The synthesis of morusin as a potent antitumor agent

Article abstract of DOI:10.1016/j.tet.2009.12.002

Morusin, which can be isolated from Chinese herbal medicine, is achieved in which the longest linear sequence is only 13 steps in 12% overall yield from commercially available phloroglucinol. In addition, the metal/EtSH reagents for regioselective demethy

Full text of DOI:10.1016/j.tet.2009.12.002

Tetrahedron 66 (2010) 1335–1340
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
The synthesis of morusin as a potent antitumor agent
,
a
a
b
b
Tsui-Hwa Tseng a , Shien-Kai Chuang , Chao-Chin Hu , Chia-Fu Chang , Yu-Chao Huang ,
*
Cheng-Wei Lin b, Yean-Jang Lee b
,
*
a School of Applied Chemistry, Chung Shan Medical University, No. 110, Section 1, Chien-Kuo N. Road, Taichung 402, Taiwan
b Department of Chemistry, National Changhua University of Education,1 Ching Der Road, Changhua 50058, Taiwan
a r t i c l e i n f o
a b s t r a c t
Article history:
Morusin, which can be isolated from Chinese herbal medicine, is achieved in which the longest linear
sequence is only 13 steps in 12% overall yield from commercially available phloroglucinol. In addition, the
metal/EtSH reagents for regioselective demethylation of polymethylated morusin were described. Con-
sequently, this strategy provided a concise route to synthesize the morusin analogues as well. Further
Received 11 September 2009
Received in revised form 2 December 2009
Accepted 2 December 2009
Available online 11 December 2009
biological studies of morusin, it exhibited strong antitumor effects with IC50 w3.0 mM on three cancer
cells.
Keywords:
Morusin
EtSLi
Ó 2009 Elsevier Ltd. All rights reserved.
SnCl2
1. Introduction
including colon, prostate, and breast and lung cancers and in me-
sothelioma.6 Additionally, mammalian 5-LOX is particularly im-
portant because of its unique ability to convert AA to leukotrienes
that act as potent mediators of inflammation, apoptosis, and tu-
morigenesis.7 Since the LOX pathway appears to be particularly
implicated in tumor growth,8 the development of selective in-
hibitors of arachidonate 5-lipoxygenase has been a subject of me-
dicinal chemical interest. Recently several studies have reported
the various biological effects of hop prenylflavonoids in vitro.8 In
addition, for investigating the effects of morusin and himanimides9
(Fig. 1) on COXs and LOXs, five known compounds, such as NS-398
(COX-2 inhibitor), indomethacin (a COX-1/COX-2 inhibitor), NDGA
(an LOX inhibitor), celecoxib (a COX-2 inhibitor), and zileuton
(a 5-LOX inhibitor), were used as the inhibitors10 of enzyme ref-
erences (Fig. 2). With the prenyl structural relative among other
flavonoid natural products, morusin4 occupies an unique position
in the realm of flavonoid chemistry. Nonetheless, despite the
structural novelty and pronounced biological activity of the mor-
usin, its total synthesis has not previously been recorded. Herein,
we undertook the total synthesis of the morusin and study its bi-
ological activities.
Our research group has exhibited a long-standing interest in the
synthesis and bioassay study of a number of unique natural prod-
ucts containing flavonoids, camphorataimides, and benzofurans.1 In
previous works, we have shown that tumor suppressor protein p53
and p38 MAPK (mitogen activated protein kinase) play a prominent
role in the caffeic acid phenethyl ester (CAPE) induced-apoptosis of
C6 glioma cells.2 Accordingly, Barron et al. reported that a library of
42 natural and synthetic flavonoids has been screened for their
effect on cell proliferation and apoptosis in a human colonic cell line
(HT29).3 In addition, studies on the constituents of the Chinese
crude drug ‘Sang-bai-pi’ (Morus root bark) have been recorded in
books of traditional Chinese herbal medicines, which are used as an
antiphlogistic, a diuretic, an expectorant, and anti-hepatitis B ac-
tivity.4 Recently, Kim et al.5 reported that the effects of 19 naturally
occurring prenylated flavonoids on cyclooxygenase (COX)-1 and
COX-2 and on 5-lipoxygenase (5-LOX) and 12-LOX were in-
vestigated. Generally, the inhibitory activity of prenylated flavo-
noids against 5-LOX was much stronger than against 12-LOX.
Arachidonic acid (AA) is released from membrane phospho-
lipids during normal and pathological processes and is metabolized
primarily by the COX, LOX and cytochrome P450 (CYP450) enzyme
system. The arachidonic acid metabolizing enzymes have been
implicated in the development of a variety of human cancers
OH
O
O
O
N
OH
N
OH
O
O
O
HO
Morusin (1)
OH
O
O
O
* Corresponding authors. Tel.: þ886 4 7232105x3511; fax: þ886 4 7211190.
E-mail addresses: tht@csmu.edu.tw (T.-H. Tseng), leeyj@cc.ncue.edu.tw (Y.-J.
Lee).
Himanimide D
Himanimide C
Figure 1. Structures of the naturally occurring products.
0040-4020/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2009.12.002

1336
T.-H. Tseng et al. / Tetrahedron 66 (2010) 1335–1340
OMe O
OH
OH
OMe O
O
OMe
O
1. LDA
Br
CO2H
MeO
NHSO2CH3
HO
HO
LDA
94%
Cl
+
O
N
O
OH
O
OH
OMe
2. H2SO4/AcOH
76%
MeO
OMe
O
Nordihydroguaiaretic acid (NDGA)
6
5
4
NO2
NS398
Cl
OMe O
O
OMe O
O
OMe
Indomethacin
O
EtSH/n-BuLi
H2NO2S
O
O
HMPA, 70 o
72%
C
N
OH
O
OH
OMe
HO
NH2
O
N
N
N
CF3
MeO
OMe
HO
N
7
O
8
S
O
OH
O
O
Zileuton
OH
O
O
1. PMBBr, acetone
K2CO3, reflux
LEE-BNOHDM
pyridine
1. AcCl,
92%
Celecoxib
89%
O
O
Figure 2. Structural features of LOXs and COXs inhibitors.
2. EtSH/n-BuLi
2. SnCl2, EtSH
CH3CN, then
OR2
R1O
9a R1=H,R2=Me
9b R1=Me,R2=H
OR2
HMPA, 70 o
73%
C
R1O
10a R1=PMB,R2=H
H2NNH2-H2O
67%
10b R1=H,R2=PMB
2. Results and discussion
OH
O
Retrosynthetic analysis suggested two possible routes for the
synthesis of this natural product: routes a and b as shown in
Scheme 1. Two approaches will be pursued to determine, which
strategy is viable. The first approach involves the Friedel–Crafts
reaction of coumarin 2 and 3 catalyzed by Lewis acid or base to
provide the precursor chalcone, followed by subsequent cyclization
and C-3 alkylation of flavone with allyl bromide to afford the de-
sired 1. The other way employs the enolate acylation11 of the
known12 2-hydroxyl acetophenone 4 and benzoyl chloride 5 to
obtain the key intermediate diketone 6, followed by subsequent
alkylation and cyclization to achieve the target molecule 1.
O
O
HO
OH
Scheme 2. Total synthesis of the morusin.
1
With known12 precursors 4 and 5 in hands, followed by the
procedure of Banerji and Luthriathe underwent the Baker–Ven-
kataraman rearrangement to give diketo 6 in high yield (94%). The
alkylation of 6 with allyl bromide provided the prenylated in-
termediate 7, and followed by cyclization with H2SO4/AcOH to
generate the expected methylated morusin 8 in 76% yield over two
steps. At this point, the final step of the synthesis was focused on
the demethylation of 8, however, attempts to treat with Me3SiI,
AlCl3, BBr3, BBr3/NaI, SiCl4/NaI, pyridinium hydrohalides, or EtSLi to
give morusin (1) were unsuccessful. The major product was
5-hydroxyl-20,40-dimethoxy morusin, the product from forming
more stable with hydrogen bond on 5-demethylation of 8. Con-
sequently, we sought to manipulate the deprotection of 8 in
a separate step, to our delight, 8 was undergone smoothly with
EtSLi/HMPA to obtain in 72% yield the desired isomer 9. Continu-
ously, in the case of 9, we attributed our failure to achieve 1 due to
the free 2-or 4-hydroxy group interfered with EtSLi/HMPA. This
problem was overcome by protection of 9 with p-methoxybenzyl
bromide (PMBBr) and subsequent treatment with EtSLi/HMPA to
give 10. It is striking that cleaving PMB group of 10 to 1 could not
be achieved despite our efforts in methods either by oxidizing
agents or Lewis acids for cleaving PMB with 1,2-dichloro-4,5-
dicyanoquinone (DDQ),13 CeCl3–7H2O/NaI,14 cerium(IV) ammo-
nium nitrate (CAN),15 ZrCl4,16 BF3–Et2O/NaBH3CN.17 In all case, they
were proceeded with decomposed and recovered starting material
10. Obviously, the target morusin was unstable under acidic and
oxidizing conditions, but it can make a detour by the first pro-
tection with AcCl and subsequent deprotection with mild SnCl2/
EtSH18 and NH2NH2 to provide 1 in moderate yield 62%. It is worth
noting that the reported method is the best among the well-
established methods by comparing thoroughly with other
methods. In addition, we also failed to achieve 1 by the use of other
starting materials 5 with easily removal protecting groups, such as
isopropyl, MOM, benzyl, acetyl, and silyl groups instead of methyl
group of 5.
OH
O
O
OH
O
OH
Br
O
Cyclization
Alkylation
O
OH
a.
OH
HO
OH
Friedel-Crafts
OH
Morusin (1)
Alkylation
Cyclization
Br
+
OMe O
O
OMe
O
OH
HO
O
O
2
3
O
OH
OMe
6
b.
OMe O
O
OMe
Cl
+
O
OH
OMe
5
4
Scheme 1. Retrosynthetic analysis of morusin.
The initial proposal was to use the phenolic 3 and the com-
mercially available 7-hydroxycoumarin 2 via the Friedel–Crafts
reaction for Morusin synthesis. Although the chromene 3 seems
a simple compound, but it has not been reported in the literature.
We started with the known12 5,7-dihydroxy-2,2-dimethyl chro-
man-4-one by reduction and dehydration with NaBH4 and 4 M HCl
to generate the desired 3 in 93% yield. In the case of 2 and 3,
however, the Friedel–Crafts reaction catalyzed with AlCl3, AlBr3,
ZnCl2, SnCl4, BF3OEt2, pyridine, piperidine, pyrrolidine, and KOH to
give the precursor chalcone or flavone was unsuccessful. To over-
come these technical difficulties, we switched to using the simple
acylation of ketone 4 and acyl chloride 5 with LDA (Scheme 2).
To probe accessibility of the active sites of P38 MAPK, c-Met
tyrosin kinase and neurokinin (NK) receptors of hepatocyte
growth factor/scatter factor (HGF/SF), and COXs/LOXs, the binding
affinity of protein–ligand complexes were first examined by
automated docking method (Table 1). The crystal structures of
proteins were from the Brookhaven Protein Data Bank (PDB)

T.-H. Tseng et al. / Tetrahedron 66 (2010) 1335–1340
1337
(http://www.rcsb.org/pdb/), for instance, 5-lipoxygenase entry
code (2Q7M). From the presented docking, CAPE and himanimides
showed significant binding on P38 kinase, c-Met and NK1 re-
ceptors, and COX-1, however, morusin gave preference to the tight
bound of COX-2 and 5-lipoxygenase. Moreover, it is worth noted
that morusin can show more specific inhibition of 5-LOX than that
of COX-2 because all 10 docking poses are clustered within 1.1 Å
(rmsd), indicating a strong consensus for a single binding mode.
The lowest-energy pose are shown in Figure 3A. The active site for
docking was defined with the amino acids Leu135, Phe131,
Ala128, Val127, and Leu124 of both B and F chains in the 5-LOX
falling within 7 Å radius of the cocrystallized ligand. The 20- and
40-hydroxy groups of morusin formed hydrogen bonds to Val127
of both B and F (2.95 and 2.09 Å), respectively. Additional ana-
lytical information, the isopropyl group of the residue Val127
(B chain) was positioned to make hydrophobic contact with
3-prenyl group of morusin, and the tricyclic core of morusin was
aligned into the groove of the residues Ala128, Val127, Leu124,
Leu135, and Phe131 for active site recognition (Fig. 3B).
higher than that of the original inhibitor (NDGA¼115
m
M). Thus,
these results imply that the presence of prenylated flavonoids and
N-hydroxyl imides assist in binding with 5-LOX. Consistently, their
inhibitory effect of CAPE, himanimide C, and morusin on 5-LOX in
vitro was also examined with an IC50 of 50.5, 26.8, and 8.7 mM, re-
spectively. Finally, the combination of hydrogen bond and hydro-
phobic pocket was validated with the designed three C2 symmetric
dimmers, and LEE-BNOHDM provided Ki value (263.98 nM) for the
strongest inhibitor of 5-LOX (Fig. 3C). The present work demon-
strates the ability of prenylated flavonoid chemistry to quickly
generate a novel series of arachidonate 5-LOX inhibitors, showing
features of high potency and selectivity, as well as ability to inhibit
inducible NO synthase (iNOS) expression in RAW 264.7 cells.20
Additionally, Lin et al. reported21 that morusin inhibited platelet
aggregation induced by platelet-activating factor (PAF), which did
not cause thromboxane formation in rabbit platelet. The combina-
tion of docking and Lin’s results enabled us to propose that morusin
may become a lead candidacy as therapeutic implications for ath-
erosclerosis and vascular diseases.
Table 1
Compound Ki values and the lowest binding energies from AutoDock Lamarckian Genetic Algorithm
Ligand
Protein(PDBentry) Binding
Energy (Kcal/mol)/Ki(
m
M)
COX-1
(2OYE)
COX-2
(1DDX)
5-LOX
(2Q7M)
P38 MAPK
(1P38)
NK1/HGF
(1NK1)
c-Met/HGF
(2RFN)
CAPE
Himanimide C
Morusin
ꢀ7.36/4.06
ꢀ7.64/2.51
38.44/NA
ꢀ4.35/652.0
ꢀ4.57/445.4
ꢀ5.24/144.41
ꢀ5.88/48.95
ꢀ4.83/286.68
ꢀ5.15/166.52
ꢀ7.37/3.95
ꢀ7.56/2.86
ꢀ7.91/1.6
0.57/NA
ꢀ4.22/813.4
ꢀ7.06/6.7
ꢀ5.86/50.61
ꢀ6.19/28.94
ꢀ5.74/61.67
ꢀ5.44/102.65
ꢀ7.91/1.59
ꢀ3.69/1970
ꢀ7.33/4.26
ꢀ6.76/11.08
ꢀ6.44/19.01
Himanimide D
NA:Unavailable.
Figure 3. Binding orientation of protein-ligand predicted by the AutoDock Lamarckian Genetic Algorithm on PyMOL viewer (A) LOX-5 (PDB entry 2Q7M)/morusin is shown as a line
model, (B) LOX-5/morusin is shown as a surface model, and (C) LOX-5/LEE-BNOHDM is shown as a surface model.
Once the docking protocol was confirmed, the other protein–
ligand complexes under analysis were docked using the same
strategy. In addition, it is well known that the 5-LOX isoenzyme play
crucial roles in the development of human cancers.19 Therefore, the
influence of prenylated flavonoids on the binding affinity of 5-LOX
for thirty-six compounds and five reference inhibitors (COXs/LOXs)
was assessed to further test the docking models. The Ki values for
compounds, 1, himanimide C and D that were estimated by auto-
To further test cell line-dependent cytotoxicity, the effect of
morusin on tumor cell proliferation/viability was examined in
MDA-MB-231 and MCF 7 breast tumor cells and in A549 lung
cancer cell (Table 2). Following 24 h incubation, morusin inhibited
MDA-MB-231 and MCF 7 cell proliferation/viability with an IC50 of
3.2 and 3.4
mM, respectively, and inhibited A549 cells with an IC50
of 3.1 M. In addition, in comparison with results of CAPE and
m
camphorataimides has clearly shown the cytotoxicity of morusin
on above three cancer cell lines. Therefore, it is worthy to further
investigate the chemopreventative or antitumorigenesis mecha-
nisms of morusin in those cancer cell lines.
mated docking on 5-LOX were 11.0, 3.95, and 19.0 mM, respectively.
In addition, the Ki values of compound 1, himanimide C and D that
were evaluated were found to range from 30–fold lower to >6–fold
Table 2
Cytotoxicity was estimated by MTT assay (24 h). The comparative cytotoxicity is expressed the 50% half maximal concentration (IC50 mM) for Morusin and camphorataimides9
Camphorataanhydride A
Camphorataimide B
Camphorataimide C
Himanimide A
Morusin
MDA-MB-231a
MCF-7a
>200
>200
>200
11.4
10.6
15.4
51.6
33.9
60.3
19.6
21.2
30.3
3.2
3.4
3.1
A549b
a
MDA-MB-231 and MCF-7 are breast cancer cell lines.
A549 is lung cancer cell.
b

1338
T.-H. Tseng et al. / Tetrahedron 66 (2010) 1335–1340
3. Conclusion
ether:hexane 1:2) to provide the Fries product (4.8 g, 85%) as
a white solid: 95–96 ꢂC (lit.23 92–93 ꢂC). To a mixture of the Fries
product (4.8 g, 19 mmol) and DDQ (4.9 g, 22 mmol) in acetic acid
(50.0 mL) was heated to 120 ꢂC for 4 h, and then the suspension
was filtered. The filtrate was directly subjected to column chro-
matography (SiO2, ether:hexane 1:4) to obtain the desired 4 (3.9 g,
81%) as a yellow solid: mp 125–126 ꢂC (lit.24 128–129 ꢂC), IR (KBr)
In summary, the first total synthesis route to morusin has been
achieved in which the longest linear sequence is only 13 steps in
12% overall yield from commercially available phloroglucinol. In
addition, it demonstrated the usefulness of the metal/EtSH re-
agents for regioselective demethylation of polymethylated mor-
usin. Furthermore, it demonstrated the usefulness of the
automated docking method for a drug lead discovery. Conse-
quently, the rationally optimized and designed LEE-BNOHDM was
identified as a potentially important inhibitor of the arachidonate
5-lipoxygenase. Moreover, morusin exhibited strong antitumor
effects in three cancer cells. Finally, the synthetic morusin is cur-
rently underway to evaluate the efficacy as a 5-LOX inhibitor or an
antitumor agent, and its biological activity will further be in-
vestigated in vitro/vivo as well.
nmax: 3450, 2973, 1634, 1600, 1265, 1124, 850 cmꢀ1
.
4.3. 1-(2,4-Dimethoxyphenyl)-3-(5-hydroxy-7-methoxy-2,2-
dimethylchromen-6-yl)-propane-1,3-dione (6)
This compound was prepared as a yellow solid from 4 and 2,4-
dimethoxybenzoyl chloride (5) according to the procedure of
Banerji and Luthria11 mp 107–108 ꢂC (lit.11 mp 107–110 ꢂC), IR (KBr)
nmax: 2978, 2938, 1634, 1600, 1265, 1124, 800 cmꢀ1
.
4. Experimental
4.4. 2-(2,4-Dimethoxyphenyl)-5-methoxy-8,8-dimethyl-3-(3-
methyl-but-2-enyl)-pyrano[2,3-f]chromen-4-one (8)25
4.1. General experimental procedures
Melting points were determined on a Mel Temp II melting point
apparatus and are uncorrected. IR spectra recorded with an Equi-
nox 55 (FTIR) spectrometer. 1H NMR and 13C NMR spectra were
obtained using a Varian-300 spectrometer. Chemical shifts are
To a solution of 6 (1.0 g, 2.4 mmol) in anhydrous THF (50.0 mL)
was cooled down to -78 ꢂC, and then added LDA (2.0 M 1.3 mL,
2.6 mmol) dropwise at that temperature for stirring 0.5 h. The
mixture was warmed up to ꢀ25 ꢂC for 1 h and cooled down again to
reported in parts per million (d, ppm) using CHCl3 (dH 7.26) as an
ꢀ78 ꢂC, and followed by addition of allyl bromide (290
mL,
internal standard. Low-resolution mass spectra (MS) and high-
resolution mass spectra (HRMS) were determined on a JEOL JMS-
HX 110 mass spectrometer from National Chung-Tsing University,
Taichung. Solvents were freshly distilled prior to use from phos-
phorus pentoxide or CaH2. THF was from sodium diphenyl ketyl. All
reactions were carried out under nitrogen atmosphere unless
otherwise stated. Silica gel (Silica gel 60, 230–400 mesh, Merck)
was used for chromatography. Organic extracts were dried over
anhydrous MgSO4.
2.5 mmol) dropwise for 1 h. The resulting solution was quenched
by addition of 1 N HCl (1.0 mL), and extracted with ether
(3ꢁ200 mL). The organic layer was concentrated in vacuo, and the
crude product 7 was directly used no further purification. A solu-
tion of 7 (0.98 g, 0.2 mmol) in acetic acid (15.0 mL) was added conc.
H2SO4 (0.02 mL) and stirred at room temperature for 15 min. The
resulting mixture was added dist. H2O (15.0 mL) and extracted with
CH2Cl2 (3ꢁ200 mL). The solvent was removed and the residue was
subjected to column chromatography (SiO2, ether:hexane 1:1) to
afford the desired 8 (0.85 g, 76%) as a white solid: mp 153–154 ꢂC
(lit.26 mp 152–154 ꢂC), IR (KBr) nmax: 3050, 2973, 2928, 2851, 1639,
1280, 1164, 850 cmꢀ1, HRMS (FAB) calcd for C28H31O6 {[MþH]þ}
463.2121, found 463.2126.
4.2. 6-Acetyl-5-hydroxy-7-methoxy-2,2-dimethylchrom-
3-ene (4)
The 5-hydroxy-7-methoxy-2,2-dimethylchroman-4-one was
prepared as previous procedure.12 The ketone (5.8 g, 26 mmol) was
dissolved in MeOH (200.0 mL), and added Zn (5.1 g, 78 mmol) in
one portion at room temperature for 15 min. The suspension was
followed by addition of conc. HCl (5 mL) dropwise and stirred at
room temperature for 1 h. The resulting mixture was filtered and
the filtrate was extracted with H2O (150.0 mL) and ethyl acetate
(3ꢁ400 mL). The organic layer was concentrated in vacuo, and then
subjected to column chromatography (Silica gel ether:hexane 1:1)
to obtain the desired chromane (5.0 g, 93%) as a yellow solid: mp
74–75 ꢂC (lit.22 78–79 ꢂC). To a solution of chromane (5.0 g,
24 mmol) in CH2Cl2 (50 mL) was added pyridine (2.0 mL, 25 mmol)
at room temperature, and then treated with AcCl (2.0 mL, 28 mmol)
dropwise. The mixture was stirred at room temperature for 0.5 h.
The resulting suspension was filtered and washed with CH2Cl2, and
the filtrate was subjected to column chromatography (SiO2,
CH2Cl2:hexane 1:2) to give the expected acetoxy chromane (5.7 g,
4.5. 2-(2-Hydroxy-4-methoxyphenyl)-5-hydroxy-8,8-
dimethyl-3-(3-methyl-but-2-enyl)-pyrano[2,3-f]chromen-
4-one (9a)27
A solution of EtSH (0.8 mL, 10.8 mmol) in HMPA (6.0 mL) was
cooled down to 0 ꢂC for 10 min, and then added n-BuLi (2 M,
10.8 mmol) under N2 at that temperature for stirring 30 min. Sub-
sequently, 8 (1.20 g, 2.4 mmol) was introduced a fresh solution of
EtSLi in HMPA, and the resulting solution was warmed at 70 ꢂC
under N2. After stirring at 70 ꢂC for 8 h, the reaction mixture was
cooled and quenched with a saturated solution of NH4Cl (4.0 mL)
and extracted with ethyl acetate (3ꢁ20.0 mL). The combined or-
ganic extracts were washed with saturated aqueous LiCl, dried over
Na2SO4, and concentrated in vacuo. The residue was purified by
flash column chromatography (SiO2, hexane:ethyl acetate 3:1) to
provide 9a (0.68 g, 60%) and 9b (135 mg,12%); 9a as yellow powder:
mp 162–164 ꢂC (lit.27 mp 162–164 ꢂC), IR (KBr) nmax: 3400, 2986,
96%) as an oil. 1H NMR
d
1.30 (s, 6H), 1.72 and 2.47 (t, J¼6.8 Hz, 4H),
2.24 (s, 3H), 3.69 (s, 3H), 6.20 and 6.26 (d, J¼2.5 Hz, 2H); 13C NMR
d
16.7, 20.4, 26.3, 31.8, 55.0, 74.2, 99.6, 100.3, 106.3, 149.6, 155.2,
2925, 1654, 1620, 1580, 1265, 1184, 850 cmꢀ1 1H NMR [(CD3)2CO]
,
158.7, 168.6.
A mixture of the acetoxy chromane (5.7 g, 23 mmol) and AlCl3
d
1.41 (s, 3H), 1.42 (s, 6H), 1.55 (s, 3H), 3.10 (d, J¼3.3 Hz, 2H), 3.82 (s,
3H), 5.09 (t, J¼3.3 Hz,1H), 5.63 (d, J¼5.4 Hz,1H), 6.14 (s,1H), 6.57 (d,
J¼5.1 Hz, 1H), 6.60 (d, J¼5.1 Hz, 1H), 6.61 (s, 1H), 7.33 (d, J¼5.4 Hz,
(9.1 g, 68 mmol) in anhydrous dioxane (50.0 mL) was heated to
120 ꢂC for 5 h. The solvent was removed and neutralized with
NaHCO3 (80.0 mL), then followed by extraction with CH2Cl2
(3ꢁ100 mL), and the organic layer was concentrated in vacuo. The
residue was subjected to column chromatography (Silica gel
1H), 9.00 (br s, 1H), 12.79 (br s, 1H); 13C NMR
d 17.7, 24.3, 25.6, 28.2,
28.2, 55.5, 78.0, 99.9, 100.9, 102.1, 104.8, 107.2, 112.2, 114.7, 120.8,
121.3, 127.1, 131.3, 133.3, 152.0, 155.1, 159.3, 159.4, 161.4, 162.8, 182.0,
HRMS (EI) calcd for C26H26O6 (Mþ) 434.1729, found 434.1728.

T.-H. Tseng et al. / Tetrahedron 66 (2010) 1335–1340
1339
4.6. 2-(4-Hydroxy-2-methoxyphenyl)-5-hydroxy-
4.9. 2-(2,4-Dihydroxyphenyl)-5-hydroxy-8,8-dimethyl-3-
(3-methyl-but-2-enyl)-pyrano[2,3-f]chromen-4-one (1)28
8,8-dimethyl-3-(3-methyl-but-2-enyl)-pyrano[2,3-f]chromen-
4-one (9b)27
To a solution of 10a (0.30 g, 0.6 mmol) in pyridine (4.0 mL) was
added acetyl chloride (39 L) with dropwise at room temperature,
Compound 9b as yellow powder: mp 195–197 ꢂC (lit.27 mp 198–
199 ꢂC), IR (KBr) nmax: 3400, 2975, 2928, 1650, 1620, 1580, 1265,
m
and followed by stirring for 1 h. The resulting reaction mixture was
quenched with H2O (0.2 mL) and extracted with CH2Cl2 (3ꢁ5.0 mL).
The combined organic extracts were dried over Na2SO4, and con-
centrated in vacuo. The residue was purified by flash column
chromatography (SiO2, hexane:ethyl acetate 3:1) to obtain the
desired product (0.30 g, 92%) as an oil. Subsequently, a mixture of
acetoxy product and SnCl2 (0.11 g, 0.6 mmol) in acetonitrile
(5.0 mL) was added EtSH (0.2 mL) at room temperature under N2,
and monitored by TLC. The solvent was removed and the residue
was subjected to column chromatography (SiO2, ether:hexane 1:1)
to give the demethoxybenzyl acetoxy product isomers, which was
dissolved in THF (5.0 mL), and followed by deacetylation with
1184, 850 cmꢀ1, 1H NMR [(CD3)2CO]
d 1.38 (s, 3H), 1.42 (s, 6H), 1.55
(s, 3H), 3.02 (d, J¼3.3 Hz, 2H), 3.80 (s, 3H), 5.06 (t, J¼3.3 Hz, 1H),
5.62 (d, J¼5.1 Hz, 1H), 6.14 (s, 1H), 6.54 (d, J¼5.1 Hz, 1H), 6.58 (dd,
J¼5.1, 1.2 Hz, 1H), 6.64 (d, J¼1.2 Hz, 1H), 7.27 (d, J¼5.1 Hz, 1H), 9.09
(br s, 1H), 13.2 (br s, 1H); 13C NMR [(CD3)2CO]
d 16.7, 23.7, 24.9, 27.3,
27.3, 55.1, 77.8, 98.9, 99.2, 100.7, 104.6, 107.3, 113.1, 114.4, 120.8,
121.5, 127.2, 131.2, 131.4, 152.3, 158.7, 159.1, 161.0, 161.4, 161.8, 182.2,
HRMS (EI) calcd for C26H26O6 (Mþ) 434.1729, found 434.1723.
4.7. 2-[2-(4-Methoxybenzyloxy)-4-hydroxyphenyl]-
5-hydroxy-8,8-dimethyl-3-(3-methyl-but-2-enyl)-pyrano
[2,3-f]chromen-4-one (10a)
NH2NH2–H2O (29 mL, 0.6 mmol) at room temperature. The mixture
was stirred for 20 min, and the resulting solution was diluted with
CH2Cl2 and filtered by silica gel. The organic solvent was concen-
trated in vacuo, and the residue was purified by flash column
chromatography (SiO2, ether:CH2Cl2 1:20) to achieve 1 (156 mg,
67%) as light yellow solid: mp 149–150 ꢂC (ether/hexane) (lit.29
147–149 ꢂC), IR (KBr) nmax: 3400, 2978, 2928, 1648, 1620, 1580,
The mixture of 9a (2.10 g, 4.8 mmol) and K2CO3 (3.00 g,
21.7 mmol) in acetone (50.0 mL) was added PMBBr (0.5 mL,
5.3 mmol) at room temperature, and then refluxed at 70 ꢂC for 1 h by
TLC monitoring. After cooling, the solution was evaporated in vacuo,
and the brown residue was subjected to flash chromatography (SiO2,
hexane:ethyl acetate 4:1) togive the protected product (2.39 g, 89%).
Subsequently, the protected product was treated by adding a solu-
tion of EtSLi in HMPA (6.0 mL) at room temperature, and the
resulting mixture was heated at 70 ꢂC under N2. After stirring at
70 ꢂC for 2 h, the reaction mixture was cooled and quenched with
a saturated solution of NH4Cl (4.0 mL) and extracted with ethyl ac-
etate (3ꢁ20 mL). The combined organic extracts were washed with
saturated aqueous LiCl, dried over Na2SO4, and concentrated in
vacuo. The residue was purified by flash column chromatography
(SiO2, hexane:ethyl acetate 3:1) to afford 10a (1.70 g, 73%) as an oil:
IR (KBr) nmax: 3400, 2972, 2932, 1654, 1620, 1580, 1265, 1184, 850,
1265, 1184, 790 cmꢀ1, 1H NMR [(CD3)2CO]
d 1.42 (s, 9H), 1.55 (s, 3H),
3.10 (d, J¼6.9 Hz, 2H), 4.55 (br s,1H), 4.58 (br s,1H), 5.11 (t, J¼6.9 Hz,
1H), 5.63 (d, J¼8.1 Hz, 1H), 6.13 (s, 1H), 6.50 (dd, J¼9.6, 2.4 Hz, 1H),
6.55 (d, J¼2.4 Hz, 1H), 6.58 (d, J¼9.6 Hz, 1H), 7.23 (d, J¼8.1 Hz, 1H),
13.2 (br s, 1H); 13C NMR
d 17.6, 24.2, 25.6, 28.1, 28.1, 78.1, 99.8, 101.0,
103.8, 104.9, 108.3, 112.3, 114.7, 120.8, 121.3, 127.1, 131.6, 133.2, 152.1,
155.2, 159.2, 159.3, 159.9, 161.2, 182.3, HRMS (EI) calcd for C25H24O6
(Mþ) 420.1573, found 420.1566.
4.10. Computational methods. Preparation of ligand and
protein complexes input structures
780 cmꢀ1, 1H NMR
d 1.41 (s, 3H), 1.46 (s, 6H), 1.59 (s, 3H), 3.09 (d,
J¼5.4 Hz, 2H), 3.75 (s, 3H), 4.99 (s, 2H), 5.11 (t, J¼5.4 Hz,1H), 5.48 (d,
J¼8.1 Hz, 1H), 6.26 (s, 1H), 6.53 (dd, J¼9.9, 2.1 Hz, 1H), 6.60 (d,
J¼9.9 Hz, 1H), 6.63 (d, J¼2.1 Hz, 1H), 6.80 (d, J¼8.4 Hz, 2H), 7.20 (d,
J¼8.4 Hz, 2H), 7.21 (d, J¼8.1 Hz,1H),13.19 (br s,1H); 13C NMR (CDCl3)
First, protein structures were extracted from the Protein Data-
bank (pdb) file (containing only non-hydrogen atoms). The pro-
tein’s pdb files were edited with AutoDockTools-1.5.1 (on Linux
system) to protonate polar hydrogen atoms, add Kollman charges,
and save as a pdbqt file. Furthermore, ligand structures were
optimized by Chem 3D Ultra version 8.0 (Run MOPAC minimize
energy to minimum rms Gradient 0.001) to get the structure (pdb
file). Alternatively, the X-ray structure of morusin was extracted
from Cambridge Chemical Database (pdb file). The ligand file was
input into AutoDockTools-1.5.1 program, detected the root of
molecule, and output as a pdbqt file as well.
d
17.6, 24.3, 25.6, 28.1, 28.1, 55.2, 70.2, 77.9, 99.6, 100.9, 100.9, 105.1,
107.6, 113.9, 114.6, 115.0, 121.1, 121.4, 126.7, 128.2, 128.6, 131.6, 132.3,
152.4, 157.7, 159.2, 159.3, 161.2, 161.2, 161.4, 182.6, 13C NMR
[(CD3)2CO]
d 16.8, 23.8, 24.9, 27.4, 27.4, 54.5, 69.8, 77.8, 98.9, 100.6,
100.8, 104.6, 107.5, 113.5, 113.7, 114.5, 120.7, 121.6, 127.1, 128.6, 128.9,
131.4,131.5,152.2,157.9, 159.1,159.5, 160.9, 161.6,161.8,182.2, HRMS
(EI) calcd for C33H32O7 (Mþ) 540.2148, found 540.2158.
4.8. 2-[2-Hydroxy-4-(4-methoxybenzyloxy)phenyl]-
5-hydroxy-8,8-dimethyl-3-(3-methyl-but-2-enyl)-pyrano-
[2,3-f]chromen-4-one (10b)
4.10.1. Grid and docking studies. To predict the energetically fa-
vorable positions of ligand molecules in the interior structure of the
protein target, a rectangular grid box of 21.75ꢁ21.75ꢁ21.75 Å3 with
grid points separated by 0.333 Å, centered on the midpoint of the
ligand binding pocket. It was essential to modify the AutoGrid
program to enable docking in the presence of metals and hetero-
atoms. Subsequently, Automated docking studies were performed
with Lamarckian Genetic Algorithm to provide the binding energies
and conformations of protein–ligand complexes.
This compound was prepared as light yellow oil according to the
previous procedures. IR (KBr) nmax: 3400, 2975, 2924, 1650, 1620,
1580,1265,1184, 850,780 cmꢀ1,1HNMR
d1.36(s, 3H),1.41(s, 6H),1.60
(s, 3H), 3.11 (d, J¼6.6 Hz,2H), 3.81(s, 3H), 5.01(s, 2H), 5.08(t, J¼6.6 Hz,
1H), 5.48 (d, J¼8.4 Hz,1H), 6.15 (s,1H), 6.58 (d, J¼9.9 Hz,1H), 6.63 (dd,
J¼9.9, 2.1 Hz, 1H), 6.64 (d, J¼2.1 Hz, 1H), 6.80 (br s, 1H), 6.92 (d,
J¼8.4 Hz, 2H), 7.22 (d, J¼8.4 Hz,1H), 7.36 (d, J¼8.4 Hz, 2H),13.19 (br s,
4.11. 5-Lipoxygenase assay
1H); 13C NMR (CDCl3)
d 17.7, 24.2, 25.6, 28.2, 28.2, 55.3, 69.9, 78.1, 99.7,
101.1, 103.0, 104.8, 107.9, 112.5, 114.0, 114.8, 120.9, 121.3, 127.0, 128.4,
129.3, 131.3, 133.0, 152.1, 155.3, 159.3, 159.5, 159.9, 161.1, 161.9, 182.2,
Possible inhibition of 5-lipoxygenase activity was determined
by the method of Sircar et al.30 and modified by Evans.31 All con-
centrations refer to final concentrations in 3 mL cuvettes main-
tained at 25 ꢂC in a thermostated bath. The assay mixture contained
13C NMR [(CD3)2CO]
d 16.8, 23.7, 24.9, 27.3, 27.3, 54.6, 69.5, 77.9, 98.8,
100.7, 102.6, 104.7, 106.4, 112.9, 113.8, 114.4, 120.9, 121.5, 127.1, 128.8,
129.4, 131.4, 131.5, 152.3, 156.4, 159.1, 159.6, 161.2, 161.8, 161.9, 182.3,
HRMS (EI) calcd for C33H32O7 (Mþ) 540.2148, found 540.2150.
10
mL of various chemicals (including CAPE, himanimide C, and
morusin) dissolved in DMSO, 0.1 M potassium phosphate buffer

1340
T.-H. Tseng et al. / Tetrahedron 66 (2010) 1335–1340
3. Daskiewicz, J. B.; Depeint, F.; Viornery, L.; Bayet, C.; Comte-Sarrazin, G.; Comte,
G.; Gee, J. M.; Johnson, I. T.; Ndjoko, K.; Hostettmann, K.; Barron, D. J. Med.
Chem. 2005, 48, 2790–2804.
4. (a) Fukai, T.; Satoh, K.; Nomura, T.; Sakagami, H. Fitoterapia 2003, 74, 720–724;
(b) Shi, Y. Q.; Fukai, T.; Sakagami, H.; Chang, W. J.; Yang, P. Q.; Wang, F. P.;
Nomura, T. J. Nat. Prod. 2001, 64, 181–188 and references therein.
5. Chi, Y. S.; Jong, H. G.; Son, K. H.; Chang, H. W.; Kang, S. S.; Kim, H. P. Biochem.
Pharmacol. 2001, 62, 1185–1191.
6. Li, N.; Sood, S.; Wang, S.; Fang, M.; Wang, P.; Sun, Z.; Yang, C. S.; Chen, X. Clin.
Cancer Res. 2005, 11, 2089–2096.
7. (a) Ghosh, J.; Myers, C. E. Adv. Exp. Med. Biol. 1999, 469, 577–582; (b) Ghosh, J.;
Myers, C. E. Adv. Exp. Med. Biol. 2002, 507, 415–420.
(pH 6.3,2.95 mL) prepared with analytical grade reagents and
100
M linoleic acid (ꢃ99%) from Fluka. The reaction was initiated
m
with the addition of 35 U isolated 5-lipoxygenase (Cayman) diluted
with equal volume potassium phosphate buffer maintained at 4 ꢂC.
The increase in absorbance at 234 nm was recorded for 10 min with
a single beam spectrophotometer (Hitachi U-2000). The percentage
inhibition of enzyme activity was calculated by comparison with
the negative control (DMSO). The concentration of each chemical
that caused 50% enzyme inhibition (IC50) was determined.
8. (a) Miranda, C. L.; Stevens, J. F.; Helmrich, A.; Henderson, M. C.; Rodriguez, R. J.;
Yang, Y. H.; Deinzer, M. L.; Barnes, D. W.; Buhler, D. R. Food Chem. Toxicol. 1999,
37, 271–285; (b) Tobe, H.; Muraki, Y.; Kitamura, K.; Komiyama, O.; Sato, Y.;
Sugioka, T.; Maruyama, H. B.; Matsuda, E.; Nagai, M. Biosci. Biotechnol. Biochem.
1997, 61, 158–159; (c) Tabata, N.; Ito, M.; Tomoda, H.; Omura, S. Phytochemistry
1997, 46, 683–687.
9. Cheng, C. F.; Lai, Z. C.; Lee, Y. J. Tetrahedron 2008, 64, 4347–4353.
10. Pommery, J.; Pommery, N.; Henichart, J. P. Prostaglandins Leukot. Essent. Fatty
Acids 2005, 73, 411–417.
11. (a) Banerji, A.; Goomer, N. C. Synth. Commun. 1985, 15, 1165–1170; (b) Banerji,
A.; Luthria, D. L. Indian J. Chem., Sect. B 1990, 29, 163–165; (c) Bbadhwar, I. C.;
Kang, K. S.; Venkataraman, K. J. Chem. Soc. 1932, 54, 1107–1112; (d) Baker, W. J.
Chem. Soc. 1933, 55, 1381–1389; (e) Mahal, H. S.; Venkataraman, K. J. Chem. Soc.
1934, 56, 1767–1771; (f) Gissaub, M. A.; Tarafdar, S. A. Indian J. Chem., Sect. B
1998, 37, 540–543; (g) Rajendra Prasad, K. J.; Periasamy, P. A.; Vijayalakshmi, C.
S. J. Nat. Prod. 1993, 56, 208–214 and references therein.
4.12. Cytotoxicity assay
3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide
(MTT) colorimetric assay was developed to monitor mammalian
cell survival and proliferation. MTT assay was measured by de-
hydrogenase activity as described by Mosmann,32 with minor
modification (Alley et al.).33 The MTT solution was prepared in
5 mg/mL of phosphate-buffered saline just before use and filtered
through a 0.22
m
M filter. Cells were seeded 5ꢁ104 cells per well into
24-well culture plates and incubated for 24 h. After a 24 h in-
cubation in various concentrations of morusin or synthetic com-
12. Fang, Y. T.; Lin, C. N.; Wu, H. T.; Lee, Y. J. J. Chin. Chem. Soc. 2007, 54, 817–822 and
references therein.
13. Horita, K.; Yoshioka, T.; Tanaka, T.; Oikawa, Y.; Yonemitsu, O. Tetrahedron 1986,
42, 3021–3028.
14. Cappa, A.; Marcantoni, E.; Torregiani, E. J. Org. Chem. 1999, 64, 5696–5699.
15. Johansson, R.; Samuelsson, B. J. Chem. Soc., Perkin Trans. 1 1984, 2371–2374.
16. Sharma, G. V. M.; Reddy, C. G.; Krishna, P. R. J. Org. Chem. 2003, 68, 4574–4575.
17. Srikrishna, A.; Viswajanani, R.; Sattigeri, J. A.; Vijaykumar, D. J. Org. Chem. 1995,
60, 5961–5962.
pounds, 10 mL of MTT solution was added to each well. Plates were
incubated in a CO2 incubator for 4 h, and then the blue crystals,
which are the metabolized product of MTT were extracted by iso-
propanol. Absorbance at 563 nm was determined and used for the
measurement of the proportion of surviving cells. The percentage
of the dehydrogenase activity at each concentration, compared
with that of the control, was calculated from the absorbance values.
18. Bouzide, A.; Sauve, G. Synlett 1997, 1153–1154.
19. (a) Romano, M.; Claria, J. FASEB J. 2003, 17, 1986–1995; (b) Pommery, N.; Tav-
erne, T.; Telliez, A.; Goossens, L.; Chaelier, C.; Pommery, J.; Goossens, J. F.;
Houssin, R.; Durant, F.; Henichart, J. P. J. Med. Chem. 2004, 47, 6195–6206.
20. Cheon, B. S.; Kim, Y. H.; Son, K. S.; Chang, H. W.; Kang, S. S.; Kim, H. P. Planta
Med. 2000, 66, 596–600.
21. Ko, H. H.; Yu, S. M.; Ko, F. N.; Teng, C. M.; Lin, C. N. J. Nat. Prod. 1997, 60,
1008–1011.
22. Lahey, F. N.; Stick, R. V. Aust. J. Chem. 1973, 26, 2311–2314.
23. Ahluwalia, V. K.; Arora, K. K.; Jolly, R. S. Synthesis 1981, 7, 527–529.
24. Manandhar, M. D.; Hussaini, F. A.; Kapil, R. S.; Shoeb, A. Phytochemistry 1985, 24,
199–200.
Acknowledgements
We thank the National Science Council of the Republic of China
for financial support (Research Grant NSC 96-2113-M-018-001-
MY2). We also acknowledge with appreciation the NSC-supported
HRMS spectrometry facility for providing high-resolution mass
spectra.
25. Nomura, T.; Fukai, T.; Yamada, S.; Katayanagi, M. Chem. Pharm. Bull. 1978, 26,
1394–1402.
26. Nomura, T.; Fukai, T.; Amagai, M. Heterocycles 1979, 12, 1529–1533.
27. Nomura, T.; Fukai, T.; Yamada, S.; Katayanagi, M. Chem. Pharm. Bull. 1978, 26,
1431–1436.
Supplementary data
Supplementary data associated with this article can be found in
the online version at doi:10.1016/j.tet.2009.12.002.
28. Nomura, T.; Fukai, T.; Yamada, S.; Katayanagi, M. Chem. Pharm. Bull. 1976, 24,
2898–2900.
29. Boonphong, S.; Baramee, A.; Kittakoop, P.; Puangsombat, P. Chiang Mai J. Sci.
2007, 34, 339–344.
30. Sircar, U. C.; Shwender, C. J.; Johnson, E. A. Prostaglandins 1983, 25, 393–396.
31. Evans, A. T. Biochem. Pharmacol. 1987, 36, 2035–2037.
32. Mosmann, T. J. Immunol. Methods 1983, 65, 55–63.
References and notes
1. Tseng, T. H.; Lee, Y. J. Anticancer Agents Med. Chem. 2006, 6, 347–365.
2. (a) Lee, Y. J.; Kuo, H. C.; Chu, C. Y.; Wang, C. J.; Tseng, T. H. Biochem. Pharmacol.
2003, 66, 2281–2289; (b) Kuo, H. C.; Kuo, W. H.; Lee, Y. J.; Lin, W. L.; Chou, F. P.;
Tseng, T. H. Cancer Lett. 2006, 234, 199–208.
33. Alley, M. C.; Scudiero, D. A.; Monks, A.; Hursey, M. L.; Czerwinski, M. J.; Fine, D.
L.; Abbott, B. J.; Shoemaker, R. H.; Boyd, M. R. Cancer Res. 1988, 48, 589–592.