Synthesis of potential antidipsotropic isoflavones: Inhibitors of the mitochondrial monoamine oxidase - Aldehyde dehydrogenase pathway

Article abstract of DOI:10.1021/jm0101390

Recently we have shown that daidzin, the major active principle of an ancient herbal treatment for quot;alcohol addictionquot; suppresses ethanol intake in alcohol-preferring laboratory animals. Further, we have identified the monoamine oxidase (MAO) - aldehyde dehydrogenase (ALDH-2) pathway of the mitochondria as the potential site of action of daidzin. Daidzin analogues that potently inhibit ALDH-2 but have no or little effect on MAO are most antidipsotropic, whereas those that also inhibit MAO exhibit little, if any, antidipsotropic activity. Therefore, in the design and synthesis of more potent antidipsotropic analogues, structural features important for the inhibition of both ALDH-2 and MAO must be taken into consideration. To gain further information on the structure-activity relationships at the inhibitor binding sites of ALDH-2 and MAO, we prepared 44 analogues of daidzin and determined their potencies for ALDH-2 and MAO inhibition. Results indicate that a sufficient set of criteria for a potent antidipsotropic analogue is an isoflavone with a free 4′-OH function and a straight-chain alkyl substituent at the 7 position that has a terminal polar function such as -OH, -COOH, or -NH₂. The preferable chain lengths for the 7-O-ω-hydroxy, 7-O-ω-carboxy, and 7-O-ω-amino subsitutents are 2 ≤ n ≤ 6, 5 ≤ n ≤ 10, and n ≥ 4, respectively. Analogues that meet these criteria have increased potency for ALDH-2 inhibition and/or decreased potency for MAO inhibition and therefore are likely to be potent antidipsotropic agents.

Full text of DOI:10.1021/jm0101390

3320
J . Med. Chem. 2001, 44, 3320-3328
Syn th esis of P oten tia l An tid ip sotr op ic Isofla von es: In h ibitor s of 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
Guang-Yao Gao, Dian-J un Li, and Wing Ming Keung*
Center for Biochemical and Biophysical Sciences and Medicine, Harvard Medical School, Boston, Massachusetts 02115
Received March 29, 2001
Recently we have shown that daidzin, the major active principle of an ancient herbal treatment
for “alcohol addiction”, suppresses ethanol intake in alcohol-preferring laboratory animals.
Further, we have identified the monoamine oxidase (MAO)-aldehyde dehydrogenase (ALDH-
2) pathway of the mitochondria as the potential site of action of daidzin. Daidzin analogues
that potently inhibit ALDH-2 but have no or little effect on MAO are most antidipsotropic,
whereas those that also inhibit MAO exhibit little, if any, antidipsotropic activity. Therefore,
in the design and synthesis of more potent antidipsotropic analogues, structural features
important for the inhibition of both ALDH-2 and MAO must be taken into consideration. To
gain further information on the structure-activity relationships at the inhibitor binding sites
of ALDH-2 and MAO, we prepared 44 analogues of daidzin and determined their potencies for
ALDH-2 and MAO inhibition. Results indicate that a sufficient set of criteria for a potent
antidipsotropic analogue is an isoflavone with a free 4′-OH function and a straight-chain alkyl
substituent at the 7 position that has a terminal polar function such as -OH, -COOH, or
-NH2. The preferable chain lengths for the 7-O-ω-hydroxy, 7-O-ω-carboxy, and 7-O-ω-amino
subsitutents are 2 e n e 6, 5 e n e 10, and n g 4, respectively. Analogues that meet these
criteria have increased potency for ALDH-2 inhibition and/or decreased potency for MAO
inhibition and therefore are likely to be potent antidipsotropic agents.
In tr od u ction
5-hydroxyindole-3-acetic acid (5-HIAA) and 3,4-dihy-
droxyphenylacetic acid (DOPAC), in isolated hamster
or rat liver mitochondria.11 This, together with the
finding that daidzin does not affect the rates of mito-
chondria-catalyzed oxidative deamination of these
monoamines, suggests that the ethanol intake suppres-
sive (antidipsotropic) activity of daidzin is not mediated
by the monoamines but rather by their reactive biogenic
aldehyde intermediates such as 5-hydroxyindole-3-ac-
etaldehyde (5-HIAL) and/or 3,4-dihydroxyphenylacetal-
dehyde (DOPAL), which accumulate in the presence of
daidzin.11 Correlation studies using structural ana-
logues of daidzin revealed a positive correlation between
the antidipsotropic activities of these analogues and
their abilities to increase 5-HIAL accumulation during
5-HT metabolism in isolated liver mitochondria.12 These
studies also showed that daidzin analogues that po-
tently inhibit ALDH-2 but have little or no effect on
monoamine oxidase (MAO) are most antidipsotropic,
whereas those that also potently inhibit MAO exhibit
little, if any, antidipsotropic activity. On the basis of
these results, we proposed that the mitochondrial
MAO-ALDH-2 pathway is the site of action of daidzin
and that a biogenic aldehyde derived from the action of
MAO may directly or indirectly mediate its antidipso-
tropic action (Figure 1). The level of biogenic aldehyde
in isolated mitochondria is determined by the relative
activities of MAO and ALDH-2. Therefore, in the design
and synthesis of more potent antidipsotropic analogues,
structural features important for the inhibition of both
enzymes needed to be considered. To gain more infor-
mation on the structure-activity relationships at the
For more than a millennium, herbalists practicing
traditional Chinese medicine have used extracts of the
root and flower of kudzu (Pueraria lobata) to treat
“alcohol addiction”. The efficacy of the root extract was
first demonstrated in alcohol-preferring golden ham-
sters under strictly controlled laboratory conditions.
Further, the isoflavone daidzin was isolated from the
root and identified as its major active principle.1,2 These
findings were subsequently confirmed by us and other
investigators in additional animal models commonly
used in alcohol research.3-6 These provide for the first
time a scientific basis for the traditional use of kudzu
in the treatment of alcohol addiction.
In vitro, daidzin is a potent and selective inhibitor of
mitochondrial aldehyde dehydrogenase (ALDH-2),7 the
major ALDH isozyme that catalyzes the detoxification
of ethanol-derived acetaldehyde.8 This finding has led
to the suggestion that daidzin may act as an alcohol-
sensitizing agent.1,2 However, later in vivo studies
showed that daidzin, at doses that significantly suppress
hamster alcohol intake, does not inhibit overall ethanol
and acetaldehyde metabolism in this animal species.9,10
On the basis of these results, we suggested that daidzin
does not act by alcohol sensitization but may act by
modulating the activity of an as-yet-undefined physi-
ological pathway catalyzed by ALDH-2.
Recently, we have shown that daidzin inhibits the
conversion of monoamines such as serotonin (5-HT) and
dopamine (DA) into their respective acid metabolites,
* To whom correspondence should be addressed. Phone: (617) 432-
4001. Fax: (617) 566-3137. E-mail: wingming_keung@hms.harvard.edu.
10.1021/jm0101390 CCC: $20.00 © 2001 American Chemical Society
Published on Web 08/30/2001

Structure/ Activity of Antidipsotropic Isoflavones
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 20 3321
information data are listed in Table 1. The purities of
all daidzin analogues prepared for this study are greater
than 99.6% as judged by data from elementary analyses.
Str u ctu r e-Activity Rela tion sh ip s (SARs): AL-
DH-2 In h ibition . 7-O-Su bstitu tion Is Cr u cia l to
ALDH-2 In h ibition . In previous studies,9,12 we have
shown that daidzein (7,4′-dihydroxyisoflavone) is a
relatively poor ALDH-2 inhibitor (IC50 ) 9 µM) whereas
its 7-O-substituted derivatives, including daidzin (IC50
) 0.04 µM), inhibit the enzyme very potently (10-1000
times more potent). This result suggests that a variety
of 7-O-substituted daidzein derivatives would be ap-
propriate targets for designing novel daidzin analogues
that could be more potent for ALDH-2 inhibition and
therapeutically useful for the treatment of alcohol abuse
and alcoholism. To gain further SAR information, we
have synthesized and tested the potencies for ALDH-2
inhibition of nine series of 7-O-substituted daidzeins,
namely, the 7-O-alkyl (3-6), 7-O-ω-carboxyalkyl (7-15),
7-O-ω-hydroxyalkyl (16-20), 7-O-ω-bromoalkyl (21-
23), ethyl 7-O-ω-carboxyalkyl (24-26), 7-O-ω-alkenyl
(27, 28), 7-O-R-carboxyalkyl (29-31), ethyl 7-O-R-car-
boxyalkyl (32-34), and 7-O-ω-aminoalkyl (35, 36) de-
rivatives of daidzein. The results, expressed in IC50
values, are listed in Table 2. To facilitate dissolution,
stock solutions of all analogues were prepared in DMSO/
EtOH (9/1, v/v), and the final concentrations of these
solvents in assay media were 0.09% and 0.01%, respec-
tively. The IC50 values determined in the presence of
these solvents are in general higher than those deter-
mined in their absence.12
F igu r e 1. 5-HT metabolism in hamster liver mitochondria:
a proposed site of action for daidzin.
inhibitor binding sites of ALDH-2 and MAO, we syn-
thesized a series of analogues of daidzin and determined
their potencies for ALDH-2 and MAO inhibition.
Resu lts a n d Discu ssion
Syn th esis, P u r ifica tion , a n d Str u ctu r a l Id en ti-
fica tion of An a logu es of Da id zin . The 7-O-mono- and
7,4′-O-disubstituted derivatives of daidzein were pre-
pared via Williamson synthesis, by reacting the 7-O-
mono- and 7,4′-O-diaroxide anions of daidzein with an
appropriate alkyl halide RX (X ) Br, Cl), respectively.
The pKa values of the 7- and 4′-hydroxyl groups of
daidzein are 7.6 and 11.2, respectively, different enough
to allow selective ionization of the 7-OH group under
appropriate conditions ([base]/[daidzein] e 1), thus
permitting regioselective functionalization.13 The 7,4′-
disubstituted derivatives could be obtained by reacting
daidzein with excessive amounts of base and RX. In this
study, both the 7-O-mono- and 7,4′-O-disubstituted
derivatives were obtained from one-pot reactions in
which 1 < [base]/[daidzein] < 2. Reactions were carried
out in acetone or DMF using KOH or K2CO3 as the
bases, respectively.14,15 The mono- and disubstituted
species formed in the reaction mixtures were separated
and purified by solvent extraction, column chromatog-
raphy on Sephadex LH-20 and/or silica gel, and recrys-
tallization from specified solvents and compositions (see
the Experimental Section for details).
Daidzein derivatives with a 7-O-triethyleneglycol (20)
or a medium-chain-length (C-5 to C-10) 7-O-ω-carboxy-
alkyl (9-13) function are the most potent ALDH-2
inhibitors (IC50 values <0.06 µM), followed by the 7-O-
ω-hydroxyalkyl (16, 17), 7-O-alkyl (3, 4), 7-O-glucosyl
(1), and 7-O-ω-aminohexyl (36) derivatives (0.06 µM <
IC50 e 0.12 µM) and the 7-O-ω-bromoalkyl (21-23), 7-O-
ω-alkenyl (27, 28), 7-O-ω-aminobutyl (35), and ethyl
7-O-R-carboxyethyl (32) derivatives (0.2 µM < IC50
e
0.4 µM). Daidzein and the long-chain 7-O-alkyl (5, 6),
7-O-carboxyalkyl (15), 7-O-hydroxyalkyl (18, 19), and
7-O-R-carboxyalkyl (29-31, 33, 34) derivatives of daid-
zein are the least potent with IC50 values ranging from
0.7 to .9 µM. Except for 6 and 15, which have very long
7-O-substituents, alkylation of the 7-OH function of
daidzein (2) improves potency for ALDH-2 inhibition.
The structures and chemical compositions were de-
1
termined by H NMR, 13C NMR, MS, and elementary
analysis. NMR signals were assigned on the basis of
standard spectra of daidzein.12 Alkylation of the 7- and
4′-OH functions of daidzein resulted in additional H-
1
and C-signals at high field (1.2-4.1 ppm for H NMR,
24-33 ppm for 13C NMR). The identities of various
substituents were further verified on the basis of signals
arising from their characteristic functional groups. For
instance, (i) an R- or ω-carboxyl function on the 7- and
4′-substituents of 7-15, 49, and their ethyl esters 24-
26, 32-34, and 43-45 was indicated by the presence
of a carbonyl signal in the low field of 13C NMR (168-
174 ppm), (ii) terminal hydroxyl groups of 16-20 and
38 were identified by signals contributed by the hy-
droxymethylene function, which appeared at around 60
ppm, (iii) 7- and 4′-alkyl substituents in 4-6 were
indicated by the signals of their terminal methyl groups,
which appeared at 13.9 ppm for 6 and 21.6 ppm for 4,
and (iv) terminal double bonds on the side chains of 28
and 42 were verified by signals appearing at 5.01 (m,
2H) and 5.83 (m, 2H) ppm (1H NMR), and at 138 ppm
(13C NMR) (see the Experimental Section for details).
Molecular weights, melting points, and structural
Analogue 3 has the smallest 7-O-substituent, and yet
it is as potent as daidzin (1) or other daidzin analogues
with long-chain 7-O-substituents. This suggests that an
increase in the size and/or length of the 7-O-substituent
per se does not improve analogue binding. The fact that
daidzein is a poor ALDH-2 inhibitor may stem from the
weak electron-donating property of its free 7-OH group.
Therefore, replacing 7-OH with stronger electron-donat-
ing groups such as the alkoxyls should enhance enzyme
binding. Indeed, all 7-O-substituted daidzeins synthe-
sized in this study have increased electron density in
the A ring as suggested by the deshielding effect
observed in their NMR spectra (signals of H-5, H-6, and
H-8 were shifted to lower field in 7-O-substituted
derivatives), and all but 6 and 15 are more potent
ALDH-2 inhibitors than daidzein. Alternatively, binding
of daidzein to its binding site on ALDH-2 may bring its

3322 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 20
Ta ble 1. Physical Constants of the Compounds Synthesized
Gao et al.
Ta ble 2. Inhibition of Hamster Liver MAO and ALDH-2
Activities by Daidzin and Its Structural Analoguesa
IC50, µM IC50, µM
IC50, µM
ALDH-
ALDH-
ALDH-
compd
2
MAO compd
2
MAO compd
2
MAO
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
0.08
9
0.08
0.08
1.5
.9
1.2
0.1
0.06
0.05
0.04
0.05
0.04
0.13
ni
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
0.04
0.26
0.27
0.3
0.13
0.24
0.28
0.18
0.15
2.5
1
2
0.32
4.5
7.2
ni
0.15
4
2
ni
>9
5.3
0.45
5
ni
ni
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
ni
ni
ni
ni
ni
ni
ni
>6
ni
ni
3.5
7
ni
ni
ni
ni
ni
ni
ni
ni
0.03
0.12
7
4.5
13
0.5
ni
ni
ni
ni
14
0.3
0.2
ni
ni
.9
5
4
2.1
6.5
10
13
ni
ni
.9
.9
.9
>9
.9
ni
ni
1.5
1.5
0.15
ni
>9
ni
0.07
0.12
0.7
1.5
>9
>9
ni
0.25
0.1
.9
ni
ni
1.7
ni
a
MAO and ALDH-2 activity were assayed as described in the
Experimental Section, using 10 µM 5-HT and 0.3 mM formalde-
hyde as the stubstrates, respectively. IC50 values were estimated
graphically with inhibition data determined at more than six
inhibitor concentrations. ni ) no inhibition up to 10 µM.
free 7-OH group in close contact with a nonpolar region
in the inhibitor binding pocket of the enzyme. Therefore,
replacing the 7-OH of daidzein with a less polar alkoxyl
function should lead to more potent inhibition. Further,
this nonpolar region in the enzyme binding pocket must
be relatively short and narrow because shifting the OH
function two carbon atoms away from the 7 position
increases potency of inhibition dramatically (from IC50
) 9 µM for daidzein to IC50 ) 0.07 µM for 16), whereas
replacing the terminal hydroxymethylene function in 16
with a carboxyl group in 7 dramatically decreases its
ability to inhibit ALDH-2 (from 0.07 to .9 µM).
Potencies for ALDH-2 inhibition of the 7-O-alkyl
series are inversely proportional to the alkyl chain
lengths. The IC50 value for ALDH-2 inhibition increases
from 0.08 to 1.5 and .9 µM as the chain length
increases from C-2 (3, 4) to C-6 (5) and C-12 (6),
respectively (Table 2). This trend also pertains to the
series of 7-O-ω-hydroxyalkyl, ethyl 7-O-R-carboxyalkyl,
and 7-O-ω-carboxyalkyl derivatives. However, potencies
for ALDH-2 inhibition of the 7-O-alkyl derivatives with
terminal hydroxyl and carboxyl functions do not drop
off significantly until chain lengths are greater than 6
(18, 19) and 11 (15) carbon atoms, respectively. Further,
addition of either a bromo (23) or an amino (36) end
group also improves ALDH-2 inhibition. These, together
with the fact that derivatives with 7-O-glucosyl (1),
triethylene glycol (20), or O-ω-carboxyalkyl (9-13)
functions are all potent inhibitors of ALDH-2, suggest
that the subsite of the binding pocket of ALDH-2 that
accommodates the 7-O-substituent of a daidzin analogue
begins with a nonpolar bottleneck region, which is about
2-3 carbon atoms long measuring from the 7 position
of the chromone structure, followed by a long, spacious
and relatively hydrophilic pocket. It accommodates
poorly nonpolar alkyls having chain lengths greater
than three carbon atoms but receives very well long 7-O-
substituents with hydrophilic end groups.
a
Commercial products, over 99.5% pure as judged by HPLC
analysis. b Compounds synthesized in a previous study (see ref 12).
c Compounds synthesized and provided by Dr. X. Shen (current
address: American Cyanamid Co., P.O. Box 400, Princeton, NJ
08543-0400).
Th e Isofla von e Sk eleton Is Im p or ta n t for AL-
DH-2 In h ibition . Flavonoid is one of the most abun-

Structure/ Activity of Antidipsotropic Isoflavones
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 20 3323
F r ee 4′-OH Is Im p or ta n t for ALDH-2 In h ibition .
To study the effect of 4′-O-substitution on the potency
for ALDH-2 inhibition, we have synthesized nine 7,4′-
O-disubstituted derivatives of daidzein, and tested and
compared their ALDH-2 inhibitory activities with those
of their 7-O-monosubstituted analogues. As shown in
Table 2, none of the disubstituted derivatives inhibit
ALDH-2. These results are consistent with our earlier
finding that replacement of the 4′-OH of daidzin with a
methoxy group (as in 46) increases the inhibition
constant by more than 100-fold.7 On the basis of these
results, we conclude that blocking the free 4′-OH of an
isoflavone abolishes its ability to inhibit ALDH-2.
Str u ctu r e-Activity Rela tion sh ip s: MAO In h ibi-
t ion . 7-O-Su b st it u t ion a n d MAO In h ib it ion . The
structural requirements of the 7-O-substituent for MAO
inhibition appear to be similar to those for ALDH-2
except that, in MAO, the 7 position of a bound inhibitor
is located in a more spatially restricted area. MAO, like
ALDH-2, is relatively resistant to daidzein inhibition
(Table 2). The free 7-OH group again is the culprit for
poor inhibition because its replacement with a short-
chain hydrophobic alkyl (3, 4), bromoalkyl (21), or allyl
(27) moiety greatly enhances the potency for MAO
inhibition. Long-chain 7-O-alkyl derivatives of daidzein
(5, 6), however, are not inhibitory at all. This trend also
pertains to the 7-O-ω-hydroxyalkyl (16-19), ethyl 7-O-
R-carboxyalkyl (29-31), and 7-O-ω-carboxyalkyl (10-
15) series. Replacement of the alkyl side chains of the
noninhibitory analogues (5, 6) with ω-carboxyalkyls of
equal chain lengths (9, 10) improves the potency for
MAO inhibition. Nevertheless, none of the 7-O-ω-
carboxyalkyl derivatives (7-15), nor for that matter the
7-O-ω-hydroxyalkyl, 7-O-ω-bromoalkyl, and 7-O-ω-ami-
noalkylderivatives of daidzein with chain lengths greater
than three carbon atoms, potently inhibit MAO. It
appears that, in MAO, the 7-O-substituent of a bound
inhibitor is located in a spatially restricted area. This
area is likely to be hydrophobic because short-chain
carboxyalkyl (7) and hydroxyalkyl (16) derivatives are
poor MAO inhibitors.
F igu r e 2. Structures of some isoflavones and chemically or
biologically related compounds.
dant classes of phenolic compounds found in the plant
kingdom that share a 1,3-diphenylpropane (2-pheny-
lated) skeleton. They are structurally and biogenetically
related to the isoflavonoids that have a 1,2-diphenyl-
propane (3-phenylated) skeleton. Despite their struc-
tural similarities, none of the flavones tested so far
suppress alcohol intake in alcohol-preferring Syrian
golden hamsters.1 In an early study, we surveyed 29
commercially available flavones and isoflavones and
showed that all flavones studied exhibit low, if any,
ALDH-2 inhibitory activity.7 To further evaluate the
importance of the 1,2-diphenylpropane skeleton for
ALDH-2 inhibition, we synthesized 7-O-ω-carboxypent-
yl-4′-hydroxyflavone (49) (Figure 2), a flavone analogue
of 7-O-ω-carboxypentyl-4′-hydroxyisoflavone (9), and
tested and compared their potencies for ALDH-2 inhibi-
tion. Results show that the isoflavone (9) is at least 50
times more potent than its flavone analogue (49) (IC50
) 0.06 and 3.5 µM, respectively). This, together with
the fact that none of the flavones tested (47, 48) inhibit
ALDH-2, suggests that the isoflavone skeleton is im-
portant for ALDH-2 inhibition (Table 2).
Potencies for ALDH-2 inhibition of five coumarins
(52-56), one chalcone (50), and one flavanone (51) (see
Figure 2 for the structures) have been studied before7
and were reexamined under the same assay conditions
described in this study. Results are consistent with the
previous finding: only 4-phenyl (52, 53) and 3-phenyl
(54) substituted coumarins inhibit ALDH-2 (Table 2).
In this context, it is of interest to note that, among the
coumarin compounds tested, 54 is the most potent and
has a skeletal ring structure similar to that of an
isoflavone. Nevertheless, a direct comparison between
this particular coumarin (54) and the isoflavones could
be misleading because its saturated cyclohexyl ring is
both spatially and electronically different from that of
the aromatic A ring of an isoflavone so that its mode of
binding may differ significantly from the SAR derived
from the isoflavones.
4′-O-Su bstitu tion a n d MAO In h ibition . The effects
of 4′-O-substitution on the potency for MAO inhibition
were evaluated by testing and comparing the MAO
inhibitory activities of a series of 7-O-monosubstituted
derivatives (4-7, 18, 25, 26, 28, and 32) and their 7,4′-
O-disubstituted counterparts (39-41, 37, 38, 43, 44, 42,
and 45). It appears that the free 4′-OH groups of these
analogues are important for MAO inhibition because
replacement of the 4′-OH of an MAO inhibitory analogue
(4, 26, or 28) with their respective 7-O-substituents (39,
44, and 42) invariably destroys their abilities to inhibit
MAO (Table 2).
F la von es Ra th er Th a n Isofla von es Ar e Better
MAO In h ibitor s. Unlike ALDH-2, MAO appears to be
more sensitive to flavone than isoflavone inhibition.
Daidzein, a 7,4′-O-dihydroxyisoflavone, is a poor MAO
inhibitor (IC50 ) 14 µM), whereas 47, its flavonoid
counterpart, inhibits MAO very potently (IC50 ) 0.03
µM). Like its isoflavone counterpart, substituting 7-OH
with a six-carbon carboxyalkyl function (49) is detri-
mental to its potency for MAO inhibition (IC50 ) 7 µM).
7-Hydroxyflavone (48) is also a potent inhibitor of MAO
(IC50 ) 0.12 µM). Therefore, unlike the isoflavones, the

3324 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 20
Gao et al.
7-O-ω-carboxydecyldaidzein (13), 7-O-ω-bromopropyldaidzein
(21), 7-O-ω-bromobutyldaidzein (22), 7-O-ω-bromohexyldaid-
zein (23), and 7-O-allyldaidzein (27) were synthesized as
described in a previous study.12 The remaining compounds (4-
8, 11, 14-20, 24-26, 28-34, and 37-45) were synthesized
as described below. Serotonin (5-HT) was purchased from
Research Biochemical International (Natick, MA), and its
metabolic intermediate 5-HIAL was produced in this labora-
tory by MAO-catalyzed oxidative deamination of 5-HT using
rat liver mitochondrial membrane as a source of MAO.16 All
other reagents used were of the best grade available.
4′-OH function of a flavonoid compound might not be
critical for MAO inhibition. Coumarin compounds with
proper substituents could be potent MAO inhibitors. For
instance, 7-hydroxy-4-phenylcoumarin (52) is a fairly
potent inhibitor of MAO (IC50 ) 0.5 µM), whereas
7-methyl-4-phenylcoumarin (53) does not inhibit MAO
at all. This suggests that the mode of binding of the
flavones to the enzyme is significantly different from
that of the coumarins.
MAO a n d ALDH-2 Assa ys. The membrane and super-
natant fractions of the lysate of a gradient-purified hamster
liver mitochondria preparation were used as sources of MAO
and ALDH-2, respectively. ALDH-2 and MAO activity assays
were performed according to the procedures reported before.12
To facilitate dissolution, stock solutions of all test compounds
were prepared in DMSO/ethanol (90/10, v/v), and their final
concentrations in all assay media, including controls, were
0.09% and 0.01%, respectively.
Syn th esis. Gen er a l Meth od s. 1H and 13C NMR spectra
were recorded on a Bruker AMX 500 BQ spectrometer at 500
MHz and a Bruker AM-500 spectrometer at 126 MHz (NuMega
Resonace Laboratories Inc., San Diego, CA), respectively, using
DMSO as solvent and as internal standard (2.50 and 39.51
ppm for 1H and 13C, respectively) unless otherwise indicated.
Mass spectra were measured on a Perkin-Elmer PE-SCIEX
API 100 mass spectrometer by infusion. Samples were ionized
by electrospray, and spectra were recorded in positive mode.
Melting points were determined with a Hoover capillary
melting point apparatus. Elementary analyses were performed
by NuMega Resonance Laboratories Inc. Crude synthetic
products were purified by one or a combination of the following
methods: chromatography on a Sephadex LH-20 (Fluka, 25-
100 µm) or a silica gel 60 (70-230 mesh, EM Science) column
and recrystallization from acetone or chloroform/methanol of
various proportions. Analytical thin-layer chromatography
(TLC) was performed on Kiselgel 60F254 plates (Merck KgaA,
Darmstadt, Germany).
7-O-Isop r op yld a id zein (4) a n d 7,4′-Di-O-isop r op yl-
d a id zein (39). To a suspension of 5.1 g of 2 (20.08 mmol) and
60 mL of acetone was added 15 mL of 2 N KOH (30 mmol)
with vigorous stirring until 2 was completely dissolved. To this
solution was added 6 mL of 2-bromopropane (63.9 mmol), and
the reaction mixture was refluxed with gentle stirring for 72
h. Solvent was removed on a rotary evaporator, and the residue
was suspended in water and extracted with ethyl acetate.
Ethyl acetate was removed, and the residue was fractionated
on a Sephadex LH-20 column equilibrated in chloroform/
methanol (7/3, v/v). Fractions that contained 4 or 39 were
pooled separately (monitored by TLC), dried, and recrystallized
from chloroform/methanol (7/3) to give 1.8 g of 4 and 1.31 g of
39. Analyses (4): colorless crystalline plates; yield 30.3% (4/2,
mol/mol); mp169-171 °C; 1H NMR (DMSO-d6) δ 1.27 (m, 3H,
-CH3), 1.31 (m, 3H, -CH3), 4.85 (m, 1H, -OCH), 6.81 (dd,
2H, J ) 8.15, 1.7 Hz, H-3′, H-5′), 7.02 (dd, 1H, J ) 8.86, 2.1
Hz, H-6), 7.11 (d, 1H, J ) 2.5 Hz, H-8), 7.39 (dd, 2H, J ) 8.15,
1.68 Hz, H-2′, H-6′), 8.0 (d, 1H, J ) 8.86 Hz, H-5), 8.34 (s, 1H,
H-2), 9.55 (s, 1H, OH-4′); 13C NMR δ 21.5 (-CH3), 21.8 (-CH3),
70.4 (-OCH<), 101.7 (C-8), 114.9 (C-6), 115.1 (C-3′, C-5′), 117.3
(C-10), 122.4 (C-1′), 123.6 (C-3), 127.0 (C-5), 130.0 (C-2′, C-6′),
153.0 (C-2), 157.1 (C-9), 157.2 (C-4′), 161.9 (C-7), 174.6 (C-4);
MS m/z 297.5 (M + H)+. Anal. Calcd (C18H16O4): C, 72.96; H,
5.44. Found: C, 72.96; H, 5.25. Analyses (39): colorless
crystalline needles; yield 19.3%; mp134-136 °C; 1H NMR and
13C NMR similar to those of 4 and consistent with the
interpretation that 39 is a 7,4′-disubstituted analogue of 4;
MS m/z 339.5 (M + H)+. Anal. Calcd (C21H22O4): C, 74.54; H,
6.55. Found: C, 74.70; H, 6.68.
Su m m a r y
We have prepared a series of daidzein (2) derivatives
in which the 7-OH group of the parent molecule was
replaced with 7-O-alkyl (3-6), 7-O-ω-carboxyalkyl (7-
15), 7-O-ω-hydroxyalkyl (16-19), 7-O-ω-bromoalkyl
(21-23), ethyl 7-O-ω-carboxyalkyl (24-26), 7-O-ω-allyl
(27, 28), 7-O-R-carboxyalkyl (29-31), ethyl 7-O-R-car-
boxyalkyl (32-34), and 7-O-ω-aminoalkyl (35, 36) sub-
stituents. Their potencies for ALDH-2 and MAO inhi-
bition were determined and compared among each other
and with those of the 7-O-glucosyl derivative daidzin
(1). Results reveal that alkylation of the 7-OH of 2 is
essential for potent inhibition of both ALDH-2 and
MAO. However, long-chain 7-O-substituents have no or
decreased inhibitory activities toward both enzymes.
The optimal chain length varies for both ALDH-2 and
MAO inhibition, depending on the chemical nature of
the end group and the enzyme. It appears that a
straight-chain alkyl substituent with a polar end group
at the 7 position of an isoflavone is the key to a potent
and yet selective inhibitor of ALDH-2 and hence a
potent antidipsotropic agent.
We have also begun to explore the effect of 4′-
substitution and the isoflavone skeleton on ALDH-2 and
MAO inhibition. Preliminary results suggest that a free
-OH at the 4′ position may be crucial for both ALDH-2
and MAO inhibition and that isoflavones, as opposed
to flavones, inhibit ALDH-2 more potently than MAO
and hence are more likely to be antidipsotropic.
Three-dimensional structures of ALDH-2- and MAO-
isoflavonoid inhibitor complexes are still unknown at
this time, which hence precludes the discovery of
potential lead compounds based solely on rational
design. However, interactive docking using selected
analogues and the X-ray structure of ALDH-217,18 should
provide useful information on the mode of enzyme-
inhibitor binding, and this approach is currently being
pursued.
Exp er im en ta l Section
General chemicals were purchased from either Aldrich
Chemical Co. (Milwaukee, WI) or Lancaster Synthesis Inc.
(Windham, NH). All organic solvents used were of HPLC grade
and were supplied by J .P. Baker (Phillipsburg, NJ ) or Fisher
Scientific Co. (Pittsburgh, PA). Daidzin (1) was purchased from
LC Laboratories (Woburn, MA). Daidzein (2) was synthesized
by Tyger Scientific Inc. (Princeton, NJ ). Ononin (46), 7,4′-
dihydroxyflavone (47), 7-hydroxyflavone (48), 4′-hydroxychal-
cone (50), and 7-hydroxyflavanone (51) were purchased from
Indofine Chemical Co. (Somerville, NJ ). 7-Hydroxy-4-phenyl-
coumarin (52), 7-methyl-4-phenyl-3,4-dihydrocoumarin (53),
7-chloro-4A,5,6,7,8A-hexahydro-4-methyl-3-phenylcoumarin (54),
warfarin (55), and 6,7-dimethoxycoumarin (56) were pur-
chased from Aldrich Chemical Co. (Milwaukee, WI). 7-O-
Ethyldaidzein (3), 7-O-ω-carboxypentyldaidzein (9), 7-O-ω-
carboxyhexyldaidzein (10), 7-O-ω-carboxynonyldaidzein (12),
7-O-Hexyldaidzein (5) an d 7,4′-Di-O-h exyldaidzein (40).
The monohexyl (5) and dihexyl (40) daidzeins were synthesized
and purified by the same methods described for 4 and 39 using
6-bromohexane (3.5 mL) as the alkylating agent. Analyses
(5): colorless crystalline needles; yield 26.9%; mp145-148 °C;

Structure/ Activity of Antidipsotropic Isoflavones
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 20 3325
1H NMR (DMSO-d6) δ 0.874 (t, 3H, -CH3), 1.30 (m, 4H, -CH2-
CH2-), 1.45 (m, 2H, -CH2-), 1.75 (m, 2H, -CH2-), 4.09 (m,
-OCH2-), 6.81 (dd, 2H, J ) 8.7, 1.77 Hz, H-3′, H-5′), 7.03 (dd,
1H, J ) 8.86, 1.9 Hz, H-6), 7.10 (d, 1H, J ) 1.9 Hz, H-8), 7.40
(dd, 2H, J ) 8.7, 1.77 Hz, H-2′, H-6′), 8.0 (d, 1H, J ) 8.86 Hz,
H-5), 8.34 (s, 1H, H-2), 9.53 (s, 1H, OH-4′); 13C NMR δ 13.9
(-CH3), 22.0 (-CH2-), 25.1 (-CH2-), 28.4 (-CH2-), 30.9
(-CH2-), 68.4 (-OCH2-), 100.9 (C-8), 114.9 (C-6), 114.9 (C-3′,
C-5′), 117.5 (C-10), 122.4 (C-1′), 123.7 (C-3), 126.9 (C-5), 130.0
(C-2′, C-6′), 153.0 (C-2), 157.2 (C-9), 157.3 (C-4′), 163.0 (C-7),
174.7 (C-4); MS m/z 339.3 (M + H)+. Anal. Calcd (C21H22O4):
C, 74.54; H, 6.55. Found: C, 74.60; H, 6.82. Analyses (40):
colorless crystalline plates; yield 4.2%; mp 119-121 °C; MS
m/z 425.5 (M + H)+; 1H NMR and 13C NMR spectra similar to
those of 5 and consistent with the interpretation that 40 is a
7,4′-disubstituted analogue of 5. Anal. Calcd (C27H34O4): C,
76.74; H, 8.11. Found: C, 76.30; H, 7.92.
(M + H)+, 369.3 (M - H)+, 393.2 (M + Na)+. Anal. Calcd
(C19H14O8): C, 61.63; H, 3.81. Found: C, 61.46; H, 3.71.
7-O-ω-Ca r boxybu tyld a id zein (8), Its Eth yl Ester (25),
a n d 7,4′-Di-O-ω-ca r boxybu tyld a id zein (43). Compounds
25, 43, and 8 were synthesized and purified according to the
procedures described for 7, 24, and 37, respectively, using ethyl
5-bromovalerate as the alkylating agent. Analyses (25): color-
1
less crystalline plates; yield 8.8%; mp 146-148 °C; H NMR
(DMSO-d6) δ 1.17 (t, 3H, -CH3), 1.69 (m, 2H, -CH2-), 1.77
(m, 2H, -CH2-), 2.38 (t, 2H, -CH2-), 4.05 (m, 2H, -CH2O-),
4.12 (m, 2H, -CH2O-), 6.81 (dd, 2H, J ) 8.67, 1.5 Hz, H-3′,
H-5′), 7.04 (dd, 1H, J ) 8.89, 2.1 Hz, H-6), 7.11 (d, 1H, J ) 2.1
Hz, H-8), 7.39 (dd, 2H, J ) 8.7, 1.5 Hz, H-2′, H-6′), 8.0(d, 1H,
J ) 8.89 Hz, H-5), 8.35 (s, 1H, H-2), 9.55 (s,1H, HO-4′); 13C
NMR δ 14.1 (-CH3), 21.1 (-CH2-), 27.7 (-CH2-), 59.7 (-CH2O-
), 68.1 (-CH2O-), 100.9 (C-8), 114.9 (C-6), 114.9 (C-3′, C-5′),
117.5 (C-10), 122.4 (C-1′), 123.7 (C-3), 126.9 (C-5), 130.1 (C-
2′, C-6′), 153.1 (C-2), 157.3 (C-9), 157.4 (C-4′), 162.9 (C-7), 172.7
(OdCO-), 174.7 (C-4); MS m/z 383.3 (M + 1)+. Anal. Calcd
(C22H22O6): C, 69.09; H, 5.80. Found: C, 68.60; H, 5.84.
Analyses (43): colorless crystalline plates; yield 30%; mp 96-
7-O-Dod ecyld a id zein (6) a n d 7,4′-Di-O-d od ecyl-
d a id zein (41). The monododecyl (6) and didodecyl (41) deriva-
tives of daidzein were also synthesized and purified according
to the methods described for 4 and 39 using 1-bromododecane
as the alkylating agent. Analyses (6): colorless crystalline
plates; yield 42%; mp 131-133 °C; 1H NMR and 13C NMR
spectra virtually identical to those of 5 except signals contrib-
uted by the alkyl group on the seventh position; MS m/z 423.5
(M + H)+. Anal. Calcd (C27H34O4): C, 76.74; H, 8.11. Found:
C, 77.11; H, 8.10. Analyses (41): colorless crystalline plates;
yield 30%; mp 140-141 °C; 1H NMR and 13C NMR spectra
similar to those of 6 and consistent with the interpretation
that it is the 7,4′-disubstituted analogue of 6; MS m/z 591.8
(M + H)+. Anal. Calcd (C39H58O4): C, 79.28; H, 9.89. Found:
C, 79.18; H, 9.91.
1
98 °C; H NMR and 13C NMR spectra similar to those of 25
and consistent with the interpretation that 43 is a 7,4′-
disubstituted analogue of 25; MS m/z 511 (M + H)+. Anal.
Calcd (C29H34O8): C, 68.22; H, 6.71. Found: C, 68.11; H, 6.35.
Analyses (8): white amorphous powder; yield 20.6%; mp 232-
235 °C; 1H NMR and 13C NMR spectra with the characteristic
features of those of 25 and consistent with the interpretation
that 8 is the free acid of 25; MS m/z 355.2 (M + H)+. Anal.
Calcd (C20H18O6): C, 67.79; H, 5.12. Found: C, 66.09; H, 4.67.
7-O-ω-Ca r boxyh ep tyld a id zein (11) was synthesized ac-
cording to the procedure described before.1 In this preparation,
5.1 g of 2 (20.02 mmol), 20 mL of 2 N KOH (40 mmol), and 4.4
g of 8-bromo-1-octanoic acid (19.72 mmol) were refluxed in 60
mL of acetone for 72 h. The reaction mixture was refrigerated
for 12 h, and the precipitates were collected on a fritted funnel
and recrystallized from acetone to give 100 mg of 11. Analyses
(11): white amorphous powder; yield 1.6%; mp192-193 °C;
1H NMR (DMSO-d6) δ 1.31 [m, 4H, -(CH2)2-], 1.41 (m, 2H,
-CH2-), 1.51 (m, 2H, -CH2-), 1.74 (m, 2H, -CH2-), 2.20 (t,
2H, -CH2-), 4.10 (t, -CH2O-), 6.80 (dd, 2H, J ) 8.2, 1.7 Hz,
H-3′, H-5′), 7.05 (dd, 1H, J ) 8.85, 1.7 Hz, H-6), 7.12 (d, 1H,
J ) 1.7 Hz, H-8), 7.39 (dd, 2H, J ) 8.2, 1.7 Hz, H-2′, H-6′), 8.0
(d, 1H, J ) 8.85 Hz, H-5), 8.35 (s, 1H, 2-H), 9.55 (s, 1H, OH-
4′), 12.0 (s, 1H, -COOH); 13C NMR δ 24.4 (-CH2-), 25.3
(-CH2-), 28.3 (-CH2-), 28.4 (-CH2-), 28.4 (-CH2-), 33.6
(-CH2-), 68.4 (-OCH2-), 100.9 (C-8), 114.9 (C-6), 114.9 (C-3′,
C-5′), 117.5 (C-10), 122.4 (C-1′), 123.7 (C-3), 126.9 (C-5), 130.0
(C-2′, C-6′), 153.1 (C-2), 157.2 (C-9), 157.4 (C-4′), 163.0 (C-7),
174.5 (C-4, -COOH); MS m/z 397 (M + H)+. Anal. Calcd
(C23H24O6): C, 69.68; H, 6.10. Found: C, 69.52; H, 5.99.
7-O-ω-Ca r boxyu n d ecyld a id zein (14) was synthesized
according to the procedure described for 11 using 12-bromo-
1-dodecanoic acid (5.83 g, 20.87 mmol) as the alkylating agent.
The precipitates collected after refrigeration and purified on
a Sephadex LH-20 column in chloroform/methanol (7/3). The
pooled fractions were dried and recrystallized in chloroform/
methanol (6/4) to give 500 mg of 14. Analyses (14): white
amorphous powder; yield 5.3%; mp 238-240 °C; 1H NMR
(DMSO-d6) δ 1.22-1.24 [m, 10H, (-CH2-)5], 1.40 (m, 2H,
-CH2-), 1.50 (m, 2H, -CH2-), 1.75 (m, 2H, -CH2-), 2.17 (t,
2H, -CH2-), 3.40 (m, 2H, -CH2-), 4.09 (t, -CH2O-), 6.80 (dd,
2H, J ) 8.4, 1.7 Hz, H-3′, H-5′), 7.04 (dd, 1H, J ) 8.9, 1.9 Hz,
H-6), 7.10 (d, 1H, J ) 1.9 Hz, H-8), 7.40 (dd, 2H, J ) 8.4, 1.7
Hz, H-2′, H-6′), 8.0(d, 1H, J ) 8.9 Hz, H-5), 8.34 (s, 1H, H-2),
9.5 (s, 1H, OH-4′), 12.0 (br, 1H, -COOH); 13C NMR δ 24.5
(-CH2-), 25.4 (-CH2-), 28.4 (-CH2-), 28.5 (-CH2-), 28.7
(-CH2-), 28.7 (-CH2-), 28.9 (-CH2-), 28.9 (-CH2-), 29.0
(-CH2-), 33.7 (-CH2-), 68.4 (-OCH2-), 100.9 (C-8), 115.0 (C-
6), 115.0 (C-3′, C-5′), 117.5 (C-10), 122.4 (C-1′), 123.7 (C-3),
126.9 (C-5), 130.0 (C-2′, C-6′), 153.0 (C-2), 157.2 (C-9), 157.4
(C-4′), 163.0 (C-7), 174.5 (C-4, -COOH); MS m/z 453.4 (M +
H)+. Anal. Calcd (C27H32O6): C, 71.66; H, 7.13. Found: C,
71.32; H, 7.13.
7-O-Ca r boxym eth yld a id zein (7), Its Eth yl Ester (24),
a n d 7,4′-Di-O-ca r boxym eth yld a id zein (37). To a solution
of 5.1 g of 2 (20.08 mmol), 60 mL of acetone, and 10 mL of 2
N KOH (20 mmol) was added 2.5 mL of ethyl bromoacetate
(21.54 mmol). The reaction mixture was refluxed under gentle
stirring for 72 h and refrigerated for 12 h. The precipitates
were collected on a fritted funnel and recrystallized from
acetone to yield 1.42 g of pure 24. Analyses (24): colorless
crystalline needles; yield 20.9%; mp169-171 °C; 1H NMR
(DMSO-d6) δ 1.22 (t, 3H, -CH3), 4.19 (dd, 2H, -CH2O-), 4.98
(s, 2H, -CH2O-), 6.81 (dd, 2H, J ) 8.47, 1.6 Hz, H-3′, H-5′),
7.10 (dd, 1H, J ) 8.88, 2.6 Hz, H-6), 7.16 (d, 1H, J ) 2.6 Hz,
H-8), 7.40 (dd, 2H, J ) 8.47, 1.6 Hz, H-2′, H-6′), 8.03(d, 1H, J
) 8.88 Hz, H-5), 8.36 (s, 1H, H-2), 9.55 (s, 1H, OH-4′); 13C NMR
δ 14.02(-CH3), 60.87 (-CH2O-), 65.08 (-CH2O-), 101.6 (C-
8), 114.8 (C-6), 114.9 (C-3′, C-5′), 118.1 (C-10), 122.3 (C-1′),
123.7 (C-3), 127.1 (C-5), 130.1 (C-2′, C-6′), 153.2 (C-2), 157.1
(C-9), 157.3 (C-4′), 161.9 (C-7), 168.1 (-OCO-), 174.7 (C-4);
MS m/z 341.2 (M + H)+. Anal. Calcd (C19H16O6): C, 67.05; H,
4.73. Found: C, 67.20; H, 4.67. The filtrate was concentrated
and fractionated on a Sephadex LH-20 column in chloroform/
methanol (7/3). Fractions that contained the diethyl ester of
37 were pooled and recrystallized in acetone to give 0.5 g of
white crystalline 7,4′-diethylcarbonylmethoxyisoflavone. The
mono- and dicarboxymethyldaidzeins 7 and 37 were obtained
by hydrolyzing 0.5 g of their respective ethyl esters in 10 mL
of methanol, 1 mL of 2 N KOH, and 5 mL of water. The
reaction mixtures were refluxed with gentle stirring for 2 h.
Methanol was evaporated, and the residue was partitioned in
water and ethyl acetate. The ethyl acetate layer was collected,
dried, and recrystallized from acetone to yield pure 7 or 37.
Analyses (7): colorless crystalline plates; yield 22.7%; mp 244-
245 °C; 1H NMR (DMSO-d6) δ 4.98 (s, 2H, -CH2O-), 6.81 (d,
2H, J ) 8.47 Hz, H-3′, H-5′), 7.10 (dd, 1H, J ) 8.88, 2.6 Hz,
H-6), 7.16 (d, 1H, J ) 1.94 Hz, H-8), 7.40 (d, 2H, J ) 8.33 Hz,
H-2′, H-6′), 8.03 (d, 1H, J ) 8.86 Hz, H-5), 8.36 (s, 1H, H-2),
9.55 (s, 1H, OH-4′), 12.0 (br, 1H, -COOH); MS m/z 313.1 (M
+ H)+, 311.3 (M - H)+. Anal. Calcd (C17H12O6): C, 65.39; H,
3.87. Found: C, 64.16; H, 3.87. Analyses (37): colorless
crystalline plates; yield 3.2%; mp 268-270 °C;1H NMR spec-
trum similar to that of 7 and consistent with the interpretation
that it is the 7,4′-disubstituted analogue of 7; MS m/z 371.2

3326 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 20
Gao et al.
7-O-ω-Ca r boxyp en ta d ecyld a id zein (15) was synthesized
and purified according to the procedures described for 14 using
6-bromo-1-hexadecanoic acid (5.15 g, 15.36 mmol) as the
alkylating agent. Analyses (15): white amorphous powder;
Analyses (38): white amorphous powder; yield 13%; mp141-
142 °C;
1H NMR and 13C NMR spectra with the characteristic
features of those of 18 and consistent with the interpretation
that 38 is a 7,4′-disubstituted analogue of 18; MS m/z 539.8
(M + H)+. Anal, Calcd (C33H46O6): C, 73.58; H, 8.61. Found:
C, 73.68; H, 8.66.
1
yield 36%; mp151-152 °C; H NMR (DMSO-d6) δ 1.22-1.24
[m, 20H, (-CH2-)10], 1.40 (m, 2H, -CH2-), 1.46 (m, 2H,
-CH2-), 1.70 (m, 2H, -CH2-), 2.16 (t, 2H, -CH2-), 3.97 (t,
-CH2O-), 6.82 (dd, 2H, J ) 8.7, 1.7 Hz, H-3′, H-5′), 7.03 (dd,
1H, J ) 8.86, 1.9 Hz, H-6), 7.10 (d, 1H, J ) 1.9 Hz, H-8), 7.48
(dd, 2H, J ) 8.7, 1.7 Hz, H-2′, H-6′), 8.0 (d, 1H, J ) 8.86 Hz,
H-5), 8.34 (s, 1H, H-2), 9.5 (s, 1H, OH-4′); 13C NMR δ 24.5
(-CH2-), 25.4 (-CH2-), 25.5 (-CH2-), 28.9-29.0 [(-CH2-)10],
33.8 (-CH2-), 67.4 (-OCH2-), 100.9 (C-8), 114.1 (C-6), 114.9
(C-3′, C-5′), 117.5 (C-10), 122.4 (C-1′), 124.1 (C-3), 126.9 (C-5),
130.0 (C-2′, C-6′), 152.9 (C-2), 157.2 (C-9), 157.4 (C-4′), 162.9
(C-7), 174.6 (C-4, -COOH); MS m/z 509.7 (M + H)+. Anal.
Calcd (C31H40O6): C, 73.2; H, 7.93. Found: C, 73.41; H, 7.83.
7-O-ω-Hyd r oxyd od ecyld a id zein (19) was synthesized
and purified according to the procedures described for 18 using
12-bromo-1-dodecanol (5 g, 18.85 mmol) as the alkylating
agent. Analyses (19): white amorphous powder; yield 20.5%;
mp 142-144 °C; 1H NMR (DMSO-d6) δ 1.23 [m, 14H, (-CH2-)7],
1.37 (m, 4H, -CH2CH2-), 1.73 (m, 2H, -CH2-), 3.36 (m, 2H,
-CH2O-), 4.08 (t, -CH2O-), 6.81 (dd, 2H, J ) 8.69, 1.58 Hz,
H-3′, H-5′), 7.04 (dd, 1H, J ) 8.86, 1.9 Hz, H-6), 7.09 (d, 1H,
J ) 1.9 Hz, H-8), 7.40 (dd, 2H, J ) 8.69, 1.58 Hz, H-2′, H-6′),
7.98 (d, 1H, J ) 8.86 Hz, H-5), 8.34 (s, 1H, H-2), 9.55 (s,1H,
OH-4′); 13C NMR δ 25.4 [(-CH2-)2], 25.5 (-CH2-), 28.4
(-CH2-), 28.7 (-CH2-), 28.9 (-CH2-), 29.1 (-CH2-), 32.5
(-CH2-), 60.7 (-CH2O-), 68.4 (-OCH2-), 100.9 (C-8), 114.8
(C-6), 114.8 (C-3′, C-5′), 117.5 (C-10), 122.4 (C-1′), 123.7 (C-3),
126.9 (C-5), 130.0 (C-2′, C-6′), 153.0 (C-2), 157.2 (C-9), 157.4
(C-4′), 163.0 (C-7), 174.7 (C-4); MS m/z 439.6 (M + H)+. Anal.
Calcd (C27H34O5): C, 73.95; H, 7.81. Found: C, 74.20; H, 7.61.
7-O-ω-Hyd r oxyeth yld a id zein (16). To a suspension of 5.3
g of 2 (20.87 mmol) and 60 mL of acetone was added with
vigorous stirring 13 mL of 2 N KOH (26 mmol). After 2 was
completely dissolved, 2 mL of 2-bromoethanol (28.21 mmol)
was added, and the reaction mixture was refluxed with gentle
stirring for 72 h. Solvent was evaporated, and the residue was
partitioned in water and ethyl acetate. The ethyl acetate layer
was further purified on a silica gel column in petroleum ether/
ethyl acetate/methanol (6/3/1) followed by a Sephadex LH-20
column in methanol. Fractions that contained the major
product were pooled, dried, and recrystallized from acetone
to give 1.7 g of 16. Analyses (16): white amorphous powder;
yield 28.5%; mp 210-212 °C; MS m/z 299.1 (M + H)+. Anal.
Calcd (C17H14O5): C, 68.45; H, 4.73. Found: C, 67.86; H, 4.23.
7-O-ω-Hyd r oxyh exyld a id zein (17) was synthesized ac-
cording to the method described for 16 using 6-bromo-1-
hexanol (3 mL, 22.93 mmol) as the alkylating agent. After
reflux, the reaction mixture was refrigerated for 12 h, and the
precipitates were collected and fractionated on a Sephadex LH-
20 column in methanol. Fractions that contained pure product
were pooled, dried, and recrystallized from acetone to give 1.16
g of 17. Analyses (17): white amorphous powder; yield 16.4%;
mp177-178.5 °C; 1H NMR (DMSO-d6) δ 1.35 (m, 2H, -CH2-),
1.43 (m, 4H, -CH2CH2-), 1.75 (m, 2H, -CH2-), 3.40 (m, 2H,
-CH2-), 4.1 (t, -CH2O-), 6.81 (dd, 2H, J ) 8.7, 1.6 Hz, H-3′,
H-5′), 7.04 (dd, 1H, J ) 8.86, 1.98 Hz, H-6), 7.10 (d, 1H, J )
1.99 Hz, H-8), 7.40 (dd, 2H, J ) 8.7, 1.58 Hz, H-2′, H-6′), 8.0-
(d, 1H, J ) 8.86 Hz, 5-H), 8.34 (s, 1H, H-2), 9.53 (s, 1H, OH-
4′); 13C NMR δ 25.2 (-CH2-), 25.3 (-CH2-), 28.5 (-CH2-), 32.5
(-CH2-), 60.6 (-CH2O-), 68.4 (-OCH2-), 100.9 (C-8), 114.9
(C-6), 114.9 (C-3′, C-5′), 117.5 (C-10), 122.4 (C-1′), 123.7 (C-3),
126.9 (C-5), 130.1 (C-2′, C-6′), 153.1 (C-2), 157.2 (C-9), 157.4
(C-4′), 163.0 (C-7), 174.7 (C-4); MS m/z 355.3 (M + H)+. Anal.
Calcd (C21H22O5): C, 71.17; H, 6.26. Found: C, 70.82, H, 6.44.
7-O-ω-H yd r oxyet h yl-2-(2-oxyet h yl)oxyet h yld a id zein
(20). To a solution of 5.1 g of 2 (20.08 mmol) in 60 mL of DMF
were added 4.15 g of K2CO3 (30.03 mmol) and 6.74 g of 2-(2-
(2-chloroethoxy)ethoxy)ethanol (40.0 mmol), and the mixture
was refluxed (80 °C) with gentle stirring for 12 h. After reflux,
the reaction mixture was poured into 100 mL of ice/water
followed by ethyl acetate extraction. Ethyl acetate was evapo-
rated, and the residue was fractionated on a Sephadex LH-20
column in chloroform/ methanol (7/3). Fractions that contained
20 were pooled, dried, and recrystallized from acetone to yield
3.08 g of pure product. Analyses (20): colorless crystalline
1
needles; yield 39.9%; mp 149-150 °C; H NMR (DMSO-d6) δ
3.43 (m, 2H, -CH2-), 3.48 (m, 2H, -CH2-), 3.55 (m, 2H,
-CH2-), 3.60 (m, 2H, -CH2-), 3.79 (m, 2H, -CH2-), 4.25 (m,
2H, -CH2-), 6.81 (d, 2H, J ) 8.66 Hz, H-3′, H-5′), 7.08 (d, 1H,
J ) 8.86 Hz, H-6), 7.15 (d, 1H, J ) 1.77 Hz, H-8), 7.40 (d, 2H,
J ) 8.66 Hz, H-2′, H-6′), 8.01 (d, 1H, J ) 8.18 Hz, H-5), 8.35
(s, 1H, H-2), 9.54 (s, 1H, OH-4′); 13C NMR 60.2 (-CH2O-),
68.1 (-CH2O-), 68.6 (-CH2O-), 69.8 (-CH2O-), 69.9 (-CH2-
O-), 72.4 (-CH2O-), 101.1 (C-8), 114.9 (C-6), 115.0 (C-3′, C-5′),
117.6 (C-10), 122.4 (C-1′), 123.7 (C-3), 126.9 (C-5), 130.1 (C-
2′, C-6′), 153.1 (C-2), 157.2 (C-9), 157.3 (C-4′), 162.8 (C-7), 174.7
(C-4); MS m/z 387 (M + H)+. Anal. Calcd (C21H22O7): C, 65.28;
H, 5.74. Found: C, 65.15; H, 5.59.
7-O-ω-Eth oxyca r bon ylp en tyld a id zein (26) a n d 7,4′-Di-
O-ω-eth oxyca r bon ylp en tyld a id zein (44). To a solution of
7.7 g of 2 (30.31 mmol), 30 mL of 2 N KOH (60 mmol), and
120 mL of acetone was added 12.5 mL of ethyl-6-bromo-1-
hexanoate (70.60 mmol). The mixture was refluxed with gentle
stirring for 72 h, and the products were allowed to precipitate
under refrigeration. The precipitates were collected on a fritted
funnel and fractionated on a Sephadex LH-20 column in
chloroform/methanol (7/3). Fractions that contained 26 or 44
were pooled separately, dried, and recrystallized from acetone
to give 670 mg of 26 and 5.39 g of 44. Analyses (26): white
crystalline needles; yield 5.6%; mp 130-131 °C; 1H NMR
(DMSO-d6) δ 1.17 (t, 3H, -CH3), 1.45 (m, 2H, -CH2-), 1.58
(m, 2H, -CH2-), 1.74 (m, 2H, -CH2-), 2.31 (t, 2H, -CH2-),
4.04 (m, 2H, -CH2O-), 4.09 (m, 2H, -CH2O-), 6.82 (dd, 2H,
J ) 9.76, 2.7 Hz, H-3′, H-5′), 7.04 (dd, 1H, J ) 8.87, 2.79 Hz,
H-6), 7.09 (d, 1H, J ) 2.1 Hz, H-8), 7.40 (dd, 2H, J ) 9.76, 2.7
Hz, H-2′, H-6′), 8.0 (d, 1H, J ) 8.86 Hz, H-5), 8.34 (s, 1H, H-2),
9.55 (s, 1H, OH-4′); 13C NMR δ 14.1(-CH3), 24.2 (-CH2-), 24.9
(-CH2-), 28.1 (-CH2-), 33.4 (-CH2-), 59.7 (-CH2O-), 68.3
(-CH2O-), 100.9 (C-8), 114.9 (C-6), 115 (C-3′, C-5′), 117.5 (C-
10), 122.4 (C-1′), 123.7 (C-3), 126.9 (C-5), 130.1 (C-2′, C-6′),
153.1 (C-2), 157.2 (C-9), 157.4 (C-4′), 163.0 (C-7), 172.8
(-OCO-), 174.7 (C-4); MS m/z 397.3 (M + H)+. Anal. Calcd
(C23H24O6): C, 69.68; H, 6.10. Found: C, 69.84; H, 6.15.
Analyses (44): yield 50%; mp 99-101 °C; MS m/z 539.6 (M +
7-O-ω-Hyd r oxyn on yld a id zein (18) a n d 7,4′-d i-O-ω-h y-
d r oxyn on yld a id zein (38) were synthesized according to the
procedure described for 16 using 9-bromo-1-nonanol (5 g, 22.41
mmol) as the alkylating agent. After reflux, the reaction
mixture was refrigerated, and the precipitates formed were
collected on a fritted funnel and further purified on a Sephadex
LH-20 column in chloroform/methanol (7/3). Fractions that
contained 18 or 38 were pooled separately, concentrated, and
recrystallized from acetone to yield 305 mg of 18 and 1.4 g of
38, respectively. Analyses (18): colorless crystalline plates;
1
yield 3.8%; mp 165-167 °C; H NMR (DMSO-d6) δ 1.27 [m,
8H, (-CH2-)4], 1.40 (m, 4H, -CH2CH2-), 1.73 (m, 2H, -CH2-),
3.37 (m, 2H, -CH2-), 4.09 (t, -CH2O-), 6.81 (d, 2H, J ) 8.75,
1.58 Hz, H-3′, H-5′), 7.04 (dd, 1H, J ) 8.87, 2.1 Hz, H-6), 7.10
(d, 1H, J ) 2.1 Hz, H-8), 7.39 (d, 2H, J ) 8.75, 1.58 Hz, H-2′,
H-6′), 8.0 (d, 1H, J ) 8.86 Hz, H-5), 8.34 (s, 1H, H-2), 9.53 (s,
1H, OH-4′); 13C NMR δ 25.4 (-CH2-), 25.5 (-CH2-), 28.4
(-CH2-), 28.7 (-CH2-), 28.9 (-CH2-), 29.0 (-CH2-), 32.5
(-CH2-), 60.7(-CH2O-), 68.4 (-OCH2-), 100.9 (C-8), 114.9 (C-
6), 114.9 (C-3′, C-5′), 117.5 (C-10), 122.4 (C-1′), 123.7 (C-3),
126.9 (C-5), 130.0 (C-2′, C-6′), 153.0 (C-2), 157.2 (C-9), 157.4
(C-4′), 163.0 (C-7), 174.7 (C-4); MS m/z 397.5 (M + H)+. Anal.
Calcd (C24H28O5): C, 72.71, H, 7.12. Found: C, 73.11; H, 7.13.

Structure/ Activity of Antidipsotropic Isoflavones
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 20 3327
H)+. Anal. Calcd (C31H38O8): C, 69.13; H, 7.11. Found: C,
69.34; H, 7.11.
C, 69.68, H, 6.10. Found: C, 69.98, H, 5.83. The free acid 30
was prepared by hydrolyzing 1 g of 33 under the conditions
described for the preparation of 29 to give 855 mg of product.
Analyses (30): white amorphous powder; yield 86%; mp162-
163 °C; MS m/z 369 (M + 1)+. Anal. Calcd (C21H20O6): C, 68.47;
H, 5.47. Found: C, 68.08; H, 5.45.
7-O-ω-Hexen yld a id zein (28) a n d 7,4′-Di-O-ω-h exen yl-
d ia d zein (42). The ω-hexenyl derivatives of daidzein were
synthesized and purified according to the same procedure
described for 26 and 44 using 6-bromo-1-hexene (5 g, 30.66
mmol) as the alkylating agent. Analyses (28): yield 39%; mp
156-157 °C; 1H NMR (DMSO-d6) δ 1.51 (m, 2H, -CH2-), 1.74
(m, 2H, -CH2-), 2.08 (m, 2H, -CH2-), 4.09 (t, -OCH2-), 5.01
(m, 2H, -CH2d), 5.83 (m, 1H, -CHd), 6.81 (dd, 2H, J ) 8.66,
1.53 Hz, H-3′, H-5′), 7.03 (dd, 1H, J ) 8.88, 2.6 Hz, H-6), 7.09
(d, 1H, J ) 1.8 Hz, H-8), 7.40 (dd, 2H, J ) 8.66, 1.53 Hz, H-2′,
H-6′), 8.0 (d, 1H, J ) 8.88 Hz, H-5), 8.33 (s, 1H, H-2), 9.54 (s,
1H, OH-4′); 13C NMR δ 24.6 (-CH2-), 27.8 (-CH2-), 32.8
(-CH2-), 68.2 (-OCH2-), 100.9 (C-8), 114.9 (C-6), 114.9 (C-3′,
C-5′), 117.5 (C-10), 122.4 (C-1′), 123.7 (C-3), 126.9 (C-5), 130.0
(C-2′, C-6′), 138.4 (CH2dCH-), 153.0 (C-2), 157.2 (C-9), 157.4
(C-4′), 162.9 (C-7), 174.7 (C-4); MS m/z 337.5 (M + H)+. Anal.
Calcd (C21H20O4): C, 74.98; H, 5.99. Found: C, 74.52; H, 6.08.
7-O-r-Ca r boxyh ep tyld a id zein (31) a n d Its Eth yl Ester
(34). Compounds 31 and 34 were synthesized and purified
according to the methods described for 30 and 33 using ethyl
2-bromo-1-octanoate as the alkylating agent. Analyses (34):
1
yield 80%; mp118-120 °C; H NMR (DMSO-d6) δ 0.85(t, 3H,
J ) 6.66 Hz, -CH3), 1.18(t, 3H, J ) 7.35 Hz, -CH3), 1.26-
1.35 [(m, 6H, (-CH2-)3], 1.45 (m, 2H, -CH2-), 1.90 (m, 2H,
-CH2-), 4.16 (m, 2H, -CH2-), 5.12 (t, 1H, J ) 6.08 Hz, -CHO),
6.81 (dd, 2H, J ) 8.49, 1.6 Hz, H-3′, H-5′), 7.06 (dd, 1H, J )
8.8, 2.8 Hz, H-6), 7.08 (d, 1H, J ) 2.55 Hz, H-8), 7.39 (dd, 2H,
J ) 8.72, 1.6 Hz, H-2′, H-6′), 8.03 (d, 1H, J ) 8.79 Hz, H-5),
8.36 (s, 1H, H-2), 9.55 (s, 1H, HO-4′); 13C NMR (DMSO-d6) δ
13.9 (-CH3), 14.0 (-CH3), 21.9 (-CH2-), 24.4 (-CH2-), 28.2
(-CH2-), 31.0 (-CH2-), 31.8 (-CH2-), 61.0 (-OCH2-), 75.8
(-CHO-), 101.8 (C-8),114.9 (C-6), 115.0 (C-3′, C-5′), 118.1 (C-
10), 122.3 (C-1′), 123.7 (C-3), 127.2 (C-5), 130.1 (C-2′, C-6′),
153.2 (C-2), 157.1 (C-9), 157.3 (C-4′), 161.8 (C-7), 170.2
(CdO), 174.6 (C-4); MS m/z 425 (M + H)+. Anal. Calcd
(C25H28O6): C, 70.74, H, 6.65. Found: C, 70.82, H, 6.51.
Analyses (31): yield 96%; mp 123-125 °C; MS m/z 397 (M +
H)+. Anal. Calcd (C23H24O6): C, 69.68; H, 6.10. Found: C,
68.05; H, 6.11.
1
Analyses (42): yield 20%; mp 112-115 °C; H NMR and 13C
NMR spectra with the characteristic features of those of 28;
MS m/z 419.4 (M + H)+. Anal. Calcd (C27H30O4): C, 77.48; H,
7.22. Found: C, 77.50; H, 7.27.
7-O-r-Ca r boxyeth yld a id zein (29), Its Eth yl Ester (32),
a n d th e Dieth yl Ester of 7,4′-Di-O-r-ca r boxyeth yld a id -
zein (45). The ethyl esters 32 and 45 were synthesized
according to the method described for 20 using ethyl 2-bromo-
propionate (3 mL, 23.1 mmol) as the alkylating agent. The
reaction mixture was refluxed at 80 °C for 2 h and poured into
100 mL of ice/water. The products were extracted with ethyl
acetate and further purified on a Sephadex LH-20 column in
chloroform/methanol (3/7). Fractions that contained 32 or 45
were pooled separately, concentrated, and recrystallized from
acetone to give 5.615 g of 32 and 866 mg of 45. Analyses (32):
7-O-ω-Ca r boxyp en tylfla von e (49). A mixture of 600 mg
of 7,4′-hydroxyflavone (3.15 mmol), 515 mg of anhydrous K2-
CO3 (3.7 mmol), and 611 mg of 6-bromo-1-hexanoic acid (3.13
mmol) were refluxed in 40 mL of anhydrous DMF with
vigorous stirring for 4 h. The mixture was poured into 100
mL of ice/water and extracted with ethyl acetate. The organic
phase was dried and further purified on a Sephadex LH-20
column in chloroform/methanol (7/3). Fractions that contained
49 were pooled and solvents removed on a rotary evaporator.
The residue was recrystallized from acetone to yield 50 mg of
49. Analyses (49): yellow crystalline needles; yield 60%; mp
209-211 °C; 1H NMR (DMSO-d6) δ 1.44 (m, 2H, -CH2-), 1.55
(m, 2H, -CH2-), 1.74 (m, 2H, -CH2-), 2.23 (t, 2H, -CH2-),
4.05 (t, 2H, -CH2O-), 6.78 (s, 1H, H-3), 6.90 (dd, 1H, J )
8.16, 2.0 Hz, H-6), 6.98 (d, 1H, J ) 2.0 Hz, H-8), 7.07 (dd, 2H,
J ) 8.82, 1.6 Hz, H-3′, H-5′), 7.86 (d, 1H, J ) 8.16 Hz, H-5),
7.98 (dd, 2H, J ) 8.82, 1.6 Hz, H-2′, H-6′); 13C NMR δ 24.3
(-CH2-), 25.1 (-CH2-), 28.3 (-CH2-), 33.6 (-CH2-), 67.7
(-OCH2-), 102.5 (C-8), 105.0 (C-3), 114.8 (C-6), 114.9 (C-3′,
C-5′), 116.1 (C-10), 123.3 (C-1′), 126.5 (C-5), 128.0 (C-2′, C-6′),
157.4 (C-9), 161.3 (C-4′), 162.0 (C-2), 162.6(C-7), 174.5 (-COOH),
176.3 (C-4); MS m/z 369 (M + H)+. Anal. Calcd (C21H20O6): C,
68.47; H, 5.47. Found: C, 68.05; H, 5.85.
1
colorless crystalline needles; yield 79%; mp 167-169 °C; H
NMR (DMSO-d6) δ 1.18 (t, 3H, J ) 7.04 Hz, -CH3), 1.57 (d,
3H, J ) 6.84 Hz, -CH3), 4.17 (dd, 2H, J ) 6.95, 14.07 Hz,
-CH2-), 5.23 (d, 1H, J ) 6.80 Hz, -CH-), 6.82 (dd, 2H, J )
8.76, 1.6 Hz, H-3′, H-5′), 7.06 (dd, 1H, J ) 8.2, 1.83 Hz, H-6),
7.08 (d, 1H, J ) 1.83 Hz, H-8), 7.40 (dd, 2H, J ) 8.65, 1.6 Hz,
H-2′, H-6′), 8.03 (d, 1H, J ) 8.2 Hz, H-5), 8.35 (s, 1H, H-2),
9.55 (s, 1H, HO-4′); 13C NMR (DMSO-d6) δ 13.9 (-CH3), 17.9
(-CH3), 61.1 (-OCH2-), 72.1 (-CH-), 101.8 (C-8), 114.9 (C-
6), 114.9 (C-3′, C-5′), 118.1 (C-10), 122.3 (C-1′), 123.7 (C-3),
127.2 (C-5), 130.1 (C-2′, C-6′), 153.2 (C-2), 157.1 (C-9), 157.3
(C-4′), 161.5 (C-7), 170.7 (CdO), 174.6 (C-4); MS m/z 355 (M
+ H)+. Anal. Calcd (C20H18O6): C, 67.79; H, 5.12. Found: C,
67.65, H, 5.17. Analyses (45): colorless crystalline plates; yield
9.5%; mp 97-99 °C. Anal. Calcd (C25H26O8): C, 66.07; H, 5.77.
Found: C, 66.07; H, 5.77. The free acid 29 was prepared by
hydrolyzing 2.5 g of 32 (7.06 mmol) in 20 mL of methanol, 5.0
°
mL of 2 N KOH, and 10 mL of water at 60-80 C for 4 h. After
Abbr evia tion s
removal of methanol, the white residue was recrystallized from
acetone to give 2.0 g of 29. Analyses (29): yield 86.9%; MS
m/z 327 (M + 1)+. Anal. Calcd (C18H14O6): C, 66.24; H, 4.33.
Found: C, 66.19%, H, 4.23.
The abbreviations used are ADH, alcohol dehydroge-
nase; ALDH, aldehyde dehydrogenase; DA, dopamine;
DOPAC, 3,4-dihydroxyphenylacetic acid; DOPAL, 3,4-
dihydroxyphenylacetaldehyde; 5-HIAL, 5-hydroxyin-
dole-3-acetaldehyde; 5-HIAA, 5-hydroxyindole-3-acetic
acid; 5-HTOL, 5-hydroxytryptophol; 5-HT, 5-hydroxy-
tryptamine (serotonin); and MAO, monoamine oxidase.
7-O-r-Ca r boxyp en tyld a id zein (30) a n d Its Eth yl Ester
(33). The ethyl ester 33 was synthesized and purified by the
same methods described for 32 using ethyl 2-bromo-1-hex-
anoate (8 mL, 0.044 mol) as the alkylating agent. Analyses
(33): colorless crystalline needles; yield 85%; mp 129-131 °C;
1H NMR (DMSO-d6) δ 0.88 (t, 3H, J ) 7.09 Hz, -CH3), 1.18
(t, 3H, J ) 7.07 Hz, -CH3), 1.35 (m, 2H, -CH2-), 1.43 (m, 2H,
-CH2-), 1.91 (m, 2H, -CH2-), 4.17 (m, 2H, -CH2-), 5.12 (t,
1H, J ) 6.52 Hz, -CHO), 6.81 (dd, 2H, J ) 8.76, 1.55 Hz,
H-3′, H-5′), 7.07 (dd, 1H, J ) 8.63, 2.3 Hz, H-6), 7.08 (d, 1H,
J ) 2.57 Hz, H-8), 7.39 (dd, 2H, J ) 8.63, 1.52 Hz, H-2′, H-6′),
8.03 (d, 1H, J ) 8.63 Hz, H-5), 8.36 (s, 1H, H-2), 9.55(s, 1H,
HO-4′); 13C NMR (DMSO-d6) δ 13.8 (-CH3), 14.0 (-CH3), 21.7
(-CH2-), 26.6 (-CH2-), 31.5 (-CH2-), 61.0 (-OCH2-), 75.8
(-CHO-), 101.8 (C-8),114.9 (C-6), 114.9 (C-3′, C-5′), 118.1 (C-
10), 122.3 (C-1′), 123.7 (C-3), 127.2 (C-5), 130.1 (C-2′, C-6′),
153.2 (C-2), 157.1 (C-9), 157.3 (C-4′), 161.8 (C-7), 170.2 (Cd
O), 174.6 (C-4); MS m/z 397 (M + H)+. Anal. Calcd (C23H24O6):
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) Keung, W. M.; Vallee, B. L. Therapeutic lessons from traditional
Oriental medicine to contemporary Occidental pharmacology.
EXS 1994, 71, 371-381.

3328 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 20
Gao et al.
(3) Heyman, G. M.; Keung, W. M.; Vallee, B. L. Daidzin decreases
ethanol consumption in rats. Alcohol.: Clin. Exp. Res. 1996, 20,
1083-1087.
(4) 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.
(5) 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.
(6) Overstreet, D. H.; Lee, D. Y.-W.; Chen, Y. T.; Rezvani, A. H.
The Chinese herbal medicine NPI-028 suppresses alcohol intake
in alcohol-preferring rats and monkeys without inducing taste
aversion. Perfusion 1998, 11, 381-390.
(7) 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.
(8) Svanas, G. W.; Weiner, H. Aldehyde dehydrogenase activity as
the rate-limiting factor for acetaldehyde metabolism in rat liver.
Arch. Biochem. Biophys. 1985, 236, 36-46.
(11) Keung, W. M.; Vallee, B. L. Daidzin and its antidipsotropic
analogs inhibit serotonin and dopamine metabolism in isolated
mitochondria. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 2198-2203.
(12) Rooke, N.; Li, D. J .; Li, J .; Keung, W. M. The mitochondrial
monoamine oxidase-aldehyde dehydrogenase pathway: a po-
tential site of action of daidzin. J . Med. Chem. 2000, 43, 4169-
4179.
(13) 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.
(14) 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.
(15) 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.
(16) Nilsson, G. E.; Tottmar, O. Biogenic aldehydes in brain: On their
preparation and reactions with rat brain tissue. J . Neurochem.
1987, 48, 1566-1572.
(17) Steinmetz, C. G., Xie, P.; Weiner, H.; Hurley, T. D. Structure of
mitochondrial aldehyde dehydrogenase: the genetic component
of ethanol aversion. Structure 1997, 7, 701-711.
(18) Ni, Li.; Zhou, J .; Hurley, T. D.; Weiner, H. Human liver
mitochondrial aldehyde dehydrogenase: Three-dimensional struc-
ture and the restoration of solubility and activity of chimeric
forms. Protein Sci. 1999, 8, 2784-2790.
(9) Keung, W. M.; Klyosov, A. A.; Vallee, B. L. Daidzin inhibits
mitochondrial aldehyde dehydrogenase and suppresses ethanol
intake of Syrian golden hamsters. Proc. Natl. Acad. Sci. U.S.A.
1997, 94, 1675-1679.
(10) 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.
J M0101390