Free-radical-scavenging, antityrosinase, and cellular melanogenesis inhibitory activities of synthetic isoflavones

Article abstract of DOI:10.1002/cbdv.201400208

In this study, we examined the potential of synthetic isoflavones for application in cosmeceuticals. Twenty-five isoflavones were synthesized and their capacities of free-radical-scavenging and mushroom tyrosinase inhibition, as well as their impact on cell viability of B16F10 murine melanoma cells and HaCaT human keratinocytes were evaluated. Isoflavones that showed significant mushroom tyrosinase inhibitory activities were further studied on reduction of cellular melanin formation and antityrosinase activities in B16F10 melanocytes in vitro. Among the isoflavones tested, 6-hydroxydaidzein (2) was the strongest scavenger of both ABTS^.+ and DPPH^. radicals with SC₅₀ values of 11.3±0.3 and 9.4±0.1 μM, respectively. Texasin (20) exhibited the most potent inhibition of mushroom tyrosinase (IC₅₀ 14.9±4.5 μM), whereas retusin (17) showed the most efficient inhibition both of cellular melanin formation and antityrosinase activity in B16F10 melanocytes, respectively. In summary, both retusin (17) and texasin (20) exhibited potent free-radical-scavenging capacities as well as efficient inhibition of cellular melanogenesis, suggesting that they are valuable hit compounds with potential for advanced cosmeceutical development.

Full text of DOI:10.1002/cbdv.201400208

CHEMISTRY & BIODIVERSITY – Vol. 12 (2015)
963
Free-Radical-Scavenging, Antityrosinase, and Cellular Melanogenesis
Inhibitory Activities of Synthetic Isoflavones
by Tzy-Ming Lu*a), Horng-Huey Kob), Lean-Teik Ngc), and Yen-Pin Hsiehd)
a) Department of Pharmacy, Tajen University, No. 20 Wei-Xin Rd., Yanpu, Pingtung 907, Taiwan, ROC
(phone: þ886-8-7624002 ext. 2727; fax: þ886-8-7625308; e-mail: tzminglu@gmail.com,
cmlu@tajen.edu.tw)
b) Department of Fragrance and Cosmetic Science, Kaohsiung Medical University, Kaohsiung 807,
Taiwan, ROC
c) Department of Agricultural Chemistry, National Taiwan University, Taipei 106, Taiwan, ROC
d) Department of Pharmacy, Fooyin University Hospital, Pingtung 917, Taiwan, ROC
In this study, we examined the potential of synthetic isoflavones for application in cosmeceuticals.
Twenty-five isoflavones were synthesized and their capacities of free-radical-scavenging and mushroom
tyrosinase inhibition, as well as their impact on cell viability of B16F10 murine melanoma cells and
HaCaT human keratinocytes were evaluated. Isoflavones that showed significant mushroom tyrosinase
inhibitory activities were further studied on reduction of cellular melanin formation and antityrosinase
activities in B16F10 melanocytes in vitro. Among the isoflavones tested, 6-hydroxydaidzein (2) was the
.
.
þ
strongest scavenger of both ABTS and DPPH radicals with SC50 values of 11.3Æ0.3 and 9.4Æ0.1 mm,
respectively. Texasin (20) exhibited the most potent inhibition of mushroom tyrosinase (IC50 14.9Æ
4.5 mm), whereas retusin (17) showed the most efficient inhibition both of cellular melanin formation and
antityrosinase activity in B16F10 melanocytes, respectively. In summary, both retusin (17) and texasin
(20) exhibited potent free-radical-scavenging capacities as well as efficient inhibition of cellular
melanogenesis, suggesting that they are valuable hit compounds with potential for advanced
cosmeceutical development.
Introduction. – Excessive solar exposure usually causes the generation of cellular
reactive oxygen species (ROS), which may lead to skin lesions including aging,
wrinkles, melasma, hyperpigmentation, and tumors [1]. Although antioxidants have
been widely applied both in skin protection and beautification, the increase of
consumer demands has encouraged cosmetic producers to launch more research in the
development of new cosmetic ingredients and formulations since the 1980ꢀs [2]. The
major concerns of consumers about cosmetic formulations are the substantive efficacy
and safety of the antioxidants or cosmeceuticals [2][3]. Many botanicals are considered
as safer ingredients when applied in cosmetics [3][4]. Natural products currently used
in commercial cosmetics such as arbutin and kojic acid are popular for their skin
whitening properties, but they are deficient in antioxidant capacity. Therefore, the
development of safer and more effective cosmeceuticals for skin care from botanicals
continues to be an important research target [5].
Flavonoids, a family of well-known polyphenols, have long received a great deal of
attention for use in the cosmeceutical development [4–7]. They are known to have
numerous advantages when used as cosmetic ingredients: i) they are effective in blood
vessel protection, antiplatelet aggregation, and capillary permeability decrease [6], ii)
ꢁ 2015 Verlag Helvetica Chimica Acta AG, Zürich

964
CHEMISTRY & BIODIVERSITY – Vol. 12 (2015)
numerous flavonoids have demonstrated powerful tyrosinase inhibitory activities,
suggesting their potential in skin lightening applications [5], and iii) flavonoids are
believed to be innocuous and pharmacologically beneficial botanicals that can satisfy
the safety prerequisite in cosmeceutical applications [3][4][7].
Isoflavones represent a subfamily of the flavonoid. Genistein, the most distin-
guished isoflavone, is known as cardiovascular-protecting phytoestrogen. It can
eradicate UV-induced skin aging [8], and its glucoside is also capable of inhibiting
melanoma cell growth [9], but it lacks the tyrosinase inhibitory activity [10–12]. In
addition, some ortho-dihydroxylated isoflavones obtained from fermented soy food-
stuffs were shown to have notable pharmacological benefits in cosmeceutical
applications [10–16]. For example, 7,8,4’-trihydroxyisoflavone (1) and 5,7,8,4’-tetrahy-
droxyisoflavone were suicide substrates of mushroom tyrosinase [10][13][14] and
effective cellular melaninogenetic inhibitors [12][13], 6,7,4’-trihydroxyisoflavone (2)
was reported to be a potent mushroom tyrosinase inhibitor [10][11], whereas 7,3’,4’-
trihydroxyisoflavone showed potent inhibitory efficacy both on mushroom tyrosinase
(EC 1.14.18.1) and on melanin formation of B16 melanocytes [12]. Calycosin, an
isoflavone isolated from Astragalus membranaceus, also exhibited effective inhibition
of melanin biosynthesis [16]. Furthermore, some of these isoflavones were reported to
possess free-radical-scavenging activities [15][17][18].
Our previous works have shown that dexoybenzoins, the cyclization precursors and
potential metabolites of isoflavones [19][20], are potent antioxidants [19] and powerful
porcine arterial relaxants [20]. In this study, we further synthesized a series of
polyphenolic isoflavones and inspected their potential for cosmeceutical application.
.
.
þ
Twenty-five isoflavones were assayed on ABTS and DPPH radical scavenging,
antityrosinase activity, and impact on cell viability of B16F10 murine melanoma cells
and HaCaT human keratinocytes. Those isoflavones that showed significant mushroom
tyrosinase inhibitory activities were also assayed on reduction of cellular melanin
generation and tyrosinase inhibitory activity in B16F10 melanocytes in vitro.
Results and Discussion. – Design and Synthesis of Isoflavones. The reported one-pot
reaction facilitated the synthesis of polyoxygenated isoflavones [21]. In general, a
compound with one or two phenolic OH groups could be a suitable substrate for
tyrosinase by mimicking the ring identity of its primary substrates, tyrosine and dopa
[5]. Furthermore, compounds with more phenolic OH groups possibly have greater
antioxidant capacities [19]. StructureÀactivity relationship (SAR) studies can be
conducted to validate the biological potency contributed by various O-bearing
substituents in isoflavones. Based on the reported approach [21], in the present work,
25 isoflavones were successfully synthesized, in which the O-bearing substituents on
ring A were located at C(6), C(7), C(5) and at C(7), C(6) and at C(7), or at C(7) and
C(8), combined with various substituents at ring B [21]. In addition, the O-bearing
substituents on ring B were primarily focused at C(3’) and C(4’).
Isoflavones 3–9 were obtained by coupling of related phenols with various
phenylacetic acids with BF3 ·Et2O as catalyst, followed by cyclization with MeS-
O2ClÀDMF complex (Scheme) [21]. Because 2-hydroxydeoxybenzoins are key
intermediates in this synthesis, the major synthetic work in this study was the
previously reported cyclization of deoxybenzoins [19][20]. Besides, 7-hydroxyisofla-

CHEMISTRY & BIODIVERSITY – Vol. 12 (2015)
965
vone (10) was a product reported previously [22], isoflavone 11 was the demethylated
product of mutabilein (12), which was heated under reflux in HBr (48%)/AcOH 1:1,
and prunetin (13) was obtained by monomethylation of genistein (14). The isoflavones
used in this study can be categorized according to the position of the O-bearing
substituents on ring A as 6-oxygenated (i.e., 3), 7-oxygenated series (i.e., 4, 5, 10, and
15), 7,8-dioxygenated series (i.e., 1, 6–8, and 16–19), 6,7-dioxygenated series (i.e., 2, 9,
and 20–22), and 5,7-dioxygenated series (i.e., 11–14 and 23–25) with various
substituents on ring B (Scheme). Isoflavones 5, 8, 9, 13, 19, and 21 can also be
alternatively classified as 7-methoxylated series. The 7-methoxylated series is used for
comparing substitutionÀactivity relationship with those isoflavones with a MeO group
at C(4’) (p-anisyl ring B). Among the investigated isoflavones, 7, 18, and 19 were newly
synthesized. Isoflavone 11 was already reported as a natural product, but NMR data
were lacking [23]. Generally, the yields of synthetic isoflavones are satisfactory and
Scheme. One-Pot Synthesis and Structures of Isoflavones 1–25
a) BF3 ·Et2O, 908, 2 h, N2. b) BF3 ·DMF, MeSO2ClÀDMF, 708, 2 h. * Synthesized via process b from
related deoxybenzoins reported previously [19][20]. D Demethylated product of 12. q Methylated
product of 14.

966
CHEMISTRY & BIODIVERSITY – Vol. 12 (2015)

CHEMISTRY & BIODIVERSITY – Vol. 12 (2015)
967
they can be easily obtained. Table 1 shows the 13C-NMR data of the synthesized
isoflavones 1–9 and 11–25.
Ingredients of commercial cosmetics with dual efficacy, antioxidant and antityr-
osinase activities, are rare; if they show both activities, the performance will be better in
cosmeceutical applications. Accordingly, the evaluation assays of these isoflavones
focused on free-radical-scavenging capacity as well as mushroom tyrosinase inhibitory
.
.
þ
effect. Artificial free radicals, such as ABTS and DPPH , are commonly used in
antioxidant assays, which can provide a reliable antioxidant profile of tested
compounds. The screening results of the synthesized isoflavones are presented below.
.
.
þ
þ
ABTS Radical-Scavenging Activities. The ABTS radical-scavenging results
indicated that isoflavones with two OH groups in ortho position to each other (in 1, 2,
6–8, 16–18, and 20), either on ring A (6,7-dihydroxy or 7,8-dihydroxy) or ring B (3’,4’-
.
þ
dihydroxy; in 8), demonstrated potent ABTS radical-scavenging activities as shown
.
þ
in Table 2. These isoflavones exhibited ABTS SC50 values comparable to the positive
control vitamin C (SC50 19.1Æ0.0 and 19.0Æ0.1 mm). 6-Hydroxydaidzein (2) was the
.
þ
most active ABTS scavenger, with SC50 values of 20.1Æ1.4 and 11.3Æ1.3 mm at two
incubation intervals.
The ortho-dihydroxylated isoflavones 1, 2, 6–8, 16–18, and 20 showed the highest
.
þ
ABTS radical-scavenging activities in this study as shown in Table 2. The acidic OH
group at C(7) (or at C(4’) alternatively) could have a counterion effect after ionization,
.
þ
which may play an important role in ABTS radical scavenging, and the effect was
similar to that of the OH group at C(4) of deoxybenzoins [19]. The counterion effect
.
may have facilitated the contact of the cationic free radical (ABTS þ ) with the anionic
.
þ
scavengers [19]. The ABTS radical-scavenging activities of ortho-dihydroxylated
isoflavones were abolished thoroughly after O-methylation at C(7), which also
terminated the counterion phenomenon, such as in the cases of 1 vs. 19 and 20 vs. 22.
.
DPPH Radical-Scavenging Activities. The results indicated that ortho-dihydroxy-
.
lated isoflavones (i.e., 1, 2, 6–8, and 16–20) are usually potent DPPH radical
scavengers (Table 2). Among these isoflavones, 6-hydroxydaidzein (2) exhibited the
.
strongest DPPH radical-scavenging activity at two check intervals with SC50 values of
9.1Æ0.0 and 9.4Æ0.1 mm, which were three times stronger than for the positive control
vitamin C (SC50 28.6Æ2.5 and 28.6Æ1.7 mm). Retusin (17; SC50 12.3Æ0.6 and 11.8Æ
0.8 mm) and texasin (20; SC50 13.7Æ0.5 and 11.7Æ0.4 mm) both showed nearly 2.5-fold
.
higher DPPH radical-scavenging activities than the positive control. 8-Hydroxydaid-
zein (1; SC50 16.5Æ0.5 and 14.3Æ1.1 mm), the isomer of 6-hydroxydaidzein (2), also
.
exhibited a DPPH radical-scavenging activity of about twice the potency of vitamin C.
Furthermore, the other ortho-dihydroxylated isoflavones including 6–8, 16, and 18
.
generally showed DPPH radical-scavenging activities comparable to that of vitamin C
.
(Table 2). Briefly, the sequence in DPPH radical-scavenging potency based on the
SC50 values was 2>20ꢀ17>1>18>8>16>7ꢀvitamin Cꢀ6.
.
The obtained powerful DPPH radical scavengers including 8-hydroxydaidzein (1),
6-hydroxydaidzein (2), retusin (17), and texasin (20) suggested that the two OH groups
in ortho position of an isoflavone play an important role in the activity. Although
isoflavone 8, which has an ortho-dihydroxylated ring B, also exhibited significant
.
DPPH radical-scavenging activity, its potency was only about half of the former
mentioned isoflavones. Thus, the free-radical-scavenging capability of an isoflavone

968
CHEMISTRY & BIODIVERSITY – Vol. 12 (2015)
Table 2. SC50 Values of Isoflavones on ABTS· þ and DPPH· Free Radical Scavenging Activity, and IC50
Values on Mushroom Tyrosinase Inhibitory Assaya)
Compound
SC50 [mm]
IC50 [mm]
Tyrosinase
30 min
.
.
þ
ABTS
7 min
DPPH
30 min
20 min
60 min
60 min
1
2
3
4
5
6
7
8
9
10
11
12
13
14
24.9Æ0.4
24.7Æ0.6
16.5Æ0.5
14.3Æ1.1
45.0Æ0.9
15.6Æ2.4
45.8Æ1.8
110.9Æ10.2
38.1Æ8.6
53.9Æ11.9
96.4Æ6.3
129.9Æ2.3
83.5Æ4.2
94.6Æ11.2
181.3Æ0.9
180.4Æ7.3
113.9Æ8.5
ns
42.3Æ6.6
24.3Æ4.2
52.3Æ8.0
129.2Æ2.2
54.4Æ1.9
54.2Æ9.2
100.2Æ6.3
146.8Æ5.4
118.8Æ7.8
108.9Æ6.4
197.7Æ4.9
169.2Æ10.0
105.9Æ11.9
ns
20.1Æ1.4
11.3 Æ 1.3
9.1Æ0.0
9.4Æ0.1
nsc)
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
25.5Æ0.4
25.2Æ0.4
30.0Æ1.2
29.3Æ1.2
25.2Æ0.0
24.9Æ0.1
28.8Æ0.3
28.3Æ0.2
25.2Æ1.8
23.2Æ1.5
23.6Æ1.1
20.3Æ0.9
ndd)
nd
nd
nd
ns
ns
ns
nd
ns
ns
ns
nd
ns
ns
ns
nd
ns
ns
ns
nd
nd
nd
nd
nd
15
nd
nd
nd
nd
ns
ns
16
17
18
19
20
21
22
23
28.4Æ1.0
28.1Æ0.8
24.2Æ0.4
23.8Æ0.5
11.8Æ0.8
16.9Æ1.0
79.2Æ5.6
11.7Æ0.4
ns
ns
ns
ns
ns
76.1Æ7.6
50.9Æ1.8
82.8Æ5.0
77.2Æ10.7
8.1Æ0.2
74.0Æ9.2
51.8 Æ 4.6
105.5Æ10.7
85.2Æ7.4
14.9Æ4.5
199.9Æ22.8
154.3Æ19.3
199.6Æ15.9
166.0Æ24.0
177.7Æ8.2
–
26.3Æ1.6
26.5Æ1.7
12.3Æ0.6
26.1Æ0.6
25.9Æ0.9
17.5Æ0.9
ns
ns
113.0Æ8.4
44.3Æ5.7
27.8Æ0.0
13.7Æ0.5
nd
nd
ns
145.9Æ19.6
141.1Æ4.4
170.0Æ10.4
141.3Æ4.7
171.1Æ4.2
–
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
24
25
Vitamin Ce)
Arbutinf)
19.1Æ0.0
19.0Æ0.1
28.6Æ2.5
28.6Æ1.7
–
–
–
–
154.0Æ4.9
146.2Æ8.3
c
a) Data are presented as mean ÆSD (nꢀ4). b) SC50, concentration of 50% free radical scavenged. ) ns,
no significance (IC50 >300 mm). d) nd, not detected due to poor solubility (for 9, 13, and 21) or neglected,
such as genistein (14) and daidzein (15). e) Vitamin C (ascorbic acid) was used as positive control of
f
radical scavengers. ) Arbutin, a commercial whitening agent, was used as positive control in the anti-
tyrosinase assay.
seems to be better when the two OH groups in ortho position are located on ring A. In
addition, the free OH group at C(7) that keeps the two ortho-positioned OH groups of
ring A intact can be crucial for the free-radical-scavenging activity of an isoflavone; for
example, the free-radical-scavenging activities of both 1 and 20 disappeared when the
OH group at C(7) was methylated (in 19 and 22). According to the scaffold of an
isoflavone, the OH groups at C(7) and C(4’) are both acidic due to the electron-
withdrawing effect of g-pyronic ring C. In other words, the ring C system of an
isoflavone will keep the electrons of the acidic OH group from being taken away by a
radical but the ortho OH group can disperse the electron-withdrawing effect and
donate enough electrons for radical scavenging. The electron-donating property of the

CHEMISTRY & BIODIVERSITY – Vol. 12 (2015)
969
ortho OH group can stabilize the isoflavone C(7)-aroxyl radical formed after one
electron is lost [24]. Furthermore, the isoflavone aroxyl radical can endure an
additional electron loss to scavenge another free radical [24]; this explains why the
ortho-dihydroxylated series (i.e., 1, 2, 6–8, 16–18, and 20) exhibited stronger free-
radical-scavenging activities. In brief, the most suitable skeleton to show powerful
radical-scavenging activity of an isoflavone could be an ortho-dihydroxylated ring A
combined with an electron-rich ring B that has an electron-donating substituent, such
as OH or MeO, in para position.
Mushroom Tyrosinase Inhibitory Assay. The mushroom tyrosinase (EC 1.14.18.1)
inhibitory assay is a commonly used technique for preliminary screening of skin-
whitening agents. Table 2 shows that besides genistein (14) and daidzein (15) all other
synthetic isoflavones appeared to have antityrosinase activities, with IC50 values varying
from 14.9Æ4.5 (20) to 199.9Æ22.8 mm (21). The order of potency for isoflavones
exhibiting antityrosinase activities stronger than that of the positive control arbutin
(IC50 154.0Æ4.9 and 146.2Æ8.3 mm) was texasin (20)>6-hydroxydaidzein (2)>8-
hydroxydaidzein (1)>retusin (17)ꢀ3>6ꢀisoformononetin (5)>16>19>7>18ꢀ
13>10>9>4>arbutin. Texasin (20), 6,7-dihydroxy-4’-methoxyisoflavone, exhibited
the most potent mushroom tyrosinase inhibitory activity, which was five to nine times
stronger than for arbutin. Considering the ring A oxygenated pattern of isoflavone, 3
(the only isoflavone that was only oxygenated at C(6)) showed potent antityrosinase
activity with IC50 values of 45.8Æ1.8 and 52.3Æ8.0 mm in the two test intervals. In the 7-
oxygenated series, isoformononetin (5) also possessed notable antityrosinase activity
(IC50 38.1Æ8.6 and 54.4Æ1.9 mm), whereas the strongest inhibitor among the 7,8-
dioxygenated series (i.e., 1, 6–8, and 16–19) was 8-hydroxydaidzein (1; IC50 45.0Æ0.9
and 42.3Æ6.6 mm). Texasin (20) was the most prominent mushroom tyrosinase
inhibitor in the 6,7-dioxygenated series, whereas prunetin (13), which belongs to the
5,7-dioxygenated series, showed only moderate activity (IC50 113.9Æ8.5 and 105.9Æ
11.9 mm). Among the C(7)-methoxylated series (i.e., 5, 8, 9, 13, and 19), the order of
antityrosinase activities was 5>19>13>9>8.
Based on the results shown in Table 2, the SAR concerning the antityrosinase
activities of the tested isoflavones can be hardly defined. It can be noted that
isoflavones having free OH groups at C(6), C(7), or C(4’) usually exhibited potent
antityrosinase activities (Table 2). However, the activity becomes irregular when both
the C(7) and C(4’) were hydroxylated in the same scaffold; for example, both genistein
(14) and daidzein (15) were inactive, whereas 6,4’-dihydroxy-7,8-dimethylisoflavone
(3), 8-hydroxydaidzein (1), and 6-hydroxydaidzein (2) were powerful tyrosinase
inhibitors (Table 2). Overall, the investigated isoflavones can be divided into non-
ortho-dioxygenated and ortho-dioxygenated (i.e., 1, 2, 6–9, and 16–22) series to
differentiate their tyrosinase inhibitory properties.
Among the non-ortho-dioxygenated series, 4, 10–12, and 23–25 were isoflavones
with an OH group at C(7) that exhibited mild mushroom tyrosinase inhibitory effects.
On the other hand, isoflavones 3, 5, and 13 were C(4’)-hydroxylated isoflavones that
showed more effective inhibition of mushroom tyrosinase activities. Apparently, the
tyrosinase inhibitory activities of both genistein (14) and daidzein (15) were restored by
O-methylation either at C(7) or at C(4’). In addition, a free OH group at C(6) of an
isoflavone seems to be compatible with a OH group at C(4’) in the case of 3, which is

970
CHEMISTRY & BIODIVERSITY – Vol. 12 (2015)
also a potential tyrosinase inhibitor. The OH group at C(4’) or C(7) of an isoflavone
might have imitated the tyrosyl ring of the enzyme substrate when it contacts
tyrosinase. Briefly, in isoflavones having a C(4’)-hydroxylated ring B, the enzyme-
substrate binding seems to be more feasible than in C(7)-hydroxylated isoflavones in
the non-ortho-dihydroxylated series based on the IC50 values (Table 2).
In this study, compounds of the ortho-dihydroxylated series (i.e., 1, 2, 6–9, and 16–
22) with OH groups at C(7) and C(8) or at C(6) and C(7) appeared to be good
mushroom tyrosinase inhibitors, with the exception of the C(7)-methoxylated
derivatives (i.e., 8, 9, 19, 21, and 22). However, O-methylation at C(4’) showed an
inconsistent substitution effect on antityrosinase activity; for example, retusin (17)
retained ca. 80% of the activity of 8-hydroxydaidzein (1), whereas texasin (20) was
60% more active than 6-hydroxydaidzein (2). This outcome confirmed that the
mushroom tyrosinase inhibitory manner of the ortho-dihydroxylated series was entirely
different from that of the non-ortho-dihydroxylated series. Basically, O-bearing
substituents at C(4’) of ring B in the series seem to contribute to the antityrosinase
activity, whereas the same substituents (OH or MeO) at C(3’) and C(2’) of ring B of an
isoflavone tend to attenuate the tyrosinase inhibitory effect. Taken together, when an
isoflavone molecule is having a lipophilic end combined with a phenolic terminal, the
binding to mushroom tyrosinase is generally more feasible both in the non-ortho- and
ortho-dihydroxylated series. Consequently, the existence of a lipophilic end in
hydroxylated isoflavones may contribute to the binding to mushroom tyrosinase.
Impact on B16F10 Melanoma Cells and HaCaT Keratinocytes Viability. The critical
factors of chemicals to be used as cosmeceuticals can be the non-hazardous property to
skin cells and the lack of systemic adverse reaction when applied to human body. The
cell viability assay on HaCaT human keratinocytes was used to check compounds for
possible toxicity or occurring irritation when applied to human skin. Although B16F10
murine melanoma cells are tumor cells and if the tested compounds showed
melanomacytotoxic activities, they would be valuable in antitumor research, the task
was hereby focused on the inhibition of melanin biosynthesis by isoflavones in vitro. In
other words, the tested compounds should not kill the melanoma cells to perform the
cellular antityrosinase activity and subsequent inhibition on cellular melanogenesis.
Table 3 shows the impact on cell viability of both melanoma cells and keratinocytes by
different concentrations of isoflavones 1–13 and 16–25. Only isoflavones 6, 12, 16, 17,
and 24 showed significant inhibitory effects on the cell viability of B16F10 melanoma
cells at concentrations of 40 and 80 mm. All other isoflavones appeared to be non-
cytotoxic to melanoma cells at the highest treatment concentration (80 mm). Although
isoflavones 6, 16, and 17 also exhibited mild toxic effects to HaCaT keratinocyte cells
(< 50% inhibition of cell viability) at 40 mm, it became innocuous when the
concentration was ꢁ20 mm. Although isoflavones 6, 12, 16, 17, and 24 showed
proliferative inhibition on melanoma cells at a concentration of 40 mm, they were less
toxic to keratinocytes, suggesting that isoflavones are almost non-toxic to normal cells,
a criterion favored in cosmeceutical applications.
Inhibition of Cellular Melanin Formation and Tyrosinase Activity in B16F10
Melanocytes. The selected isoflavones 3, isoformononetin (5), 9, prunetin (13), retusin
(17), and texasin (20) were usually potent mushroom tyrosinase inhibitors among their
series except 9, which was used for comparison with C(7)-methoxylated analogs (5 and

CHEMISTRY & BIODIVERSITY – Vol. 12 (2015)
971
Table 3. Cell Viability Impact of Isoflavones 1–13 and 16–25 on B16F10 Murine Melanoma Cells and
HaCaT Human Keratinocytesa)
Compound
Cell viability [% of control]/Concentration [mm]
B16F10
40
HaCaT
20
97.1Æ4.9
80
40
94.9Æ5.0
1
2
3
4
5
6
7
8
9
10
11
12
13
16
17
18
19
20
21
22
23
24
25
101.0Æ2.8
95.1Æ1.0*
97.6Æ3.9
98.8Æ0.9
92.3Æ6.9
7.4Æ13.2
95.1Æ12.1
86.4Æ9.8
102.7Æ3.2
101.3Æ0.7
99.7Æ6.2
26.7Æ11.1
97.8Æ2.5
41.6Æ5.8
35.7Æ1.6
99.2Æ10.2
70.7Æ7.3**
94.6Æ2.8*
99.9Æ3.8
99.2Æ3.2
91.5Æ7.7
86.9Æ7.2
84.6Æ3.0**
98.3Æ4.6
63.7Æ9.4**
58.1Æ2.9**
69.3Æ1.9**
61.8Æ0.8**
5.0Æ10.0
97.5Æ3.8
102.6Æ4.8
99.1Æ10.7
98.7Æ6.0
98.1Æ6.9
94.9Æ6.7
101.5Æ7.4
98.5Æ9.0
108.0Æ2.5**
95.7Æ6.4
93.3Æ6.1
97.9Æ3.3
97.7Æ10.4
99.2Æ5.8
96.2Æ6.3
92.8Æ9.4
103.9Æ2.7
105.2Æ3.6
99.6Æ3.6
94.1Æ5.2
98.5Æ7.4
98.1Æ5.7
95.5Æ7.0
99.6Æ5.4
100.7Æ5.8
91.5Æ8.6
45.3Æ9.2
92.6Æ2.1*
93.4Æ4.9
95.2Æ5.1
99.8Æ3.6
94.2Æ6.8
73.9Æ4.7
88.6Æ6.8
46.3Æ9.1
40.4Æ7.6
96.8Æ6.7
91.0Æ6.5
84.9Æ5.1
105.0Æ2.2**
98.7Æ6.3
69.9Æ7.6
73.9Æ3.9
97.8Æ4.3
52.9Æ6.6
62.3Æ11.4**
90.1Æ5.4
82.1Æ12.4
100.2Æ3.4
9.9Æ7.5
90.8Æ9.5
28.3Æ9.5
3.2Æ14.8
78.2Æ3.3**
32.5Æ13.8**
56.6Æ11.8**
95.4Æ8.0
100.0Æ5.6
47.3Æ7.7
7.7Æ12.9
84.8Æ2.8**
a) Data are presented as mean ÆSD (n¼4). * P<0.05, ** P<0.01 indicates significant differences.
13) and its isomer, texasin (20). Because similar experiments had been performed on 8-
hydroxydaidzein (1) [12][13], retusin (17) of the 7,8-dioxygenated series was chosen.
The quality of cellular tyrosinase activity and melanogenesis of melanoma cells were
given as percent of control (set as 100%). a-Melanocyte-stimulating hormone (a-
MSH) is a melanotropin commonly used to confirm the capability of melanin
biosynthesis and is a competitive target in melanoma cells. Based on the results shown
in Table 3, the tested concentrations of selected isoflavones were the maximum non-
toxic dosages to melanoma cells; this was to assure that the inhibition of cellular
tyrosinase activity and the reduction of melanin deposits did not result from the
cytotoxic activity of isoflavones towards the cells.
Retusin (17; 10 mm) and texasin (20; 40 mm) attenuated the cellular melanin
formation from 182.6 (a-MSH) to 56.6 and 53.5%, respectively, which means ca. 70%
reduction compared to the a-MSH group (Fig. 1). The pigmentation conditions of both
retusin- (17) and texasin (20)-treated groups showed apparent lightening effects that
could be verified with the naked eye as shown in Fig. 2. Isoformononetin (5) and
prunetin (13) showed marginal inhibition of cellular melanin formation, but without
affecting the cellular tyrosinase activity when compared with that of the a-MSH group.

972
CHEMISTRY & BIODIVERSITY – Vol. 12 (2015)
Fig. 1. Cellular anti-pigmentation effect of selected isoflavones on a-MSH-stimulated B16F10 murine
melanoma cells. Treatment: 3, 5, 9, 13, and 20 were 40 mm, and 17 was 10 mm (maximum non-toxic dosages
to melanoma cells). The quality of cellular tyrosinase activity and melanin deposits influenced by
isoflavones were compared with those of the a-MSH group, values are given as percentage of control
(standardized as 100%). Data represents mean ÆSD. # Indicates significant differences from the control
group. * P<0.05, ** P<0.01 indicates significant differences from the a-MSH-treated group.
Fig. 2. Melanin deposits of B16F10 murine melanoma cells treated by isoflavones
3 (40 mm),
isoformononetin (5, 40 mm), 9 (40 mm), prunetin (13, 40 mm), retusin (17, 10 mm), and texasin (20,
40 mm) according to Fig. 1. Both the melanin coloring of retusin (17) and texasin (20) appeared lighter
than that of any other compound apparently.
Unexpectedly, isoflavones 3 and 9 were found to enhance cellular tyrosinase activity,
causing more melanogenesis of B16F10 cells (Fig. 1).
In this study, both retusin (17) and texasin (20) were demonstrated to eliminate ca.
70% of the a-MSH stimulated cellular melanin formation and tyrosinase activity at
different tested concentrations. These two isoflavones seem to have an impact on the

CHEMISTRY & BIODIVERSITY – Vol. 12 (2015)
973
tyrosinase translation. Interestingly, both 3 and 9 not only significantly elevated cellular
melanin formation (þ 76 and þ103% of the a-MSH group) but also increased
tyrosinase activity (þ28.5 and þ34.1% of the a-MSH group; Fig. 1). In addition,
isoflavone 3 was found to have a similar structure to that of 9, both have OH groups at
C(4’) and C(6) as well as a lipophilic substituent at C(7) (Me vs. MeO, resp.). This
outcome suggested that a free OH group at C(7) could be vital for inhibiting cellular
tyrosinase activity and reduction of melanin formation. In contrast, when a lipophilic
substituent was located at C(7) in a C(6)-hydroxylated isoflavone, it could become a
cellular melanin inducer. The inconsistency between mushroom tyrosinase and cellular
tyrosinase inhibition of 3, 5, 9, and 13 confirmed that the OH group at C(7) of
isoflavone could be the key functional group to perform reduction in cellular
melanogenesis when compared with both 17 and 20 as well as the reported cellular
melanin formation inhibitors including 8-hydroxydaidzein (1) [12][13], 7,3’,4’-trihy-
droxyisoflavone [12], and calycosin [16]. The binding of mushroom tyrosinase and
phenolic substrates should have been performed randomly under non-cellular
conditions. However, this binding in biosystems could have been limited by several
elements such as membrane permeability or bioavailability of the substrate, other
ligand-substrate binding competition, cellular xenobiotic metabolism, and cellular
elimination. For example, oxidative demethylating metabolism could have transformed
isoformononetin (5) and prunetin (13) into daidzein (15) and genistein (14),
respectively, which were inactive mushroom tyrosinase inhibitors (Table 2) [10][12].
To determine whether cellular tyrosinase suppressive or expressive properties were
conducted in melanoma cells, advanced tests, such as using western blotting to
determine if enzyme translation was interfered by these isoflavones, would be
necessary. Although having a higher IC50 value (51.8 Æ 4.6 mm) in mushroom tyrosinase
inhibitory assay, retusin (17) showed almost the same inhibitory effect both on cellular
tyrosinase activity and on melanogenesis of B16F10 cells at a lower treatment
concentration (10 mm) when compared with texasin (20; 40 mm). Both the interfering of
tyrosinase translation and the tyrosinase suicide-substrate effect of 7,8-dihydroxylated
isoflavones might have contributed to the cellular antityrosinase efficacy of retusin (17)
[10][13][14].
Conclusions. – 8-Hydroxydaidzein (1) and 6-hydroxydaidzein (2), both natural
products isolated from fermented soy foodstuffs [10–12], are known to be potential
mushroom tyrosinase inhibitors. However, their yields and natural sources are limited
[10–15][18]. The present one-pot synthesis of isoflavones makes the compounds of the
6,7- and 7,8-dioxygenated series more accessible [21], especially the 7,8-dihydroxylated
isoflavones that usually can be obtained in high yields. Retusin (17) and texasin (20),
the C(4’)-methoxylated derivatives of 1 and 2, might have the advantage of being
absorbed by melanoma cells due to the expression of s-receptors [25][26]. The design
of s-receptorÀligand bonded imaging agents demonstrated that 4-MeO-containing
benzamide derivatives exhibited better uptake by melanoma cells than related 4-OH
derivatives, suggesting that the s-receptorÀligand binding effect could have facilitated
the absorption of retusin (17) and texasin (20) [27]. In fact, bioavailability studies in
animal models demonstrated that O-methylated polyphenolic isoflavones such as
formononetin and biochanin A (24) showed a higher intestinal absorption rate and

974
CHEMISTRY & BIODIVERSITY – Vol. 12 (2015)
lower hepatic conjugative metabolism than related polyphenols [28]. The better
bioavailability might be the advantage of these two isoflavones to contribute to a more
powerful inhibition on cellular melanogenesis. Although 6-hydroxydaidzein (2) was
reported to be a powerful inhibitor of mushroom tyrosinase [10][11], it failed to reduce
the cellular melanin formation in another study [12]. Texasin (20) showed not only
powerful inhibition of mushroom tyrosinase but also was capable of reducing the
cellular melanin formation, confirming that the MeOÀC(4’) group (or p-anisyl ring B)
had a better bioavailability than that of the OHÀC(4’) group when compared with 6-
hydroxydaidzein (2) [12]. In summary, the present findings have revealed that retusin
(17) and texasin (20) are both potent free-radical scavengers and efficient inhibitors of
cellular melanogenesis under non-hazardous concentrations, indicating their potential
for the development of advanced cosmeceuticals.
This work was partially supported by grants of the National Science Council,
Taiwan, ROC (NSC 99-2113-M-127-001). The assistance provided by Miss Y.-C. Chang
and Miss Y.-J. Jin in the bioassay studies both of cell-free and cellular systems is also
greatly appreciated.
Experimental Part
General. The synthetic materials, including resorcinol, pyrogallol, 3-methoxycatechol, 3-methoxy-
phenol, 2,3-dimethylhydroquinone, various phenylacetic acids, (3-hydroxyphenyl)acetonitrile, and
methyl iodide, were purchased from Tokyo Chemical Industry (Tokyo, Japan). BF3 ·Et2O, 2-
methoxyhydroquinone, phloroglucinol, pyrogallol, MeSO2Cl, and DMF were purchased from Alfa
Aesar (England). Mushroom tyrosinase (EC 1.14.18.1), 2,2’-azino-bis(3-ethylbenzothiazoline-6-sulfonic
.
acid) diammonium salt (ABTS), 2,2-diphenyl-1-picrylhydrazyl (DPPH ), 2,3-bis(2-methoxy-4-nitro-5-
sulfophenyl)-2H-tetrazolium-5-carboxanilide inner salt (XTT), a-MSH, cell culture materials, arbutin,
and ascorbic acid were purchased from SigmaÀAldrich Corp. (St. Louis, MO, USA). M.p.: Electro-
thermal 9100 melting point apparatus (Electrothermal Engineering Ltd., Landon, UK); uncorrected.
Thin layer chromatography (TLC): Kieselgel 60 or GF254 plates (SiO2; Merck, Germany). Column
chromatography (CC): Kieselgel 60 (60–230 mesh; Merck, Germany). 1H- and 13C-NMR spectra: Varian
Mercury 400þ spectrometer (400 and 100 MHz, resp.; at 258; Varian Inc., Palo Alto, USA); d in ppm rel.
to Me4Si as internal standard, J in Hz. EI-MS: Shimadzu LCMS-IT-TOF mass spectrometer (Shimadzu
Corp., Japan); in m/z. HR-EI-MS: Bruker APEX II (Bruker Instrument, Germany); in m/z.
Chemistry. General Procedure for the One-Pot Synthesis of Isoflavones 3–9. Pyrogallol (1.26 g,
10 mmol) and (3-hydroxyphenyl)acetic acid (1.52 g, 10 mmol) were heated together with BF3 ·Et2O
(25 ml) at 908 for 2 h under N2 atmosphere, then the mixture was cooled to ambient temp. After slowly
adding 15 ml of DMF to the stirred mixture in ice bath (caution: Et2O evolution), the mixture was heated
gently. When the temp. reached 50–608, 5 equiv. of MeSO2ClÀDMF (2.87 g/1.83 g) were slowly added,
and then heating at 708 was continued for 2 h. After cooling, the product was poured into 150 ml of icy aq.
AcONa (12 g/100 ml) buffer and left overnight. The precipitates were extracted with AcOEt (2Â
100 ml), followed by washing with brine and H2O, the AcOEt extract was dried (Na2SO4), and then
condensed in vacuo, followed by recrystallization (purified by CC with proper solvent system if
necessary) in MeOH/H2O 1:1 to obtain 7 (1.84 g, 68%) as colorless powder.
6,4’-Dihydroxy-7,8-dimethylisoflavone (3). Yield: 1.64 g (58%). Colorless powder (MeOH). M.p.
287–2898. Rf (AcOEt/hexane/MeOH 4 :5.5 :0.5) 0.48. 1H-NMR ((D5)pyridine): 2.33 (s, ArÀMe); 2.44 (s,
ArÀMe); 7.27 (d, J¼8.4, HÀC(3’,5’)); 7.84 (d, J¼8.4, HÀC(2’,6’)); 8.05 (s, HÀC(5)); 8.24 (s, HÀC(2));
11.74 (br. s, OH); 12.12 (br. s, OH). 13C-NMR ((D5)pyridine): see Table 1. EI-MS: 282 (Mþ ) [17].
7,3’-Dihydroxyisoflavone (4). Yield: 1.51 g (61%). Colorless needles (MeOH). M.p. 3018. Rf
(AcOEt/hexane/MeOH 4 :5.5 :0.5) 0.43. 1H-NMR (CDOD3): 6.80 (ddd, J¼0.8, 2.4, 8.0, HÀC(4’));
6.85 (d, J¼2.0, HÀC(8)); 6.93 (dd, J¼2.0, 8.8, HÀC(6)); 6.96 (dt, J¼1.2, 8.0, HÀC(6’)); 6.99 (t, J¼2.0,

CHEMISTRY & BIODIVERSITY – Vol. 12 (2015)
975
HÀC(2’)); 7.23 (t, J¼8.0, HÀC(5’)); 8.05 (d, J¼8.8, HÀC(5)); 8.15 (s, HÀC(2)). 13C-NMR (CDOD3): see
Table 1. EI-MS: 254 (Mþ ) [21].
4’-Hydroxy-7-methoxyisoflavone (¼ Isoformononetin; 5). Yield: 1.29 g (48%). Colorless granules
(MeOH). M.p. 226–2288. Rf (AcOEt/hexane/acetone 3.5 :6 :0.5) 0.44. 1H-NMR ((D5)pyridine): 3.76 (s,
MeO); 6.99 (d, J¼2.4, HÀC(8)); 7.05 (dd, J¼2.4, 8.8, HÀC(6)); 7.29 (d, J¼8.8, HÀC(3’,5’)); 7.81 (d, J¼
8.8, HÀC(2’,6’)); 8.20 (s, HÀC(2)); 8.39 (d, J¼8.8, HÀC(5)); 11.75 (br., OH). 13C-NMR ((D5)pyridine):
see Table 1. EI-MS: 268 (Mþ ) [29].
7,8-Dihydroxy-3’-methoxyisoflavone (6). Yield: 2.04 g (72%). Colorless powder (MeOH/H2O). M.p.
2538. Rf (AcOEt/hexane/MeOH 4 :5.5 :0.5) 0.41. 1H-NMR ((D5)pyridine): 3.71 (s, MeO); 6.99–7.03 (m,
HÀC(4’)); 7.37–7.39 (m, HÀC(5’,6’)); 7.38 (d, J¼8.4, HÀC(6)); 7.54 (dd, J¼1.2, 2.0, HÀC(2’)); 8.10 (d, J¼
8.4, HÀC(5)); 8.36 (s, HÀC(2)). 13C-NMR ((D5)pyridine): see Table 1. EI-MS: 284 (Mþ ). HR-EI-MS:
284.0688 (Mþ, C16H12Oþ5 ; calc. 284.0685) [30].
7,8,3’-Trihydroxyisoflavone (7). Yield: 1.84 g (68%). Colorless powder (MeOH/H2O). M.p. 2988. Rf
1
(AcOEt/hexane/MeOH 4.5 :5 :0.5) 0.40. H-NMR ((D5)pyridine): 7.19 (ddd, J¼2.0, 2.4, 8.0, HÀC(4’));
7.30 (dt, J¼1.2, 7.6, HÀC(6’)); 7.33 (d, J¼8.8, HÀC(6)); 7.37 (t, J¼8.0, HÀC(5’)); 7.78 (t, J¼1.6, HÀC(2’));
8.06 (d, J¼8.8, HÀC(5)); 8.33 (s, HÀC(2)). 13C-NMR ((D5)pyridine): see Table 1. EI-MS: 270 (Mþ ).
HR-EI-MS: 270.0533 (Mþ, C15H10O5þ ; calc. 270.0528).
8,3’,4’-Trihydroxy-7-methoxyisoflavone (8). Yield: 1.95 g (65%). Colorless powder (MeOH/H2O).
1
M.p. 2808 (dec.). Rf (AcOEt/hexane/MeOH 4.5 :5 :0.5) 0.23. H-NMR ((D5)pyridine): 3.83 (s, MeO);
7.12 (d, J¼8.8, HÀC(6)); 7.31 (dd, J¼2.0, 8.0, HÀC(6’)); 7.34 (d, J¼8.0, HÀC(5’)); 7.90 (d, J¼2.0,
HÀC(2’)); 8.06 (d, J¼8.8, HÀC(5)); 8.37 (s, HÀC(2)). 13C-NMR ((D5)pyridine): see Table 1. EI-MS: 300
(Mþ ) [17].
6,4’-Dihydroxy-7-methoxyisoflavone (9). Yield: 1.28 g (45%). Colorless needles (MeOH/H2O).
1
M.p. 3258 (dec.). Rf (AcOEt/hexane/acetone 4 :5.5 :0.5) 0.30. H-NMR ((D5)pyridine): 3.87 (s, MeO);
7.09 (s, HÀC(8)); 7.31 (d, J¼8.8, HÀC(3’,5’)); 7.81 (d, J¼8.8, HÀC(2’,6’)); 8.22 (s, HÀC(5)); 8.23 (s,
HÀC(2)). 13C-NMR ((D5)pyridine): see Table 1. EI-MS: 284 (Mþ ) [17][29].
General Cyclization Procedure of 2-Hydroxydeoxybenzoins to Isoflavones. This method was carried
out using the respective deoxybenzoins reported previously to obtain major isoflavones 1, 2, 12, and 14–
25 (Scheme) [19][20]. Biochanin A (24) was synthesized by dissolving 2,4,6-trihydroxy-4’-methoxyde-
oxybenzoin (1.37 g, 5 mmol) in DMF (6 ml), followed by slowly adding BF3 ·Et2O (10 ml) in ice bath
(caution: Et2O evolution). After removing the ice bath, the mixture was gently heated at 50–608,
and then 4 equiv. of MeSO2ClÀDMF (2.29 g/1.46 g) were slowly added, followed by heating at 708
for 2 h. The quenching process and product purification were carried out as described above for the
synthesis of 7.
7,8,4’-Trihydroxyisoflavone (¼ 8-Hydroxydaidzein; 1). Yield: 880 mg (65%). Colorless powder
1
(MeOH/H2O). M.p. 331–3328 (dec.). Rf (AcOEt/hexane/MeOH 4.5 :5 :0.5) 0.37. H-NMR (CDOD3):
6.84 (d, J¼8.8, HÀC(3’,5’)); 6.95 (d, J¼8.8, HÀC(6)); 7.36 (d, J¼8.8, HÀC(2’,6’)); 7.58 (d, J¼8.8,
HÀC(5)); 8.18 (s, HÀC(2)). 13C-NMR (CDOD3): see Table 1. EI-MS: 270 (Mþ ) [10–12][14][15]
[17][18].
6,7,4’-Trihydroxyisoflavone (¼6-Hydroxydaidzein; 2). Yield: 811 mg (60%). Colorless needles
(MeOH/H2O). M.p. 322–3238 (dec.). Rf (AcOEt/hexane/MeOH 4.5 :5 :0.5) 0.43. 1H-NMR
((D5)pyridine): 7.26 (d, J¼8.8, HÀC(3’,5’)); 7.27 (s, HÀC(8)); 7.81 (d, J¼8.8, HÀC(2’,6’)); 8.16 (s,
HÀC(2)); 8.24 (s, HÀC(5)). 13C-NMR ((D5)pyridine): see Table 1. EI-MS: 270 (Mþ
) [10–
12][14][15][17][18][29].
5,7-Dihydroxy-3’-methoxyisoflavone (¼ Mutabilein; 12). Yield: 1.23 g (87%). Colorless needles
1
(MeOH/H2O). M.p. 2048. Rf (AcOEt/hexane/MeOH 4 :5.5 :0.5) 0.69. H-NMR ((D6)acetone): 3.83 (s,
MeO); 6.30 (d, J¼2.0, HÀC(6)); 6.43 (d, J¼2.0, HÀC(8)); 6.96 (ddd, J¼1.2, 2.4, 8.4, HÀC(4’)); 7.17 (ddd,
J¼0.8, 1.2, 7.6, HÀC(6’)); 7.20 (dd, J¼2.4, 2.4, HÀC(2’)); 7.35 (t, J¼8.0, HÀC(5’)); 8.25 (s, HÀC(2)); 9.64
(s, HOÀC(7)); 12.98 (s, HOÀC(5)). 13C-NMR ((D6)acetone): see Table 1. EI-MS: 284 (Mþ ) [31].
5,7,4’-Trihydroxyisoflavone (¼Genistein; 14). Yield: 1.13 g (84%). Yellowish powder (MeOH/
H2O). M.p. 296–2988 (dec.). Rf (AcOEt/hexane/MeOH 4.5 :5 :0.5) 0.71. 1H-NMR ((D5)pyridine): 6.67
(d, J¼2.4, HÀC(6)); 6.75 (d, J¼2.4, HÀC(8)); 7.28 (d, J¼8.8, HÀC(3’,5’)); 7.72 (d, J¼8.8, HÀC(2’,6’));

976
CHEMISTRY & BIODIVERSITY – Vol. 12 (2015)
8.13 (s, HÀC(2)); 11.85 (br., OH); 13.16 (br., OH); 13.67 (s, HOÀC(5)). 13C-NMR ((D5)pyridine): see
Table 1. EI-MS: 270 (Mþ ) [10–12][17][18][21].
7,4’-Dihydroxyisoflavone (¼ Daidzein; 15). Yield: 670 mg (53%). Colorless powder (MeOH/H2O).
M.p. 3258. Rf (AcOEt/hexane/MeOH 4 :5.5 :0.5) 0.28. 1H-NMR (CDOD3): 6.84 (d, J¼8.8, HÀC(3’,5’));
6.85 (d, J¼2.0, HÀC(8)); 6.94 (dd, J¼2.4, 8.8, HÀC(6)); 7.36 (d, J¼8.8, HÀC(2’,6’)); 8.05 (d, J¼8.4,
HÀC(5)); 8.13 (s, HÀC(2)). 13C-NMR (CDOD3): see Table 1. EI-MS: 254 (Mþ ) [10–12][18][21].
7,8-Dihydroxyisoflavone (16). Yield: 1.38 g (93%). Colorless powder (MeOH/H2O). M.p. 240–2418.
1
Rf (AcOEt/hexane/MeOH 4 :5.5 :0.5) 0.38. H-NMR ((D6)acetone): 7.03 (d, J¼8.4, HÀC(6)); 7.35 (dt,
J¼1.6, 7.6, HÀC(4’)); 7.41 (dt, J¼1.6, 7.2, HÀC(3’,5’)); 7.60 (dm, J¼8.0, HÀC(2’,6’)); 7.62 (d, J¼8.4,
HÀC(5)); 8.82 (br., OH); 9.23 (br., OH). 13C-NMR ((D6)acetone): see Table 1. EI-MS: 254 (Mþ )
[17][29].
7,8-Dihydroxy-4’-methoxyisoflavone (¼ Retusin; 17). Yield: 1.10 g (77%). Colorless needles
1
(MeOH/H2O). M.p. 2758. Rf (AcOEt/hexane/MeOH 4 :5.5 :0.5) 0.43. H-NMR ((D6)acetone): 3.83 (s,
MeO); 6.97 (d, J¼8.8, HÀC(3’,5’)); 7.51 (d, J¼8.8, HÀC(5)); 7.55 (d, J¼8.8, HÀC(2’,6’)); 8.19 (s,
HÀC(2)); 8.61 (s, OH); 9.07 (s, OH). 13C-NMR ((D6)acetone): see Table 1. EI-MS: 284 (Mþ )
[17][21][29].
7,8-Dihydroxy-2’-methoxyisoflavone (18). Yield: 1.22 g (86%). Colorless powder (MeOH). M.p.
254–2558. Rf (AcOEt/hexane/MeOH 4 :5.5 :0.5) 0.36. 1H-NMR ((D6)acetone): 3.77 (s, MeO); 6.98 (dt,
J¼1.2. 8.4, HÀC(5’)); 7.01 (d, J¼8.8, HÀC(6)); 7.06 (dd, J¼1.2, 8.0, HÀC(3’)); 7.30 (dt, J¼1.6, 7.6,
HÀC(6’)); 7.35 (dt, J¼1.6, 7.6, HÀC(4’)); 7.58 (d, J¼8.4, HÀC(5)); 8.07 (s, HÀC(2)); 8.86 (br., 2 OH).
13C-NMR ((D6)acetone): see Table 1. EI-MS: 284 (Mþ ). HR-EI-MS: 284.0690 (Mþ, C16H12Oþ5 ; calc.
284.0685).
8,4’-Dihydroxy-7-methoxyisoflavone (19). Yield: 1.04 g (73%). Colorless needles (MeOH). M.p.
3118. Rf (AcOEt/hexane/MeOH 4.5 :5 :0.5) 0.38. 1H-NMR ((D5)pyridine): 7.14 (d, J¼8.8, HÀC(6));
7.27 (d, J¼8.4, HÀC(3’,5’)); 7.80 (d, J¼8.4, HÀC(2’,6’)); 8.08 (d, J¼8.8, HÀC(5)); 8.34 (s, HÀC(2)).
13C-NMR ((D5)pyridine): see Table 1. EI-MS: 284 (Mþ ). HR-EI-MS: 284.0687 (Mþ, C16H12Oþ5 ; calc.
284.0685).
6,7-Dihydroxy-4’-methoxyisoflavone (¼ Texasin; 20). Yield: 782 mg (55%). Colorless needles
(MeOH/H2O). M.p. 258–2608. Rf (AcOEt/hexane/MeOH 4 :5.5 :0.5) 0.41. 1H-NMR ((D5)pyridine):
3.67 (s, MeO); 7.05 (d, J¼8.8, HÀC(3’,5’)); 7.28 (s, HÀC(8)); 7.78 (d, J¼8.8, HÀC(2’,6’)); 8.15 (s,
HÀC(5)); 8.23 (s, HÀC(2)). 13C-NMR ((D5)pyridine): see Table 1. EI-MS: 284 (Mþ ) [17][29].
6-Hydroxy-7-methoxyisoflavone (21). Yield: 870 mg (65%). Colorless needles (MeOH). M.p. 271–
2728. Rf (AcOEt/hexane/acetone 3.5 :6 :0.5) 0.38. 1H-NMR ((D5)pyridine): 3.88 (s, MeO); 7.11 (s,
HÀC(8)); 7.37 (td, J¼7.6, 2.0, HÀC(4’)); 7.46 (t, J¼8.0, HÀC(3’,5’)); 7.84 (dd, J¼2.0, 8.4, HÀC(2’,6’)); 8.20
(s, HÀC(5)); 8.24 (s, HÀC(2)). 13C-NMR ((D5)pyridine): see Table 1. EI-MS: 268 (Mþ ) [17][29].
6-Hydroxy-7,4’-dimethoxyisoflavone (22). Yield: 1.12 g (75%). Yellowish needles (MeOH). M.p.
254–2558. Rf (AcOEt/hexane/acetone 3.5 :6 :0.5) 0.33. 1H-NMR ((D5)pyridine): 3.69 (s, MeO); 3.87 (s,
MeO); 7.07 (d, J¼8.8, HÀC(3’,5’)); 7.11 (s, HÀC(8)); 7.81 (d, J¼8.8, HÀC(2’,6’)); 8.21 (s, HÀC(5)); 8.23
(s, HÀC(2)); 11.97 (br., OH). 13C-NMR ((D5)pyridine): see Table 1. EI-MS: 298 (Mþ ) [17][29].
5,7-Dihydroxyisoflavone (23). Yield: 1.03 g (81%). Yellowish needles (MeOH/H2O). M.p. 199–
2018. Rf (AcOEt/hexane/acetone 3.5 :6 :0.5) 0.61. 1H-NMR ((D6)acetone): 6.30 (d, J¼2.4, HÀC(6)); 6.43
(d, J¼2.0, HÀC(8)); 7.39–7.46 (m, HÀC(3’,4’,5’)); 7.61 (dd, J¼1.2, 8.4, HÀC(2’,6’)); 8.23 (s, HÀC(2)); 9.78
(br., HOÀC(7)); 12.96 (s, HOÀC(5)). 13C-NMR ((D6)acetone): see Table 1. EI-MS: 254 (Mþ )
[17][21][29].
5,7-Dihydroxy-4’-methoxyisoflavone (¼ Biochanin A; 24). Yield: 1.29 g (91%). Colorless powder
(MeOH/H2O). M.p. 218–2198. Rf (AcOEt/hexane/MeOH 4 :5.5 :0.5) 0.68. 1H-NMR ((D5)pyridine):
3.71 (s, MeO); 6.67 (d, J¼2.4, HÀC(6)); 6.76 (d, J¼2.4, HÀC(8)); 7.09 (d, J¼8.8, HÀC(3’,5’)); 7.71 (d, J¼
8.8, HÀC(2’,6’)); 8.14 (s, HÀC(2)); 13.61 (s, HOÀC(5)). 13C-NMR ((D5)pyridine): see Table 1. EI-MS:
284 (Mþ ) [17][21].
5,7-Dihydroxy-2’-methoxyisoflavone (25). Yield: 1.28 g (90%). Yellow solid (CH2Cl2/MeOH). M.p.
1
201–2038. Rf (AcOEt/hexane/MeOH 4 :5.5 :0.5) 0.67. H-NMR ((D6)acetone): 3.79 (s, MeO); 6.29 (d,
J¼2.4, HÀC(6)); 6.43 (d, J¼2.4, HÀC(8)); 7.00 (dt, J¼1.2, 7.6, HÀC(5’)); 7.08 (d, J¼8.4, HÀC(3’)); 7.31

CHEMISTRY & BIODIVERSITY – Vol. 12 (2015)
977
(dd, J¼2.0, 7.6, HÀC(6’)); 7.38 (dt, J¼1.6, 8.4, HÀC(4’)); 8.06 (s, HÀC(2)); 9.74 (s, OH); 12.97 (s,
HOÀC(5)). 13C-NMR ((D6)acetone): see Table 1. EI-MS: 284 (Mþ ) [32].
Preparation of 11. In brief, 12 (568 mg) was heated under reflux in HBr (48%)/AcOH 1:1 (30 ml) for
5 h. After cooling, the product was poured into 120 ml of icy water and then extracted with AcOEt (2Â
100 ml). The extract was combined and washed with brine and H2O, dried (MgSO4), and then
recrystallized to obtain 11 after condensation.
5,7,3’-Trihydroxyisoflavone (11). Yield: 680 mg (84%). Yellowish needles (MeOH/H2O). M.p. 2538.
Rf (AcOEt/hexane/MeOH 4.5 :5 :0.5) 0.73. 1H-NMR ((D6)acetone): 6.29 (d, J ¼2.4, HÀC(6)); 6.42 (d,
J¼2.4, HÀC(8)); 6.87 (ddd, J¼1.2, 2.4, 8.4, HÀC(4’)); 7.04 (ddd, J¼0.8, 1.6, 7.6, HÀC(6’)); 7.12 (dd, J¼
2.0, 2.0, HÀC(2’)); 7.25 (t, J¼8.0, HÀC(5’)); 8.20 (s, HÀC(2)); 10.30 (s, OH); 12.99 (s, HOÀC(5)).
13C-NMR ((D6)acetone): see Table 1. EI-MS: 270 (Mþ ). HR-EI-MS: 270.0530 (Mþ, C15H10Oþ5 ; calc.
270.0528) [32].
Preparation of Prunetin (13). In brief, genistein (14; 540 mg, 2 mmol) and MeI (284 mg) were
dissolved in acetone (20 ml), and then stirred in the presence of K2CO3 (138 mg) for 2 h at ambient temp.
The reaction was terminated, and the mixture was concentrated in vacuo to remove acetone, 25 ml of 1n
HCl were added, and the resulting mixture was extracted with AcOEt (2Â30 ml), followed by washing
with H2O and drying. The crude product was subjected to CC (SiO2) and 13 was obtained by
recrystallization.
5,4’-Dihydroxy-7-methoxyisoflavone (¼ Prunetin; 13). Yield: 256 mg (45%). Colorless needles
(MeOH). M.p. 2458. Rf (AcOEt/hexane/MeOH 4 :5.5 :0.5) 0.64. 1H-NMR ((D5)pyridine): 3.73 (s,
MeO); 6.55 (d, J¼2.0, HÀC(6)); 6.61 (d, J¼2.0, HÀC(8)); 7.28 (d, J¼8.8, HÀC(3’,5’)); 7.71 (d, J¼8.8,
HÀC(2’,6’)); 8.18 (s, HÀC(2)); 11.86 (br., HOÀC(4’)); 13.55 (s, HOÀC(5)). 13C-NMR ((D5)pyridine): see
Table 1. EI-MS: 284 (Mþ ) [29].
.
.
þ
þ
Bioassays. ABTS Radical Scavenging Assay. The ABTS scavenging assays of isoflavones were
conducted according to the methods reported previously [19][33][34]. Briefly, ABTS was prepared with
deionized H2O in a concentration of 7 mm, it was then mixed with 2.45 mm K2S2O8. A mixture containing
180 ml of ABTS and 40 ml of isoflavone or the negative control (phosphate buffered saline (PBS)), or the
positive control (vitamin C) was well stirred, and then incubated for 7 min and 24 min at r.t. The
absorbance of the mixture was measured at 734 nm. The 50% radical-scavenging concentration is
expressed as SC50
.
.
.
DPPH Radical Scavenging Assay. In brief, 1 ml (0.1 mm) of DPPH soln. was mixed with 3 ml of
isoflavone solns. of various concentrations or vitamin C (positive control) and dissolved in MeOH. The
mixture was then stirred vigorously and kept at 408 for 30 min in the dark. For the baseline control, 3 ml
of MeOH were used. The absorbance was measured at 517 nm. The 50% radical scavenging
concentration is expressed in SC50 [19][33][34].
Mushroom Tyrosinase Inhibition Test. Antityrosinase activities were measured according to the
method reported previously [19][33][34]. In brief, isoflavones or arbutin (positive control) were
dissolved in DMSO/MeOH before dilution to various concentrations by potassium phosphate buffer
(pH 6.8). Twenty ml of isoflavone solns. of various concentrations or arbutin soln. were mixed with 20 ml
of mushroom tyrosinase (1000 U mlÀ1) and 80 ml of l-tyrosine (2.0 mm) in a 96-well plate, then the
mixture was diluted with buffer to a volume of 200 ml. After 30 and 60 min of incubation, the absorbance
was measured at 490 nm.
Impact on B16F10 Melanoma Cells and HaCaT Keratinocytes Viability. B16F10 melanocyte (BCRC
60031, Taiwan) or HaCaT keratinocyte cells were cultured in Dulbeccoꢀs modified Eagleꢀs medium
(DMEM; HyCloneꢂ, USA) with 10% fetal bovine serum (Sigma), 2 mm l-glutamine, 200 U mlÀ1
penicillin G, 200 mg mlÀ1 streptomycin, and 0.5 mg mlÀ1 amphotericin B (Kibbutz Beit Haemek, Israel)
under a humidified atmosphere containing 5% CO2 at 378. Cells were placed in 96-well plates (1·104
cells/well). After 24 h of incubation, test samples were added and the cultures were further incubated for
48 h. Tested concentrations of isoflavones were 10, 20, 40, and 80 mm for B16F10 murine melanoma cells,
and 5, 10, 20, and 40 mm for HaCaT human keratinocyte. Cell viability was determined using the XTT
assay. Formazan formed from XTT in cell cultures during incubation was measured at 490 nm using a
multi-well scanning spectrophotometer [34–36].

978
CHEMISTRY & BIODIVERSITY – Vol. 12 (2015)
Cellular Melanin Formation and Tyrosinase Activity. The measurement of cellular melanin formation
and tyrosinase activity was conducted according to the methods described by Chen et al. [37] and Lan
et al. [34]. In brief, B16F10 cells (5·104 cells/well) together with 50 nm a-MSH were placed in 24-well
plates and then incubated for 24 h. After addition of isoflavones, samples were further incubated for 48 h,
followed by washing twice with PBS. Melanocytes were dissolved in 100 ml of 2n NaOH and digested at
808 for 1 h to solubilize the generated melanin. Absorbance was measured at 405 nm, which was used to
determine the melanin content [34][37].
For cellular tyrosinase activity inhibitory assay, B16F10 cells were treated with isoflavones for 48 h as
described above. The harvested cells were washed twice with PBS, and then centrifuged at 12,000 rpm for
10 min, followed by addition of 1% Triton X-100 to lyse the cell membrane before the mixture was
subjected to further centrifugation. The supernatants were mixed with 100 ml of l-dopa (1 mg mlÀ1) at
378 for 3 h. The absorbance was measured at 490 nm with a multi-well spectrophotometer. All
experiments were performed for at least three times (nꢀ3), and the inhibition percentage of each
compound was compared with that of control [34][37].
REFERENCES
[1] M. Wlaschek, I. Tantcheva-Poór, L. Naderi, W. Ma, L. A. Schneider, Z. Razi-Wolf, J. Schüller, K.
Scharffetter-Kochanek, J. Photochem. Photobiol., B 2001, 63, 41.
[2] M. P. Lupo, Clin. Dermatol. 2001, 19, 467.
[3] X.-H. Gao, L. Zhang, H. Wei, H.-D. Chen, Clin. Dermatol. 2008, 26, 367.
[4] H. Epstein, Clin. Dermatol. 2009, 27, 475.
[5] T.-S. Chang, Int. J. Mol. Sci. 2009, 10, 2440.
[6] J. Arct, K. Pytkowska, Clin. Dermatol. 2008, 26, 347.
[7] E. Middleton Jr., C. Kandaswami, T. C. Theoharides, Pharmacol. Rev. 2000, 52, 673.
[8] Y. N. Wang, W. Wu, H. C. Chen, H. Fang, J. Dermatol. Sci. 2010, 58, 19.
[9] A. Russo, V. Cardile, L. Lombardo, L. Vanella, R. Acquaviva, J. Nutr. Biochem. 2006, 17, 103.
[10] T.-S. Chang, H.-Y. Ding, S. S.-K. Tai, C.-Y. Wu, Food Chem. 2007, 105, 1430.
[11] T.-S. Chang, H.-Y. Ding, H.-C. Lin, Biosci., Biotechnol., Biochem. 2005, 69, 1999.
[12] J.-S. Park, D. H. Kim, J. K. Lee, J. Y. Lee, D. H. Kim, H. K. Kim, H.-J. Lee, H. C. Kim, Bioorg. Med.
Chem. Lett. 2010, 20, 1162.
[13] S. S.-K. Tai, C.-G. Lin, M.-H. Wu, T.-S. Chang, Int. J. Mol. Sci. 2009, 10, 4257.
[14] T.-S. Chang, J. Agric. Food Chem. 2007, 55, 2010.
[15] J.-S. Park, H. Y. Park, D. H. Kim, D. H. Kim, H. K. Kim, Bioorg. Med. Chem. Lett. 2008, 18,
5006.
[16] J. H. Kim, M. R. Kim, E. S. Lee, C. H. Lee, Biol. Pharm. Bull. 2009, 32, 264.
[17] H. Goto, Y. Terao, S. Akai, Chem. Pharm. Bull. 2009, 57, 346.
[18] H. Esaki, S. Kawakishi, Y. Morimitsu, T. Osawa, Biosci., Biotechnol., Biochem. 1999, 63, 1637.
[19] L.-T. Ng, H.-H. Ko, T.-M. Lu, Bioorg. Med. Chem. 2009, 17, 4360.
[20] T.-M. Lu, D.-H. Kuo, H.-H. Ko, L.-T. Ng, J. Agric. Food Chem. 2010, 58, 10027.
[21] K. Wähälä, T. A. Hase, J. Chem. Soc., Perkin Trans. 1 1991, 3005.
[22] T.-C. Wang, I-L. Chen, C.-M. Lu, D.-H. Kuo, C.-H. Liao, Chem. Biodiversity 2005, 2, 253.
[23] R. K. S. Negi, T. R. Rajagopalan, V. Batra, Indian J. Chem., Sect. B 1985, 24, 221.
[24] P. -G. Pietta, J. Nat. Prod. 2000, 63, 1035.
[25] J. Nordenberg, I. Perlmutter, G. Lavie, E. Beery, O. Uziel, C. Morgenstern, E. Fenig, A. Weizman,
Int. J. Oncol. 2005, 27, 1097.
[26] B. J. Vilner, C. S. John, W. D. Bowen, Cancer Res. 1995, 55, 408.
[27] A. Mohammed, C. Nicholl, U. Titsch, M. Eisenhut, Nucl. Med. Biol. 1997, 24, 373.
[28] S. W. J. Wang, J. Chen, X. Jia, V. H. Tam, M. Hu, Drug Metab. Dispos. 2006, 34, 1837.
[29] H. C. Jha, F. Zilliken, E. Breitmaier, Can. J. Chem. 1980, 58, 1211.
[30] G. M. Basha, S. K. Yadav, R. Srinuvasarao, S. Prasanthi, T. Ramu, N. Mangarao, V. Siddaiah, Can. J.
Chem. 2013, 91, 763.

CHEMISTRY & BIODIVERSITY – Vol. 12 (2015)
979
[31] I. Dini, O. Schettino, A. Dini, J. Agric. Food Chem. 1998, 46, 5089.
[32] S. P. Bondarenko, A. V. Levenets, M. S. Frasinyuk, V. P. Khilya, Chem. Nat. Compd. 2003, 39,
271.
[33] H.-H. Ko, Y.-J. Jin, T.-M. Lu, I.-S. Chen, Chem. Biodiversity 2013, 10, 1269.
[34] W.-C. Lan, C.-W. Tzeng, C.-C. Lin, F.-L. Yen, H.-H. Ko, Phytochemistry 2013, 89, 78.
[35] N. W. Roehm, G. H. Rodgers, S. M. Hatfield, A. L. Glasebrook, J. Immunol. Methods 1991, 142,
257.
[36] C. J. Goodwin, S. J. Holt, S. Downes, N. J. Marshall, J. Immunol. Methods 1995, 179, 95.
[37] L.-G. Chen, W.-L. Chang, C.-J. Lee, L.-T. Lee, C.-M. Shih, C.-C. Wang, Biol. Pharm. Bull. 2009, 32,
1447.
Received May 31, 2014