chloroform filtrates were
concentrated to an oil on a rotary evaporator. The desired
product was precipitated with petroleum ether, collected on a
fritted funnel, washed with small portions of CHCl3/petroleum
ether 2/1, and dried under vacuum over P2O5. The powders
were recrystallized from an appropriate solvent to give pure
desired products as off-white solids.
7-O-(3-Br om op r op yl)-3-(4-h yd r oxyp h en yl)-4H -1-b en -
zop yr a n -4-on e (7-O-ω-Br om op r op yld a id zein , 7). This com-
pound was prepared as described in the general procedure
from 2 g of 2 (7.87 mmol) and 15.898 g of 1,3-dibromopropane
(78.74 mmol) in 40 mL of acetone. The product was recrystal-
lized from 95% ethanol to yield 1.353 g of 7: mp 178-179 °C;
1H NMR (DMSO-d6) δ 9.58 (s, 1 H, OH), 8.37 (s, 1 H, H-2),
8.02 (d, J ) 8.8 Hz, 1 H, H-5), 7.39 (d, J ) 8.5 Hz, 2 H, H-2′),
7.18 (d, J ) 2.3 Hz, 1 H, H-8), 7.09 (dd, J ) 8.8, 2.3 Hz, 1 H,
H-6), 6.80 (d, J ) 8.5 Hz, 2 H, H-3′), 4.24 (t, J ) 6.1 Hz, 2 H,
CH2O), 3.68 (t, J ) 6.1 Hz, 2 H, CH2Br), 2.30 (quintet, J ) 6.1
Hz, 2 H, CH2); 13C NMR (DMSO-d6) δ 174.68 (C-4), 162.66 (C-
7), 157.35 (C-8a), 157.23 (C-4′), 153.15 (C-2), 130.05 (C-2′, 6′),
127.00 (C-5), 123.69 (C-3), 122.33 (C-1′), 117.71 (C-4a), 114.95
(C-3′, 5′), 114.89 (C-6), 101.13 (C-8), 66.26 (C-O), 31.55 (C-
Br), 30.97; MS m/ z 375 (M + H+), 377 (M + 2 + H+), 397 (M
+ Na+), 399 (M + 2 + Na+); HPLC tR 25.6 min (gradient), 30.3
min (40% acetonitrile, isocratic).
H2O (30 mL) and pH was adjusted to 7 with 1 N HCl. The
product was precipitated by stirring this aqueous solution with
50 mL of petroleum ether for a few minutes. The precipitate,
containing the product and unreacted 2, was collected on a
fritted funnel, washed successively with H2O (3 × 10 mL) and
petroleum ether (3 × 10 mL), and dried under vacuum over
P2O5. The desired product was collected by washing the
precipitate with hot chloroform (4 × 100 mL) and the combined
chloroform extracts were evaporated to a solid residue which
was recrystallized from 2-propanol to give 854 mg of pure 10
1
as a pale yellow solid: mp 178-179.5 °C; H NMR (DMSO-
d6) δ 9.52 (s, 1 H, OH), 8.35 (s, 1 H, H-2), 8.00 (d, J ) 8.8 Hz,
1 H, H-5), 7.39 (d, J ) 8.5 Hz, 2 H, H-2′), 7.13 (d, J ) 2.2 Hz,
1 H, H-8), 7.06 (dd, J ) 8.8, 2.2 Hz, 1 H, H-6), 6.80 (d, J ) 8.5
Hz, 2 H, H-3′), 4.75 (t, J ) 5.2 Hz, 1 H, CH), 4.17 (t, J ) 6.4
Hz, 2 H, CH2OAr), 4.01 (dd, J ) 11.2, 4.8 Hz, 2 H), 3.72 (dd,
J ) 11.2, 9.0 Hz, 2 H), 1.98 (td, J ) 6.4, 5.2 Hz, 2 H), 1.88 (m,
2 H); 13C NMR (DMSO-d6) δ 174.69 (C-4), 162.76 (C-7), 157.37
(C-8a), 157.23 (C-4′), 153.11 (C-2), 130.06 (C-2′, 6′), 126.96 (C-
5), 123.68 (C-3), 122.36 (C-1′), 117.62 (C-4a), 114.96 (C-3′, 5′),
114.86 (C-6), 101.03 (C-8), 98.59, 66.09, 64.17, 34.35, 25.35;
MS m/z 369 (M + H+), 367 (M - H+), 403 (M + Cl-); HPLC tR
21.5 min (gradient), 26.3 min (32% acetonitrile, isocratic).
7-O-Allyl-3-(4-h yd r oxyp h e n yl)-4H -1-b e n zop yr a n -4-
on e (7-O-Allyld a id zein , 11). To a suspension of 1.000 g of 2
(3.93 mmol) and 22 mL of acetone was added 2 mL of 11.2%
KOH (4.00 mmol) and 5.000 g of allyl bromide (3.93 mmol).
The resulting orange solution was refluxed for 18 h. The
reaction mixture was cooled to room temperature and solvent
was removed by rotary evaporation. The product was precipi-
tated by adding 25 mL of petroleum ether, 4 mL (8 mmol) of
1 N HCl and 25 mL of H2O. The precipitate formed upon
stirring was collected on a fritted funnel, washed with 2 × 10
mL of petroleum ether and 2 × 10 mL of H2O, and dried under
vacuum over P2O5. The solid, containing the desired product
and unreacted 2, was washed with hot chloroform and (2 ×
100 mL) and the combined chloroform filtrates were concen-
trated on a rotary evaporator and the desired product was
precipitated from the concentrate by adding three pipets full
of petroleum ether. The product was further purified by
recrystallization (95% ethanol) to give 441 mg of pure 11 as a
beige powder: mp 176-178 °C; 1H NMR (DMSO-d6) δ 9.57
(broad s, 1 H, OH), 8.35 (s, 1 H, H-2), 8.02 (d, J ) 8.8 Hz, 1 H,
H-5), 7.38 (d, J ) 8.5 Hz, 2 H, H-2′), 7.15 (s, 1 H, H-8), 7.09 (d,
J ) 8.8 Hz, 1 H, H-6), 6.80 (d, J ) 8.5 Hz, 2 H, H-3′), 6.06
(ddt, J ) 17.3, 10.5, 4.4 Hz, 1 H, CH)), 5.44 (d, J ) 17.3 Hz,
1 H, CH2)), 5.30 (d, J ) 10.5 Hz, 1 H, CH)), 4.72 (d, J ) 4.4
Hz, 2 H, CH2); 13C NMR (DMSO-d6) δ 174.69 (C-4), 162.47 (C-
7), 157.31 (C-8a), 157.25 (C-4′), 153.14 (C-2), 132.81, 130.07
(C-2′, 6′), 126.97 (C-5), 123.70 (C-3), 122.35 (C-1′), 118.24,
117.68 (C-4a), 115.03 (C-3′, 5′), 114.97 (C-6), 101.36 (C-8), 68.98
(C-O); MS m/z 295 (M + H+), 317 (M + Na+), 293 (M - H+);
HPLC tR 23.6 min (gradient), 43.6 min (32% acetonitrile,
isocratic).
7-O-(4-Br om obu tyl)-3-(4-h ydr oxyph en yl)-4H-1-ben zopy-
r a n -4-on e (7-O-ω-Br om obu tyld a id zein , 8). This compound
was synthesized as described in the general procedure from 2
g of 2 (7.87 mmol) and 8.501 g of 1,4-dibromobutane (39.35
mmol) in 40 mL of acetone. The product was recrystallized
from CHCl3/diethyl ether (3/1) to yield 1.310 g of 8: mp 170-
1
172 °C; H NMR (DMSO-d6) δ 9.53 (s, 1 H, OH), 8.37 (s, 1 H,
H-2), 8.02 (d, J ) 8.9 Hz, 1 H, H-5), 7.40 (d, J ) 8.5 Hz, 2 H,
H-2′), 7.15 (d, J ) 2.2 Hz, 1 H, H-8), 7.08 (dd, J ) 8.9, 2.2 Hz,
1 H, H-6), 6.81 (d, J ) 8.5 Hz, 2 H, H-3′), 4.17 (t, J ) 6.2 Hz,
2 H, CH2O), 3.63 (t, J ) 6.6 Hz, 2 H, CH2Br), 1.99 (quintet, J
) 6.8 Hz, 2 H), 1.89 (quintet, J ) 6.8 H, 2 H); 13C NMR
(DMSO-d6) δ 174.69 (C-4), 162.87 (C-7), 157.37 (C-8a), 157.23
(C-4′), 153.09 (C-2), 130.05 (C-2′, 6′), 126.92 (C-5), 123.67 (C-
3), 122.36 (C-1′), 117.57 (C-4a), 114.96 (C-3′, 5′, 6 overlap),
101.00 (C-8), 67.59 (C-O), 34.72, 28.95, 27.10; MS m/z 389
(M + H+), 391 (M + 2 + H+), 411 (M + Na+), 413 (M + 2 +
Na+), 427 (M + K+), 429 (M + 2 + K+), 387 (M - H+), 389 (M
+ 2 - H+), 423 (M + Cl-), 425 (M + 2 + Cl-); HPLC tR 27.4
min (gradient), 14.3 min (48% acetonitrile, isocratic).
7-O-(6-Br om oh exyl)-3-(4-h ydr oxyph en yl)-4H-1-ben zopy-
r a n -4-on e (7-O-ω-Br om oh exyld a id zein , 9). This compound
was prepared as described in the general procedure from 2.520
g of 2 (9.92 mmol) and 24.583 g of 1,6-dibromohexane (0.10075
mol) in 50 mL of acetone. The crude product was recrystallized
from chloroform/diethyl ether (2/1) to give 2.229 g of pure 9:
mp 140-142 °C; 1H NMR (DMSO-d6) δ 9.58 (broad s, 1 H, OH),
8.36 (s, 1 H, H-2), 8.01 (d, J ) 8.8 Hz, 1 H, H-5), 7.39 (d, J )
7.0 Hz, 2 H, H-2′), 7.13 (s, 1 H, H-8), 7.06 (d, J ) 8.8 Hz, 1 H,
H-6), 6.81 (d, J ) 7.0 Hz, 2 H, H-3′), 4.12 (t, J ) 6.5 Hz, 2 H,
CH2O), 3.54 (t, J ) 6.5 Hz, 2 H, CH2Br), 1.83 (m, 2 H, CH2),
1.76 (m, 2 H, CH2), 1.46 (m, 4 H, CH2); 13C NMR (DMSO-d6)
δ 174.67 (C-4), 163.01 (C-7), 157.38 (C-8a), 157.22 (C-4′), 153.08
(C-2), 130.04 (C-2′, 6′), 126.90 (C-5), 123.65 (C-3), 122.36 (C-
1′), 117.48 (C-4a), 114.94 (C-3′, 5′, 6 overlap), 100.93 (C-8),
68.33 (C-O), 35.07, 32.14, 28.22, 27.24, 24.57; MS m/z 417
(M + H+), 419 (M + H+ + 2); HPLC tR 29.1 min (gradient),
31.7 min (46% acetonitrile, isocratic).
7-O-[2-(1,3-Dioxa n yl)eth yl]-3-(4-h yd r oxyp h en yl)-4H-1-
ben zop yr a n -4-on e (7-O-[2-(1,3-Dioxa n yl)eth yl]d a id zein ,
10). To a solution containing 3.000 g of 2 (11.81 mmol), 6.5
mL of aqueous 2 N KOH (12.99 mmol) and 60 mL of acetone
was added 2.1 mL of 2-(2-bromoethyl)-1,3-dioxane (2.995 g,
15.35 mmol). The reaction mixture was stirred under gentle
reflux for 40 h and acetone was removed by flash evaporation.
The remaining aqueous solution was diluted with 30 mL of
7-O-(2,3-Dih yd r oxyp r op yl)-3-(4-h yd r oxyp h en yl)-4H-1-
ben zop yr a n -4-on e (7-O-(2,3-Dih yd r oxyp r op yl)d a id zein ,
12). To a solution of 647 mg of 11 (2.20 mmol) and 309 mg
methylmorpholine N-oxide (2.64 mmol) in 60 mL of acetone/
H2O (3/2) was added 1.4 mL OsO4 solution (2.5% in t-BuOH).
The resulting solution was stirred at room temperature for
48 h. Then 35 mL of a 3.6% Na2S2O4 solution was added to
the solution and the resulting black mixture was stirred at
room temperature for another 4 h. The mixture was filtered
through Celite and the filtrate was evaporated to dryness. The
residue was further dried over P2O5 and then washed with
small portions (5-10 mL) of H2O to give 486 mg of 12: mp
1
188-191 °C; H NMR (DMSO-d6) δ 9.52 (s, 1 H, 4′-OH), 8.35
(s, 1 H, H-2), 8.01 (d, J ) 8.9 Hz, 1 H, H-5), 7.39 (d, J ) 8.3
Hz, 2 H, H-2′), 7.13 (s, 1 H, H-8), 7.07 (d, J ) 8.9 Hz, 1 H,
H-6), 6.80 (d, J ) 8.3 Hz, 2 H, H-3′), 5.04 (d, J ) 5.0 Hz, 1 H,
CH-OH), 4.72 (t, J ) 5.5 Hz, 1 H, CH2-OH), 4.15 (m, 1 H, CH2-
OAr), 4.02 (m, 1 H, CH2OAr), 3.83 (m, 1 H, CH-OH), 3.46 (t,
J ) 5.5 Hz, 2 H, CH2-OH); 13C NMR (DMSO-d6) δ 174.72 (C-
4), 163.20 (C-7), 157.39 (C-8a), 157.24 (C-4′), 153.12 (C-2),

Potential Site of Action of Daidzin
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4179
(6) Keung, W. M.; Vallee, B. L. Therapeutic lessons from traditional
Oriental medicine to contemporary Occidental pharmacology.
EXS 1994, 71, 371-381.
(7) Keung, W. M.; Lazo, O.; Kunze, L.; Vallee, B. L. Daidzin
suppresses ethanol consumption by Syrian golden hamsters
without blocking acetaldehyde metabolism. Proc. Natl. Acad. Sci.
U.S.A. 1995, 92, 8990-8993.
(8) Klyosov, A. A.; Rashkovetsky, L. G.; Tahir, M. K.; Keung, W.
M. Possible role of liver cytosolic and mitochondrial aldehyde
dehydrogenases in acetaldehyde metabolism. Biochemistry 1996,
35, 4445-4456.
(9) Hald, J .; J acobsen, E. A drug densitizing the organism to ethyl
alcohol. Lancet 1948, 255, 1001-1004.
(10) Keung, W. M.; Vallee, B. L. Daidzin and its antidipsotropic
analogues inhibit serotonin and dopamine metabolism in isolated
mitochondria. Proc. Natl. Acad. Sci. U.S.A.1998, 95, 2198-2203.
(11) Wa¨ha¨la¨, K.; Valo, T.; Brunow, G.; Hase, T. A. Monoalkylation
of daidzein (7,4′-dihydroxyisoflavone): Synthesis of 7-O-(car-
boxybutyl)equal. Finn. Chem. Lett. 1989, 16, 79-83.
(12) Benedict, D. R.; Bianchi, T. A.; Cate, L. A. Synthesis of simple
unsymmetrical ethers from alcohols and alkyl halides or sul-
fates: The potassium hydroxide/dimethyl sulfoxide system.
Synthesis 1979, 428-429.
(13) J ohnstone, R. A. W.; Rose, M. E. A rapid, simple, and mild
procedure for alkylation of phenols, alcohols, amides and acids.
Tetrahedron 1979, 35, 2169-2173.
(14) Tank, A. W.; Weiner, H.; Thurman, J . A. Enzymology and
subcellular localization of aldehyde oxidation in rat liver. Bio-
chem. Pharmacol. 1981, 30, 3265-3275.
(15) Ambroziak, W.; Pietruszko, R. Human aldehyde dehydroge-
nase: Activity with aldehyde metabolites of monoamines, di-
amines, and polyamines. J . Biol. Chem. 1991, 266, 13011-
13018.
(16) Feldstein, A.; Wong, K.-K. Enzymatic conversion of serotonin
to 5-hydroxytryptophol. Life Sci. 1961, 4, 183-191.
(17) Feldstein, A. Effect of ethanol on neurohumoral amine metabo-
lism. In The Biology of Alcoholism; Kissin, B., Beleiter, H., Eds.;
Plenum: New York, 1971; pp 127-159.
(18) Arvola, A.; Forsander, O. Hamsters in experiments of free choice
between alcohol and water. Nature (London) 1961, 191, 819-
820.
(19) Li. T.-K.; Lumeng, L.; McBride, W. J .; Waller, M. B. Quantitative
correlation of ethanol elimination rates in vivo with liver alcohol
dehydrogenase activities in fed, fasted and food-restricted rats.
Drug Alcohol Depend. 1979, 4, 45-60.
(20) von Knorring, A.-L.; Bohman, M.; von Knorring, L.; Oreland, L.
Platelet MAO activity as a biological marker in subgroups of
alcoholism. Acta Psychiatr. Scand. 1985, 72, 51-58.
(21) Harada, S.; Agarwal, D. P.; Goedde, H. W.; Tagaki, S.; Ishikawa,
B. Possible protective role against alcoholism for aldehyde
dehydrogenase isozyme deficiency in J apan. Lancet 1982, 2, 827.
(22) Deitrich, R. A.; Erwin, V. G. Biogenic amine-aldehyde condensa-
tion products: Tetrahydroisoquinolines and tryptolines (â-
carbolines). Annu. Rev. Pharmacol. 1980, 20, 55-80.
(23) Nilsson, G. E.; Tottmar, O. Biogenic aldehydes in brain: On their
preparation and reactions with rat brain tissue. J . Neurochem.
1987, 48, 1566-1572.
(24) Myers, R. D. Tetrahydroisoquinolines in the brain: the basis of
an animal model of alcoholism. Alcoholism: Clin. Exp. Res. 2,
145-154.
(25) Alivisatos, S. G. A.; Tabakoff, B. Formation and metabolism of
“biogenic” aldehydes. In Chemical Modulation of Brain Function;
Sabelli, H., Ed.; Raven Press: New York, 1973; pp 41-66.
130.08 (C-2′, 6′), 126.93 (C-5), 123.70 (C-3), 122.40 (C-1′),
117.54 (C-4a), 115.04 (C-3′, 5′), 114.98 (C-6), 101.02 (C-8),
70.51, 69.76, 62.50; MS m/z 329 (M + H+), 351 (M + Na+),
367 (M + K+), 327 (M - H+), 363 (M + Cl-); HPLC tR 12.3
min (gradient), 13.4 min (16% acetonitrile, isocratic).
7-E t h oxy-3-(4-h yd r oxyp h e n yl)-4H -1-b e n zop yr a n -4-
on e (7-O-Eth yld a id zein , 13). Finely ground KOH (442 mg,
7.87 mmol) was supended in 10 mL of DMSO (dried over 4 Å
molecular sieves) for 10 min. Then 2 g of 2 (7.87 mmol) and
0.8 mL of ethyl iodide (1.596 g, 10.24 mmol) were added.
Reaction was allowed to proceed with stirring at room tem-
perature until completion as indicated by TLC (about 3.5 h).
The reaction mixture was poured into 40 mL of ice water. The
off-white precipitate was collected on a fritted funnel, washed
with 2 × 10 mL of H2O, and dried over P2O5. The dried product
was recrystallized from 95% EtOH to give 2.043 g of 13: mp
1
197-198.5 °C; H NMR (DMSO-d6) δ 9.58 (s, 1 H, OH), 8.34
(s, 1 H, H-2), 8.00 (d, J ) 8.9 Hz, 1 H, H-5), 7.38 (d, J ) 8.4
Hz, 2 H, H-2′), 7.10 (s, 1 H, H-8), 7.04 (d, J ) 8.9 Hz, 1 H,
H-6), 6.80 (d, J ) 8.4 Hz, 2 H, H-3′), 4.16 (quartet, J ) 6.4 Hz,
2 H, CH2O), 1.36 (t, J ) 6.4 Hz, 3 H, CH3); 13C NMR (DMSO-
d6) δ 174.69 (C-4), 162.89 (C-7), 157.40 (C-8a), 157.23 (C-4′),
153.07 (C-2), 130.06 (C-2′, 6′), 126.90 (C-5), 123.67 (C-3), 122.38
(C-1′), 117.48 (C-4a), 114.96 (C-3′, 5′), 114.90 (C-6), 100.89 (C-
8), 64.18 (C-O), 14.35; MS m/z 283 (M + H+), 305 (M + Na+),
281 (M - H+), 317 (M + Cl-); HPLC: tR 22.7 min (gradient),
33.7 min (32% acetonitrile, isocratic).
Abbr evia tion s: ADH, alcohol dehydrogenase; ALDH, al-
dehyde dehydrogenase; AR, aldehyde reductase; DA, dop-
amine; DOPAC, 3,4-dihydroxyphenylacetic acid; DOPAL, 3,4-
dihydroxyphenylacetaldehyde; 5-HIAL, 5-hydroxyindole-3-ac-
etaldehyde; 5-HIAA, 5-hydroxyindole-3-acetic acid; 5-HTOL,
5-hydroxytryptophol; 5-HT, 5-hydroxytryptamine (serotonin);
MAO, monoamine oxidase; AUC, area under the curve.
Ack n ow led gm en t. This work was supported by the
Endowment for Research in Human Biology, Inc. The
continuous support of Dr. Bert L. Vallee is greatly
appreciated.
Refer en ces
(1) Keung, W. M.; Vallee, B. L. Daidzin and daidzein suppress free-
choice ethanol intake by Syrian golden hamsters. Proc. Natl.
Acad. Sci. U.S.A. 1993, 90, 10008-10012.
(2) Heyman, G. M.; Keung, W. M.; Vallee, B. L. Daidzin decreases
ethanol consumption in rats. Alcohol. Clin. Exp. Res. 1996, 20,
1083-1087.
(3) Overstreet, D. H.; Lee, Y.-W.; Rezvani, A. H.; Pei, Y.-Hm.;
Criswell, H. E.; J anowsky, D. S. Suppression of alcohol intake
after administration of the Chinese herbal medicine, NPI-028,
and its derivatives. Alcohol. Clin. Exp. Res. 1996, 20, 221-227.
(4) Lin, R. C.; Guthrie, S.; Xie, C.-I.; et al. Isoflavonoid compounds
extracted from Pueraria lobata suppress alcohol preference in
a pharmacogenetic rat model for alcoholism. Alcohol. Clin. Exp.
Res. 1996, 20, 659-663.
(5) Keung, W. M.; Vallee, B. L. Daidzin: A potent, selective inhibitor
of human mitochondrial aldehyde dehydrogenase. Proc. Natl.
Acad. Sci. U.S.A. 1993, 90, 1247-1251.
J M990614I