Antitumor agents. 272. Structure-activity relationships and in vivo selective anti-breast cancer activity of novel neo-tanshinlactone analogues

Article abstract of DOI:10.1021/jm1000858

Neo-tanshinlactone (1) and its previously reported analogues, such as 2, are potent and selective in vitro antibreast cancer agents. The synthetic pathway to 2 was optimized from seven to five steps, with a better overall yield. Structure-activity relationships studies on these compounds revealed some key molecular determinants for this family of antibreast agents. Several derivatives (19-21 and 24) exerted potent and selective antibreast cancer activity with IC₅₀ values of 0.3,0.2,0.1, and 0.1 μg/mL, respectively, against the ZR-75-1 cell lines. Compound 24 was 2- to 3-fold more potent than 1 against SK-BR-3 and ZR-75-1. Importantly, 21 exhibited high selectivity; it was 23 times more active against ZR-75-1 than MCF-7. Compound 20 had an approximately 12-fold ratio of SK-BR-3/MCF-7 selectivity. In addition, analogue 2 showed potent activity against a ZR-75-1 xenograft model, but not PC-3 and MDA-MB-231 xenografts, as well as high selectivity against breast cancer cell line compared with normal breast tissue-derived cell lines. Further development of lead compounds 19-21 and 24 as clinical trial candidates is warranted.

Full text of DOI:10.1021/jm1000858

J. Med. Chem. 2010, 53, 2299–2308 2299
DOI: 10.1021/jm1000858
Antitumor Agents. 272. Structure-Activity Relationships and In Vivo Selective Anti-Breast Cancer
Activity of Novel Neo-tanshinlactone Analogues
Yizhou Dong,† Qian Shi,† Huei-Chen Pai,§ Chieh-Yu Peng,§ Shiow-Lin Pan,§ Che-Ming Teng,§ Kyoko Nakagawa-Goto,†
Donglei Yu,† Yi-Nan Liu,† Pei-Chi Wu,† Kenneth F. Bastow,*,‡ Susan L. Morris-Natschke,† Arnold Brossi,† Jing-Yu Lang,
Jennifer L. Hsu, Mien-Chie Hung, ,^ Eva Y.-H. P. Lee,# and Kuo-Hsiung Lee*,†
†Natural Products Research Laboratories, UNC Eshelman School of Pharmacy, University of North Carolina at Chapel Hill, Chapel Hill, North
Carolina 27599, ‡Division of Medicinal Chemistry and Natural Products, UNC Eshelman School of Pharmacy, University of North Carolina at
Chapel Hill, Chapel Hill, North Carolina 27599-7568, §Pharmacological Institute, College of Medicine, National Taiwan University, No. 1, Jen-Ai
Road, Sec. 1, Taipei, Taiwan, University of Texas M. D. Anderson Cancer Center, Houston, Texas 77030, ^Center for Molecular Medicine and
#
Graduate Institute of Cancer Biology, China Medical University and Hospital, Taichung, Taiwan 404, and Department of Biological Chemistry
and Department of Developmental & Cell Biology, University of California, Irvine, California 92697-4037
Received January 19, 2010
Neo-tanshinlactone (1) and its previously reported analogues, such as 2, are potent and selective in vitro
antibreast cancer agents. The synthetic pathway to 2 was optimized from seven to five steps, with a better
overall yield. Structure-activity relationships studies on these compounds revealed some key molecular
determinants for this family of antibreast agents. Several derivatives (19-21 and 24) exerted potent and
selective antibreast cancer activity with IC50 values of 0.3, 0.2, 0.1, and 0.1 μg/mL, respectively, against the
ZR-75-1 cell lines. Compound 24 was 2- to 3-fold more potent than 1 against SK-BR-3 and ZR-75-1.
Importantly, 21 exhibited high selectivity; it was 23 times more active against ZR-75-1 than MCF-7.
Compound 20 had an approximately 12-fold ratio of SK-BR-3/MCF-7 selectivity. In addition, analogue 2
showed potent activity against a ZR-75-1 xenograft model, but not PC-3 and MDA-MB-231 xenografts, as
well as high selectivity against breast cancer cell line compared with normal breast tissue-derived cell lines.
Further development of lead compounds 19-21 and 24 as clinical trial candidates is warranted.
Introduction
detection and more effective treatments, toxic side effects,
low tumor selectivity, and multidrug resistance with cancer
chemotherapy still prompt the development of novel potent
antibreast cancer agents.10
Historically, natural products have been the most significant
source of drugs and drug leads, which have led to numerous
clinically used medicines.1-4 Accordingly, our group is inter-
ested in the discovery and development of novel anticancer
drugs from natural plants. Drug discovery from medicinal
plants has played an important role in the treatment of cancer,
and about 74% of anticancer compounds are either natural
products or natural product-derived.5 Worldwide, over 10
million new cases of cancer (all types, excluding nonmelanoma
skin) and over 6 million deaths were estimated to occur in the
year 2000.6 More than 1 million women are diagnosed with
breast cancer every year, accounting for 10% of all new cancers
and 23% of all female cancer cases.7 The disease accounts for
40000 deaths each year in the United States alone.8 Tamoxifen
(TAM,a Figure 1) is the most widely used selective estrogen
receptor modulator (SERM) for the treatment of breast can-
cer.9 Other drugs, including cyclophosphamide, doxorubicin,
and paclitaxel, are also recommended to be used in combina-
tion in early breast cancer.8 Although the death rate from
breast cancer has declined significantly because of earlier
Tanshen, the rhizome of Salvia miltiorrhiza Bunge, is used
primarily in traditional Chinese medicine (TCM) for the
treatment of coronary heart diseases, inflammatory diseases,
and chronic hepatitis. Many biologically active constituents,
including neo-tanshinlactone, tanshinone I, and tanshinone
IIA, which have been studied extensively as anticancer agents,
were first isolated from the roots of Salvia miltiorrhiza.11 Neo-
tanshinlactone (1) (Figure 1), a minor component isolated
from an EtOH extract of S. miltiorrhiza, showed significant
selective in vitro antibreast cancer activity as compared to
TAM. Specifically, it was 10-fold more potent and 20-fold
more selective than TAM against ERþ and HER2þþ breast
cancer cells.12 Compound 2 (Figure 1), a congener of 1, was
about twice as active against MCF-7 and SK-BR-3 cell lines
as 1.13 Compound 2 was effective against approximately 40%
of human breast cancer cell lines and ineffective against other
cell lines tested (total 29 cell lines, unpublished data). Further-
more, preliminary in vivo evaluation in mice models suggested
that 2 is a potent and selective antibreast cancer agent. In both
BRCA1/p53 and wild-type mice treated with 2, mammary
gland side branching was dramatically reduced (unpublished
data). In addition, 2 significantly suppressed several impor-
tant protein kinases including CK2R1, ABL, and AKT1
(Table 3 in Supporting Information). Mechanism of action
studies are ongoing and will be reported in due course.
*To whom correspondence should be addressed. For K.F.B.: phone,
919-966-7633; E-mail, ken_bastow@unc.edu. For K.H.L.: phone, 919-
962-0066; fax, 919-966-3893; E-mail, khlee@unc.edu.
a Abbreviations: TAM, tamoxifen; SERM, selective estrogen recep-
tor modulator; TCM, traditional Chinese medicine; SAR, structure-
activity relationship; PPA, polyphosphoric acid; TGD, tumor growth
delay; TTE, time to end point.
r
2010 American Chemical Society
Published on Web 02/11/2010
pubs.acs.org/jmc

2300 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 5
Dong et al.
Scheme 1a
a Reagents and conditions: (a) EtMgBr, ZnCl2, THF, rt; (b) Pd/C, triglyme, reflux; (c) BBr3, CH2Cl2; (d) malonic acid, PPA (85% P2O5), 75 ꢀC, 3 h; (e)
chloroacetone, HOAc/NH4OAc, toluene/EtOH, reflux, 24 h; (f) NBS, dibenzoyl peroxide, toluene, reflux; (g) BBr3, CH2Cl2, reflux, 3 h; (h) Ac2O, Et3N,
10 h; (i) 2-chloro-N,N-dimethylethanamine, K2CO3, acetone, 12 h.
effects that can influence interaction of a ligand with its target
receptors. For our 1 analogues, we reported previously that a
C-4 substituent in ring-A is critical for antibreast cancer
activity.13 Thus, compounds 19-27 with substituents of
different sizes and electrostatic properties were designed to
find optimal groups at this position. In the B-ring, we changed
the phenyl ring to a pyridinone ring in 32 and 33 to explore a
ring system effect. The strategy of bioisoteric replacement can
be a powerful and highly productive tool in analogue design.
On the basis of this concept, the oxygens in ring-C were
changed to sulfur and nitrogen (38-39 and 40-42). Com-
pounds with different substituents on the furan D-ring
(44-53) were also designed. Moreover, the degree of satura-
tion (number of double bonds) can change the orientation of a
molecule and affect its in vitro activity and selectivity. Con-
sequently, we designed 54 and 55 to have a more saturated
dihydrofuran ring-D. Finally, the furan D-ring was changed
to a substituted phenyl ring in 58 to examine the ring system
effect and interaction volume.
Figure 1. Structures of tamoxifen, neo-tanshinlactone (1), and its
analogues 2-3.
Preliminary structure-activity relationships (SAR) showed
that a methylated furan ring-D and the C-4 substituent in ring
A are critical for antibreast cancer activity.13,14 These promis-
ing results encouraged us to continue the modification of this
series to develop novel anticancer drug candidates. To in-
crease the chemical availability, we also optimized the syn-
thetic pathway. In this paper, we describe further modifi-
cations of the A-, B-, C-, and D-rings as well as biological
evaluation of newly synthesized analogues against several
human cancer cell lines, including MCF-7 (estrogen receptor
positive, luminal-like breast cancer), SK-BR-3 (estrogen re-
ceptor negative, HER2 overexpressing luminal-like breast
cancer), ZR-75-1 (estrogen receptor positive, HER2 over-
expressing luminal-like breast cancer), MDA MB-231 (estro-
gen receptor negative, basal cell-like breast cancer), A549
(human lung cancer), DU145 (prostate cancer), KB (naso-
pharyngeal carcinoma), and KB-vin (vincristine-resistant KB
subline).
Chemistry
Synthesis of analogue 2 was achieved in five steps and an
overall yield of 18%, compared with seven steps and 3% yield
reported before12,13 (Scheme 1). A Grignard reaction of 4 in
the presence of zinc chloride gave 5 in an improved yield of
85%.15 Addition of zinc chloride increased the yield more
than 25%. Analogue 6 was obtained in one step by oxidation
of 5 with Pd/C rather than the prior two steps (hydrochloric
acid and Pd/C).13 Demethylation of 6 with boron tribromide
gave naphthol 7. Treatment of 7 with polyphosphoric acid
(PPA) in the presence of 85% P2O5 and malonic acid pro-
duced 13 in 53% yield. Phosphorus pentoxide was used to
remove water from the reaction system, and its use increased
the yield and reproducibility. Finally, analogue 2 was ob-
tained via a tandem alkylation/intramolecular Aldol reaction
with an optimized procedure (70% yield),14 which increased
the yield in this step by around 20%. This overall optimized
synthetic route can be applied to synthesize new analogues
and produce 2 in large scale for animal testing.
Design
Our general goals in drug design are to optimize the
synthesis of active analogues, systematically explore SAR,
and develop new more potent lead compounds. Thus, our first
goal in this study was to optimize the synthetic route to 2. We
aimed to reduce the number of steps and increase yields. The
optimized synthetic route would then be applied to synthesize
new analogues. Second, synthetic modifications of 1 were
considered because the resulting fundamental chemical and
physical changes may affect molecular shapes, bond angles,
and partition coefficients. Different substituents can have
different hydrophobic interactions, sizes, and electrostatic
Using the optimized synthetic pathway, target compounds
19-23 were prepared from various substituted 1-naphthols
(8-12), as shown in Scheme 1. Treating naphthols 8-12 with
malonic acid in the presence of PPA (85%P2O5) provided

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 5 2301
Scheme 2a
Target compounds 45-53 with various substituents on the
D-ring were synthesized with the same tandem alkylation/
intramolecular Aldol reaction using different bromoketones
(Scheme 4). Reduction of 3 and 2 with palladium acetate,
triethyl amine, and formic acid20 afforded 54 and 55, respec-
tively. Compound 58 was obtained from 56 by using ester-
ification21 and Heck reactions.22
Results and Discussion
a Reagents and conditions: (a) diethyl malonate, 220 ꢀC, 8 h; (b)
chloroacetone, HOAc/NH4OAc, toluene/EtOH, reflux, 24 h.
Together with 1 and previously reported analogues 2 and 3,
the newly synthesized analogues (19-27, 32-33, 37-38,
40-42, 45-53, 54-55, and 58) were evaluated for in vitro
antibreast cancer activity against two human tumor cell lines:
MCF-7 (ERþ) and SK-BR-3 (HER2þ). Compounds that
had ED50 values less than 4 μg/mL were also examined against
ZR-75-1 (ERþ, HER2þ) and MDA-MB-231 (ER-) breast
cancer cell lines (Table 1).
Scheme 3
Initially, we investigated the effects of substitutions around
the skeleton of 1 by comparing 1-3 with 19-27. The com-
pounds displayed different degrees of activity and selectivity
toward the four breast cancer cell lines.
Against the MCF-7 cell line, small alkyl groups were
favored relative to other groups at C-4 on ring-A. Analogue
2, which has a C-4 ethyl group, was the most potent com-
pound among those tested against MCF-7. Its ED50 (0.2 μg/
mL) was slightly better than that (0.6 μg/mL) of 1, which has a
C-4 methyl group. The rank order of potency for all C-4
substituted analogues was 2 (Et) > 1 (Me) > 20 (Pr) = 19
(iPr) > 21 (OMe) > 3 (H) > 23 (F) > 27 (dimethylamino) >
26 (OAc) > 25 (OH) > 22 (OEt). The substituents on the
furan (ring-D) double bond were also investigated. A methyl
group (2) was better than either an ethyl (47) or methoxyphe-
nyl (53) group at the R2 position. However, at the R3 position,
a methyl group was distinctly disfavored (48-51). These
results indicated that the optimal combination on ring-D
was methyl at R2 and hydrogen at R3.
Reagents and conditions: (a) malonic acid, PPA (85% P2O5), 75 ꢀC,
3 h; (b) chloroacetone, HOAc/NH4OAc, toluene, EtOH, reflux, 24 h; (c)
diethyl malonate, PPA (85% P2O5), 170 ꢀC, 2 h; (d) Lawesson’s reagent,
toluene, reflux, 5 h; (e) NH2OH HCl, NaOAc, MeOH, reflux, 12 h.
Most compounds were equipotent or more potent against
SK-BR-3 compared with MCF-7 cells. Compounds 2 and 1
were even more potent against SK-BR-3, with ED50 values of
0.1 and 0.2 μg/mL, respectively. However, 20 and 24 (4-
bromoethyl) were also equipotent to 2, and 19 was equipotent
to 1 against this cell line. Compounds 3, 23, 19, and 22 showed
good but lower activity (ED50 1.0, 1.1, 2.0, 2.5 μg/mL,
respectively), while 25, 26, and 27 were even less potent. These
results indicate that the size, orientation, and electronic effect
of groups at C-4 are important to the activity. Perhaps even
more importantly, certain substituents could greatly affect the
SK-BR-3/MCF-7 selectivity. Compounds 20-22 had ap-
proximately 10-fold ratios of SK-BR-3/MCF-7 selectivity.
For the D-ring analogues (45-53), most showed similar
activity against SK-BR-3 and MCF-7. An exception was
51, which showed moderate activity against SK-BR-3 (ED50
2.1 μg/mL) but was completely inactive against MCF-7.
To further explore the selectivity, compounds with ED50
values less than 4 μg/mL were further examined against two
additional breast cancer cell lines, ZR-75-1 (ERþ, HER2þ)
and MDA-MB-231 (ER-). Most compounds had similar
potency against the SK-BR-3 and ZR-75-1 cell lines. How-
ever, 19 had a 10-fold ratio of ZR-75-1/SK-BR-3 selectivity,
while 20 had a 3-fold ratio. Importantly, 21 showed a 23-fold
ratio of ZR-75-1/MCF-7 selectivity. All tested compounds
were not active against the MDA-MB-231 cell line, which
confirmed that these novel analogues were highly selective.
benzochromenones 14-18, which were converted to target
compounds 19-23 under the same conditions as for synthesis
of 2.16 Compound 24 was obtained by treatment of 2 with N-
bromosuccinimide (NBS) and dibenzoyl peroxide.17 Removal
of the methyl group of 21 with boron tribromide afforded 25,
which was esterified with acetic anhydride and alkylated with
2-chloro-N,N-dimethylethanamine under basic conditions to
give 26 and 27, respectively.
B-ring modification was achieved through a two-step reac-
tion sequence. Commercially available substituted anilines 28
and 29 were reacted with diethyl malonate at 220 ꢀC for 8 h to
give intermediates 30 and 31 (Scheme 2). The desired com-
pounds 32 and 33 were obtained through the tandem alkyla-
tion/intramolecular Aldol reaction described above and
shown in Scheme 1.
Target compounds 38 and 39, which are bioisosteres of 3,
were obtained by using the same two synthetic steps shown in
Scheme 1 for 2 from naphthol 7, except that the starting
materials were naphthalene-1-thiol 34 and naphthalen-1-
amine 35 (Scheme 3). Compounds 3 and 2 were converted
to thiolactones 40 and 41, respectively, using Lawesson’s
reagent.18 Compound 42 was obtained by treatment of 41
with sodium acetate and hydroxylamine hydrochloride.19

2302 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 5
Dong et al.
Scheme 4a
a Reagents and conditions: (a) HOAc/NH4OAc, toluene, EtOH, reflux, 24 h; (b) Et3N, formic acid, Pd/C, acetone, 12 h; (c) 2-bromo-4-methylbenzoyl
chloride, DMAP, DIEA, THF, rt, 12 h; (d) Pd(OAc)2, PPh3, NaOAc, DMF, reflux, 3 h.
As indicated by ZR-75-1/SK-BR-3 selectivity ratios, we
observed that some compounds (e.g., 20 and 21) were more
sensitive to cell lines overexpressing only HER2 (SK-BR-3
and ZR-75-1), while others (e.g., 19) weremoresensitivetocell
lines overexpressing both HER2 and ER (ZR-75-1). These
results will facilitate our studies on the mechanism(s) of
action. Because ring-A is critical to activity and selectivity,
we will further explore C1-C4 positions in the future.
and MCF10A cells, respectively, showing that2 is selective for
a subset of breast cancer-derived cell lines and is significantly
less active against normal breast-derived tissue.
We have examined the anticancer activity of compound2 in
several xenograft models such as PC-3 (androgen-indepen-
dent human prostate carcinoma cells), MDA-MB-231
(estrogen receptor negative basal-like breast cancer cells),
and ZR-75-1 (estrogen receptor positive HER2 overexpres-
sing breast cancer cells). Compound 2 was administered
intraperitoneally (ip) in a 4% benzyl alcohol/6% cremo-
phor/90% D5W solution and was given at 10 mg/kg every
otherdaytoendpoint (qodtoend). A positivereference group
received paclitaxel ip at 20 mg/kg every fourth day for 5 doses
(q4d ꢀ 5). A control group received vehicle ip on a qod to end
schedule. As shown in Figure 4, the treatment of SCID mice
with 2 resulted in inhibition of estrogen-positive ZR-75-1
tumor xenograft growth. There was significant reduction in
growth of estrogen-positive breast tumors in 2-treated ani-
mals as compared with the control group. Treatment results
were presented as percent tumor growth delay (%TGD),
which is the percent increase in the mean time to end point
(TTE) for drug-treated versus control mice. Logrank tests
determine significance of the differences between TTE values
for compound 2-treated and control mice, at P e 0.05. In ZR-
75-1 xenograft model, the mean TTE for the control group
was 15.1 days. Paclitaxel produced a mean TTE of 35.0 days,
corresponding to a %TGD of 132. Compound 2 at 10 mg/kg
produced a mean TTE of 29.5 days, corresponding to a %
TGD of 95 (p = 0.0067, logrank). Of the xenografts studied,
treatment with 2 only suppressed estrogen-dependent breast
cancer but had no effect on PC-3 (androgen-independent
human prostate carcinoma cells) or MDA-MB-231 (estro-
gen receptor negative basal-like breast cancer cells). These
findings suggest that compound 2 may be selectively used to
inhibit the growth of hormone-dependent breast cancers,
particularly regrowth of residual tumor in postmenopausal
We also investigated analogues involving skeletal modifica-
tions in ring-B, -C, or -D. Compounds 32-33 contain a
pyridinone rather than phenyl ring-B and were much less
active than 3. Bioisosteric modifications of either lactone
oxygen to nitrogen or sulfur in ring-C led to dramatically
decreased or abolished antibreast cancer activity (38-39,
40-42). The results demonstrated that the lactone ring is an
important featuretotheactivity. Compounds54and 55 witha
dihydrofuran ring-D showed decreased activity compared
with 3 and 2, respectively. Compound 58 with a substituted
phenyl rather than furan D-ring was inactive. These results,
together with our previously reported data, indicated that an
unsaturated furan is favored for ring-D. Studies on different
ring systems, such as 59 and 60, are also ongoing (Figure 2).
To further investigate human tumor-tissue-type selectivity,
compounds with ED50 values less than 4 μg/mL against breast
cancer cell lines were tested against four different human
cancer cell lines, A549 (lung), DU145 (prostate), KB
(nasopharnygeal), and KB-vin (its vincristine-resistant sub-
line) using 2 as a positive control (Table 2). All compounds,
except 22, were not active against these four tumor cell lines.
These results demonstrated that our novel analogues were
extremely selective for breast cancer cell lines.
Compound 2 was tested independently against cell lines
derived from normal breast tissue (MCF10A and 184A1)
versus SK-BR-3 as a positive breast cancer cell line control,
and results are shown in Figure 3. The interpolated ED50
values are 0.1, 4.4, and 2.7 μg/mL against SK-BR-3, 184A1,

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 5 2303
Table 1. Cytotoxicity of Compounds against Tumor Cell Linesa
compd
R1
R2
R3
X1
X2
MCF-7
SK-BR-3
ZR-75-1
MDA-MB-231
1
Me
Et
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
H
H
H
H
H
H
H
H
H
H
H
H
O
O
O
O
O
O
O
O
O
O
O
O
O
0.60
0.20
4.0
0.20
0.10
1.0
2.0
0.10
0.20
2.5
1.1
0.10
5.0
5.7
5.8
12
0.30
0.10
4.0
10
8.8
2
O
O
O
O
O
O
O
O
O
O
O
3
H
10.3
>10
>10
6.4
19
20
21
22
23
24
25
26
27
32
33
38
39
40
41
42
45
46
47
48
49
50
51
52
53
54
55
58
iPr
Pr
1.4
1.2
0.2
0.30
0.10
1.9
OMe
OEt
F
2.3
>20
4.5
9.8
0.80
0.10
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
2.3
>10
14
1-bromoethyl
NT
15.0
6.0
OH
OAc
DAE
Me
Pr
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
>10
NT
>10
NT
NT
NT
9.6
5.0
11
>20
>20
NT
>20
>20
>20
2.5
15
12
H
Me
Me
Me
Me
Me
Et
H
S
O
H
H
N
O
O
O
O
O
O
O
O
O
O
O
O
O
4.4
>20
16
H
Et
H
S
H
S
Et
H
H
NOH
O
19
H
1.8
11
Me
Et
Et
Et
H
O
7.5
1.3
NT
0.60
NT
NT
NT
2.2
H
O
1.5
6.9
9.8
12
H
Et
Me
Me
H
Me
Me
Me
Me
H
O
>20
8.0
O
H
Et
O
>20
>20
NT
>20
NT
NT
NT
H
O
2.1
5.8
>20
14
H
Et
PMP
PMP
O
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
H
O
H
Et
5.1
>20
a Mean ED50 (μg/mL), from 2 or more independent tests (standard error is listed in Supporting Information); NT: not tested; PMP: 4-methoxyphenyl,
DAE: 2-dimethylamino)ethoxy.
Table 2. Cytotoxicity of Compounds against Tumor Cell Linesa
compd
A549
DU145
KB
KBvin
2
11
11
16
11
15
14
4.7
18
18
8.7
16
14
13
11
12
13
3.7
13
13
7.5
12
14
13
7.3
11
19
20
21
22
23
24
45
47
51
12
10
12
3.5
14
5.3
15
Figure 2. Structures designed for future study.
>20
8.2
18
>20
6.6
15
breast cancer survivors receiving estrogen and progesterone
replacement therapy.
12
13
a Mean ED50 (μg/mL), from 2 or more independent tests (standard
error is listed in Supporting Information).
Conclusions
In summary, a highly efficient synthesis of 2 was accom-
plished with fewer steps and higher overall yield than those
previously reported. This synthetic pathway was used to pre-
pare new analogues. The SAR study led to the following
observations: (1) C-4 position is critical for both potency
and selectivity. The order of potency against SK-BR-3 was
ethyl=2-bromoethyl=propyl>methyl=methoxy>fluoro=
hydrogen>isopropyl > ethoxy>dimethylamino>acetate>
hydroxyl. Analogues with 4-isopropyl, -propyl, and -methoxy
groups showed high selectivity against different breast can-
cer cell lines. (2) The order of potency at the C-17 position
was methyl > ethyl>hydrogen, while the order of potency
at the C-16 position was hydrogen > methyl. (3) Pyridinone
ring is not favored for ring-B. (4) Lactone ring-C is essential for
activity; analogues with thiolactone and lactam rings were

2304 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 5
Dong et al.
inactive or less active. (5) Ring-D is preferably an unsatu-
rated furan ring. On the basis of all results, a mechanism of
action study is in progress. Because of their high selectivity
and potency, 19-21 and 24 are novel promising antibreast
cancer candidates. In addition, analogue 2 showed potent
activity against a ZR-75-1 xenograft model but not PC-3
and MDA-MB-231 xenografts. Resistance to endocrine
therapy is an important issue in the management of HER2
overexpressing estrogen receptor-positive breast cancer.23
Studies will continue to further establish the suitability of
neo-tanshinlactone analogues as clinical trials candidates
for treating breast cancer.
Experimental Section
Materials and Methods. Melting points were measured with a
Fisher Johns melting apparatus without correction. H NMR
1
spectra were measured on a 300 MHz Varian Gemini 2000
spectrometer using TMS as internal standard. The solvent used
was CDCl3 unless indicated. Mass spectra were measured on a
Shimadzu LC-MS2010 instrument. Thin-layer chromatography
(TLC) and preparative TLC were performed on precoated silica
gel GF plates purchased from Merck, Inc. Biotage Flashþ or
Isco Companion systems were used for flash chromatography.
Silica gel (200-400 mesh) from Aldrich, Inc., was used for
column chromatography. All other chemicals were obtained
from Aldrich, Inc., and Fisher, Inc. Intermediates 7-18 and
36-37 for target compounds 19-23 and 38-39 were prepared
by the optimized methods described in our previous and this
paper.13 Intermediates 30-31 were prepared by the reported
method.24,25 All final compounds are >95% pure on the basis
of two HPLC conditions.
Cell Growth Inhibition Assay. All stock cultures are grown in
T-25 flasks. Freshly trypsinized cell suspensions were seeded in
96-well microtiter plates with compounds added from DMSO-
diluted stock. The plates were incubated for an additional 72 h
after attachment and drug addition, and the assay was termi-
nated by 10% TCA. Then, 0.4% SRB dye in 1% HOAc was
added to stain the cells for 10 min. Unbound dye was removed
by repeated washing with 1% HOAc, and the plates were air-
dried. Bound stain was subsequently solved with 10 mM trizma
base and the absorbance read at 515 nm. Growth inhibition of
50% (ED50) was calculated as the drug concentration that
caused a 50% reduction in the net protein increase in control
cells during the drug incubation. The mean ED50 is the concen-
tration of agent that reduces cell growth by 50% under the
experimental conditions and is the average from at least three
independent determinations. Variation between replicates was
no more than 5% of the mean. The following human tumor cell
lines were used in the assay: A549 (nonsmall cell lung cancer),
MCF-7 (estrogen receptor positive luminal-like breast cancer),
MDA MB-231 (estrogen receptor negative basal-like breast
Figure 3. Selective in vitro anticancer activity of 2 against SK-BR-3
breast cancer versus normal breast tissue-derived cell lines
(MCF10A and 184A1). Legend: cell line description, source, and
activity determination using the MTT-dye assay are described in the
Experimental Section. Graphical data are the mean and standard
deviation of values obtained from replicates in a single experiment.
Figure 4. Anticancer activity of 2. The efficacy of 2 on the growth of human PC-3, MDA-MB-231, and ZR-75-1 xenografts in SCID mice.
Treatment with compound 2 selectively abrogated the hormone-dependent breast cancer. Tumor growth is presented as the mean tumor
volume (mm3) ( SE. Tumor volume was determined by caliper measurements and was calculated as the product of 1/2 ꢀ length ꢀ width2. Each
value represents the mean of at least five animals.

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 5 2305
cancer), SK-BR-3 (estrogen receptor negative, HER2 overex-
pressing luminal-like breast cancer), ZR-75-1 (estrogen receptor
positive breast cancer, HER2 overexpressing luminal-like breast
cancer), KB (nasopharyngeal carcinoma), KB-vin (vincristine-
resistant KB subline). All cell lines were obtained from the
Lineberger Cancer Center (UNC-CH) or from ATCC
(Rockville, MD). Cells propagated in RPMI-1640 supplemen-
ted with 10% FBS, penicillin 100 IU/mL, streptomycin 1 μg/
mL, and amphotericin B 0.25 μg/mL and were cultured at 37 ꢀC
in a humidified atmosphere of 95% air/5% CO2.
2H, CH2CH2CH3), 7.42-7.46 (m, 2H, aromatic and OCH), 7.55
(t, J = 8.4 Hz, 1H, aromatic), 7.83 (d, J = 9.0 Hz, 1H, aroma-
tic), 7.94 (d, J = 9.0 Hz, 1H, aromatic), 8.49 (d, J = 8.4 Hz,
1H, aromatic). HRMS for (Mþ þ H): calcd 293.1178, found:
293.1175.
6-Methoxy-1-methyl-11H-benzo[h]furo[3,2-c]chromen-11-one
(21). Yield 29%; mp 225-227 ꢀC. 1H NMR (300 MHz, CDCl3,
ppm): δ 2.40 (s, 3H, CH3), 4.02 (s, 3H, OCH3), 6.95 (d, J = 7.8
Hz, 1H, aromatic), 7.42 (s, 1H, OCH), 7.54 (t, J = 7.8 Hz, 1H,
aromatic), 7.78 (d, J = 9.0 Hz, 1H, aromatic), 8.15 (d, J = 9.0
Hz, 2H, aromatic). HRMS for (Mþ þ H): calcd 281.0814,
found: 281.0816.
6-Ethoxy-1-methyl-11H-benzo[h]furo[3,2-c]chromen-11-one
(22). Yield 28%; mp 201-203 ꢀC. 1H NMR (300 MHz, CDCl3,
ppm): δ 1.56 (t, J = 7.2 Hz, 3H, CH2CH3), 2.39 (s, 3H, CH3),
4.20 (q, J = 7.2 Hz, 2H, CH2CH3), 6.91 (d, J = 7.8 Hz, 1H,
aromatic), 7.40 (s, 1H, OCH), 7.50 (t, J = 8.4 Hz, 1H, aromatic),
7.75 (d, J = 9.0 Hz, 1H, aromatic), 8.11 (d, J = 8.4 Hz, 1H,
aromatic), 8.16 (d, J = 9.0 Hz, 1H, aromatic). HRMS for (Mþ þ
H): calcd 295.0970, found: 295.0970.
6-Fluoro-1-methyl-11H-benzo[h]furo[3,2-c]chromen-11-one
(23). Yield 20%; mp 215-217 ꢀC. 1H NMR (300 MHz, CDCl3,
ppm): δ 2.42 (d, J = 1.8 Hz, 3H, CH3), 7.27-7.33 (m, 1H,
aromatic), 7.48 (d, J = 1.2 Hz, 1H, OCH), 7.55-7.62 (m, 1H,
aromatic), 7.93 (d, J = 9.0 Hz, 1H, aromatic), 8.03 (d, J = 9.0
Hz, 1H), 8.40 (d, J = 8.7 Hz, 1H, aromatic). HRMS for (Mþ þ
H): calcd 269.0614, found: 269.0616.
Optimized Synthetic Procedure to Analogue 2. To a solution of
EtMgCl (3.0 M in diethyl ether, 26 mL, 78 mmol), ZnCl2 (798
mg, 6 mmol) was added at rt under argon atmosphere. This
mixture was stirred at rt for 1 h. Then, the solution was cooled to
0 ꢀC, and 5-methoxy-1-tetralone 4 (10.56 g, 60 mmol) was added
at 0 ꢀC. The mixture was stirred overnight. The reaction mixture
was quenched by saturated aqueous NH4Cl, extracted with
EtOAc, and dried over Na2SO4. The organic phase was con-
centrated under reduced pressure, and the resulting residue was
purified by silica gel column chromatography (eluent: hexane/
EtOAc), to give the desired product 5 (10.5 g, 85% yield). 10%
Pd/C (7.18 g. 33.24 mmol) was added to a solution of 5 (7.18 g) in
triglyme (15 mL), and the mixture was heated to reflux for 3 days
to afford 6 (3.71 g, 60% yield). To a solution of 6 (3.5 g, 18.8
mmol) in anhydrous CH2Cl2 (20 mL) was added a solution of
boron tribromide in CH2Cl2 (1.0 M, 56.5 mL, 56.5 mmol)
dropwise at 0 ꢀC. The mixture was stirred overnight at rt to give
the desired naphthol intermediate 7 (3.07 g, yield 95%) after
silica gel chromatography. A mixture of 7 (2.14 g, 12.47 mmol),
malonic acid (1.43 g, 13.72 mmol), and PPA (85%P2O5, 20 g)
was heated at 75 ꢀC for 3 h. After cooling, ice-water was added
to the black residue. The mixture was filtered and the solid
dissolved in MeOH. The organic layer was concentrated and
purified with flash chromatography, eluting with CH2Cl2:
MeOH = 10:1, to yield 13 as a yellow solid (1.52 g, 53% yield).
To a solution of 13 (440 mg, 1.83 mmol) in toluene (55 mL) was
added a mixture of HOAc (549 mg, 9.15 mmol) and NH4OAc
(704 mg, 9.15 mmol) in EtOH (16 mL) and chloroacetone (842
mg, 9.15 mmol). The mixture was stirred for 30 min at rt and
then heated to 60 ꢀC for 30 min. Subsequently, it was refluxed
for 24 h. After cooling, the mixture was diluted with H2O and
extracted with EtOAc. The organic layer was dried over
Na2SO4, filtered, and evaporated. The residue was purified by
column chromatography (hexane/EtOAc) to give 2 (356 mg,
70% yield).
Synthesis of Neo-tanshinlactone Analogues 19-23, 32-33,
38-39, and 45-53. To a solution of 13-18, 30-31, 36-37, or
43-44 (0.20 mmol) in toluene (8 mL) was added a mixture of
HOAc (59 mg, 1.0 mmol) and NH4OAc (75 mg, 1.0 mmol) in
EtOH (2 mL) and chloroacetone (90 mg, 1.0 mmol) or 3-
bromobutan-2-one, 2-bromopropanal, 1-bromobutan-2-one,
or 2-bromo-1-(4-methoxyphenyl)ethanone. The mixture was
stirred for 30 min at rt and then heated to 60 ꢀC for 30 min.
Subsequently, it was refluxed for 24 h. After cooling, the mixture
was diluted with H2O and extracted with EtOAc. The organic
layer was dried over Na2SO4, filtered, and evaporated. The
residue was purified by column chromatography (hexane/
EtOAc) to give a white solid.
6-Isopropyl-1-methyl-11H-benzo[h]furo[3,2-c]chromen-11-one
(19). Yield 65%; mp 155-157 ꢀC. 1H NMR (300 MHz, CDCl3,
ppm): δ 1.42 (t, J = 6.6 Hz, 6H, (CH3)2), 2.41 (d, J = 1.2 Hz, 3H,
CH3), 3.76 (h, J = 6.9 Hz, 1H, CH(CH3)2), 7.44 (d, J = 1.2 Hz,
1H, OCH), 7.56-7.64 (m, 2H, aromatic), 7.87 (d, J = 9.0 Hz,
1H, aromatic), 8.05 (d, J = 9.0 Hz, 1H, aromatic), 8.51 (d, J =
7.8 Hz, 1H, aromatic). HRMS for (Mþ þ H): calcd 293.1178,
found: 293.1168.
5-Aza-N-methyl-1-methyl-4H-benzo[h]furo[3,2,c]chromene-4,
11-dione (32). Yield 50%; mp 265-267 ꢀC. 1H NMR (300 MHz,
CDCl3, ppm): δ 2.38 (d, J = 1.5 Hz, 3H, CH3), 3.82 (s, 3H,
NCH3), 7.39 (t, J=8.1 Hz, 1H, aromatic), 7.46 (d, J=8.4 Hz,
1H, aromatic), 7.54 (d, J=1.5 Hz, 1H, OCH), 7.68-7.74 (m, 1H,
aromatic), 8.36 (dd, J=1.2, 8.4 Hz, 1H, aromatic). HRMS for
(Mþ þ H): calcd 282.0766, found: 282.0757.
5-Aza-N-propyl-1-methyl-4H-benzo[h]furo[3,2,c]chromene-4,
11-dione (33). Yield 55%; mp 238-240 ꢀC. 1H NMR (300 MHz,
CDCl3, ppm): δ 1.08 (t, J = 7.5 Hz, 3H, CH2CH2CH3), 1.83 (h,
J = 7.5 Hz, 2H, CH2CH2CH3), 2.38 (d, J = 0.9 Hz, 3H, CH3),
4.35 (t, J = 7.5 Hz, 2H, CH2CH2CH3), 7.37 (t, J = 7.8 Hz, 1H,
aromatic), 7.43 (d, J = 8.7 Hz, 1H, aromatic), 7.53 (d, J = 1.2
Hz, 1H, OCH), 7.66-7.72 (m, 1H, aromatic), 8.35 (dd, J = 1.8,
7.8 Hz, 1H, aromatic). HRMS for (Mþ þ H): calcd 310.1079,
found: 310.1068.
1-Methyl-11H-benzo[h]furo[3,2-c]thiochromen-11-one (38).
Yield 8%; mp 137-139 ꢀC. 1H NMR (300 MHz, CDCl3, ppm):
δ 2.42 (d, J=1.5 Hz, 3H, CH3), 7.44 (q, J=1.5 Hz, 1H, OCH),
7.61-7.64 (m, 2H, aromatic), 7.85 (d, J=8.4 Hz, 1H, aromatic),
7.88-7.92 (m, 1H, aromatic), 8.15 (d, J=8.7 Hz, 1H, aromatic),
8.21-8.22 (m, 1H, aromatic). HRMS for (Mþ þ H): calcd
267.0480, found: 267.0473.
1-Methylbenzo[h]furo[3,2-c]quinolin-11(10H)-one (39). Yield
1
10%; mp 135-137 ꢀC. H NMR (300 MHz, CDCl3, ppm): δ
2.53 (d, J=1.5 Hz, 3H, CH3), 7.47 (d, J=1.2 Hz, 1H, OCH),
7.61-7.71 (m, 3H, aromatic), 7.93 (d, J=8.4 Hz, 1H, aromatic),
7.98 (d, J=8.7 Hz, 1H, aromatic), 8.45 (d, J = 8.1 Hz, 1H,
aromatic). HRMS for (Mþ þ H): calcd 250.0868, found:
250.0855.
1-Ethyl-11H-benzo[h]furo[3,2-c]chromen-11-one (45). Yield
1
64%; mp 151-153 ꢀC. H NMR (300 MHz, CDCl3, ppm): δ
1.32 (t, J=7.5 Hz, 3H, CH2CH3), 2.78 (q, J=7.5 Hz, 2H, CH2-
CH3), 7.32 (s, 1H, OCH), 7.52-7.61 (m, 3H, aromatic), 7.66 (d,
J=8.4 Hz, 1H, aromatic), 7.76 (d, J = 8.7 Hz, 1H, aromatic),
8.43 (d, J = 7.5 Hz, 1H, aromatic). HRMS for (Mþ þ H): calcd
265.0865, found: 265.0865.
1-Ethyl-6-methyl-11H-benzo[h]furo[3,2-c]chromen-11-one (46).
Yield 65%; mp 183-185 ꢀC. 1H NMR (300 MHz, CDCl3, ppm):
δ 1.33 (q, J = 7.5 Hz, 3H, CH2CH3), 2.65 (s, 3H, CH3), 2.81 (q,
J=7.2 Hz, 2H, CH2CH3), 7.37 (s, 2H, aromatic and OCH), 7.45
(t, J = 8.1 Hz, 1H, aromatic), 7.71 (d, J = 8.7 Hz, 1H, aromatic),
7.78 (d, J = 9.0 Hz, 1H, aromatic), 8.36 (d, J = 8.4 Hz, 1H,
1-Methyl-6-propyl-11H-benzo[h]furo[3,2-c]chromen-11-one (20).
Yield 59%; mp 141-143 ꢀC. 1H NMR (300 MHz, CDCl3, ppm):
δ 1.04 (t, J = 7.2 Hz, 3H, CH2CH2CH3), 1.78 (m, 2H, CH2-
CH2CH3), 2.40 (d, J = 1.2 Hz, 3H, CH3), 3.06 (t, J = 7.5 Hz,

2306 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 5
Dong et al.
aromatic). HRMS for (Mþ þ H): calcd 279.1021, found:
279.1017.
6-Hydroxy-1-methyl-11H-benzo[h]furo[3,2-c]chromen-11-one
(25). To a solution of 21 (32 mg, 0.114 mmol) in DCM (3 mL)
was added BBr3 (1.12 mL, 1.12 mmol) dropwise at 0 ꢀC. The
reaction mixture was stirred and warmed to rt for 12 h. Water
was added to quench the reaction. The solution was extracted
with CHCl3. The organic layer was concentrated purified
with flash chromatography, eluting with DCM-MeOH, 15:1,
to give 25.
1,6-Diethyl-11H-benzo[h]furo[3,2-c]chromen-11-one (47). Yield
75%; mp 101-103 ꢀC. 1H NMR (300 MHz, CDCl3, ppm): δ 1.33
(q, J = 7.8 Hz, 6H, CH3), 2.83 (q, J = 7.5 Hz, 2H, CH2CH3), 3.04
(q, J = 7.8 Hz, 2H, CH2CH3), 7.35-7.40 (m, 2H, aromatic and
OCH), 7.48 (t, J = 7.5 Hz, 1H, aromatic), 7.70 (d, J = 9.3 Hz, 1H,
aromatic), 7.83 (d, J = 8.7 Hz, 1H, aromatic), 8.37 (d, J = 8.1 Hz,
1H, aromatic). HRMS for (Mþ þ H): calcd 293.1178, found:
293.1169.
1,2-Dimethyl-11H-benzo[h]furo[3,2-c]chromen-11-one (48). Yield
15%; mp 103-105 ꢀC. 1H NMR (300 MHz, CDCl3, ppm): δ 2.32
(d, J = 0.9 Hz, 3H, OCCH3), 2.42 (d, J = 0.9 Hz, 3H, CH3),
7.59-7.65 (m, 2H, aromatic), 7.72 (d, J = 8.7 Hz, 1H, aromatic),
7.82 (d, J = 8.7 Hz, 1H, aromatic), 7.85-7.88 (m, 1H, aromatic),
8.58 (d, J = 8.7 Hz, 1H, aromatic). HRMS for (Mþ þ H): calcd
265.0865, found: 265.0860.
6-Ethyl-1,2-dimethyl-11H-benzo[h]furo[3,2-c]chromen-11-one
(49). Yield 29%; mp 141-143 ꢀC. 1H NMR (300 MHz, CDCl3,
ppm): δ 1.38 (t, J = 7.5 Hz, 3H, CH2CH3), 2.29 (s, 3H, OCCH3),
2.39(s, 3H, CH3), 3.09 (q, J=7.5 Hz, 2H, CH2CH3), 7.42 (d, J=
6.9 Hz, 1H, aromatic), 7.50-7.55 (m, 1H, aromatic), 7.75-7.80
(m, 1H, aromatic), 7.88-7.91 (m, 1H, aromatic), 8.44 (d, J=
8.1 Hz, 1H, aromatic). HRMS for (Mþ þH): calcd 293.1178,
found: 293.1172.
Yield 52%. 1H NMR (300 MHz, DMSO, ppm): δ 2.30 (s, 3H,
CH3), 7.06 (d, J = 8.7 Hz, 1H, aromatic), 7.53 (t, J = 8.4 Hz,
1H, aromatic), 7.82-7.86 (m, 2H, aromatic), 7.97 (d, J = 1.2
Hz, 1H, OCH), 8.11 (d, J = 9.3 Hz, 1H, aromatic), 10.58 (s, 1H,
OH). HRMS for (Mþ - H): calcd 265.0501, found: 265.0505.
1-Methyl-11-oxo-11H-benzo[h]furo[3,2-c]chromen-6-yl Acet-
ate (26). Compound 25 (0.1 mmol) was dissolved in acetic
anhydride under argon. Triethylamine (0.14 mL, 1.0 mmol)
was added to the solution. After stirring overnight at 60 ꢀC, the
solution was washed with water and extracted with DCM and
dried (MgSO4). Removal of solvent under reduced pressure
yielded a white solid, which was purified by column chroma-
tography, eluting with EtOAc-hexane (1:4).
Yield 43%. 1H NMR (300 MHz, CDCl3, ppm): δ 2.41 (s, 3H,
CH3), 2.50 (s, 3H, COCH3), 7.39 (d, J = 6.9 Hz, 1H, aromatic),
7.46 (s, 1H, OCH), 7.65 (t, J = 8.1 Hz, 1H, aromatic), 7.80 (d,
J = 8.7 Hz, 1H, aromatic), 7.90 (d, J = 8.7 Hz, 1H, aromatic),
8.52 (d, J = 8.4 Hz, 1H, aromatic). HRMS for (Mþ þ H): calcd
309.0763, found: 309.0762.
2-Methyl-11H-benzo[h]furo[3,2-c]chromen-11-one (50). Yield
1
12%; mp 229-231 ꢀC. H NMR (300 MHz, CDCl3, ppm): δ
2.53 (s, 3H, CH3), 6.63 (s, 1H, OCCH), 7.61-7.65 (m, 2H,
aromatic), 7.75 (d, J = 8.7 Hz, 1H, aromatic), 7.83-7.90 (m,
2H, aromatic), 8.59 (d, J = 7.5 Hz, 1H, aromatic). HRMS for
(Mþ þ H): calcd 251.0708, found: 251.0703.
6-(2-(Dimethylamino)ethoxy)-1-methyl-11H-benzo[h]furo-
[3,2-c]chromen-11-one (27). Compound 25 (0.1 mmol) was dis-
solved in acetone under argon. K2CO3 (235 mg, 1.7 mmol) was
added to the solution. After stirring for 10 min, 2-chloro-N,N-
dimethylethylamine hydrochloride (30 mg, 0.2 mmol) was
added to the mixture. After refluxing for 10 h, the mixture was
filtrated and concentrated. The residue was purified by column
chromatography, eluting with EtOAc-hexane (1:2).
6-Ethyl-2-methyl-11H-benzo[h]furo[3,2-c]chromen-11-one (51).
Yield 2%; mp 175-177 ꢀC. 1H NMR (300 MHz, CDCl3, ppm):
δ 1.41 (t, J = 7.5 Hz, 3H, CH2CH3), 2.54 (s, 3H, CH3), 3.15 (q,
J = 7.5 Hz, 2H, CH2CH3), 6.64 (s, 1H, OCCH), 7.48 (d, J = 7.2
Hz, 1H, aromatic), 7.58 (t, J = 7.2 Hz, 1H, aromatic), 7.88 (d,
J = 9.0 Hz, 1H, aromatic), 8.00 (d, J = 9.0 Hz, 1H, aromatic),
8.50 (d, J = 8.4 Hz, 1H, aromatic). HRMS for (Mþ þ H): calcd
279.1021, found: 279.1017.
1
Yield 9%; mp 209-211 ꢀC. H NMR (300 MHz, CDCl3,
ppm): δ 2.41 (s, 3H, CH3), 2.44 (s, 6H, N(CH3)2), 2.96 (t, J = 5.1
Hz, 2H, OCH2CH2), 4.31 (t, J = 5.1 Hz, 2H, OCH2CH2), 6.98
(d, J = 7.5 Hz, 1H, aromatic), 7.45 (s, 1H, OCH), 7.55 (t, J = 8.1
Hz, 1H, aromatic), 7.83 (d, J = 8.7 Hz, 1H, aromatic), 8.20 (dd,
J = 7.2, 8.7 Hz, 2H, aromatic). HRMS for (Mþ þ H): calcd
338.1392, found: 338.1389.
Synthesis of Neo-tanshinlactone Analogues 40-41. A mixture
of compound 3 or 2 (0.1 mmol) and Lawesson’s reagent (81 mg,
0.2 mmol) in dry toluene (5 mL) was heated to reflux for 7 h.
After cooling, toluene was removed in vacuo and the red residue
was dissolved in EtOAc and partitioned with H2O. The organic
phase was separated and dried over MgSO4. Removal of solvent
in vacuo afforded an oily residue, which was purified by column
chromatography (EtOAc-hexane) to give a yellow solid.
1-Methyl-11H-benzo[h]furo[3,2-c]chromene-11-thione (40). Yield
90%; mp 267-269 ꢀC. 1H NMR (300 MHz, CDCl3, ppm): δ 2.53
(s, 3H, CH3), 7.46 (s, 1H, OCH), 7.67-7.69 (m, 2H, aromatic),
7.79-7.92 (m, 3H, aromatic), 8.75 (d, J = 7.2 Hz, 1H, aromatic).
HRMS for (Mþ þ H): calcd 267.0480, found: 267.0471.
1-(4-Methoxyphenyl)-11H-benzo[h]furo[3,2-c]chromen-11-one
(52). Yield 20%; mp 173-175 ꢀC. 1H NMR (300 MHz, CDCl3,
ppm): δ 3.88 (s, 3H, CH3), 7.02 (d, J = 9.0 Hz, 2H, aromatic),
7.64-7.68 (m, 2H, aromatic and OCH), 7.59-7.80 (m, 4H,
aromatic), 7.90-7.94 (m, 2H, aromatic), 8.62-8.65 (m, 1H,
aromatic). HRMS for (Mþ þ H): calcd 343.0970, found:
343.0975.
6-Ethyl-1-(4-methoxyphenyl)-11H-benzo[h]furo[3,2-c]chro-
1
men-11-one (53). Yield 34%; mp 185-187 ꢀC. H NMR (300
MHz, CDCl3, ppm): δ 1.39 (t, J = 7.5 Hz, 3H, CH2CH3), 3.11
(q, J = 7.2 Hz, 2H, CH2CH3), 3.86 (s, 3H, OCH3), 6.85-7.01
(m, 2H, aromatic), 7.46 (d, J = 6.6 Hz, 1H, aromatic), 7.56 (t,
J = 7.8 Hz, 1H, aromatic), 7.72-7.77 (m, 3H, aromatic and
OCH), 7.86 (d, J = 8.7 Hz, 1H, aromatic), 7.95 (d, J = 8.7 Hz,
1H, aromatic), 8.47 (d, J = 8.1 Hz, 1H, aromatic). HRMS for
(Mþ þ H): calcd 371.1283, found: 371.1291.
6-(1-Bromoethyl)-1-methyl-11H-benzo[h]furo[3,2-c]chromen-
11-one (24). To a solution of 2 (27 mg, 0.1 mmol) in CCl4 (3 mL)
was added N-bromosuccinimide (18 mg, 0.1 mmol) and diben-
zoyl peroxide (2 mg). The reaction mixture was stirred and
heated at reflux for 9 h. After the mixture was cooled in an ice
bath, the solid was removed by filtration and washed with CCl4.
Concentration and silica gel flash column chromatography
(hexane-EtOAc, 8:1) gave 24 (18 mg, 52%) as a white solid.
6-Ethyl-1-methyl-11H-benzo[h]furo[3,2-c]chromene-11-thione
(41). Yield 84%; mp 189-191 ꢀC. 1H NMR (300 MHz, CDCl3,
ppm): δ 1.38 (t, J = 7.5 Hz, 3H, CH2CH3), 2.49 (d, J = 1.2 Hz,
3H, CH3), 3.09 (q, J = 7.5 Hz, 2H, CH2CH3), 7.40 (d, J = 1.5
Hz, 1H, OCH), 7.46 (d, J = 7.2 Hz, 1H, aromatic), 7.56 (t, J =
7.2, 8.1 Hz, 1H, aromatic), 7.78 (d, J = 9.0 Hz, 1H, aromatic),
7.95 (d, J = 9.0 Hz, 1H, aromatic), 8.55 (d, J = 8.7 Hz, 1H, aro-
matic). HRMS for (Mþ þ H): calcd 295.0793, found: 295.0779.
6-Ethyl-1-methyl-11H-benzo[h]furo[3,2-c]chromen-11-one Oxime
(42). A mixture of 41 (22 mg. 0.075 mmol), hydroxylamine
hydrochloride (10.4 mg, 0.15 mmol), sodium acetate (12 mg,
0.15 mmol), and MeOH (5 mL) was refluxed overnight and then
filtered. The filtrate was concentrated under reduced pressure to
give an oil. Purification by the column chromatography
(EtOAc-hexane) on silica gel gave 42 as a white in 87% yield.
1
Yield 52%; mp 173-175 ꢀC. H NMR (300 MHz, CDCl3,
ppm): δ 2.29 (d, J = 6.9 Hz, 3H, CHBrCH3), 2.42 (d, J = 1.2 Hz,
3H, CH3), 5.97 (q, J = 7.5 Hz, 1H, CHCH3), 7.47 (d, J = 1.2 Hz,
1H, OCH), 7.65 (t, J = 7.8 Hz, 1H, aromatic), 7.88 (d, J = 6.9
Hz, 1H, aromatic), 8.00 (d, J = 9.0 Hz, 1H, aromatic), 8.15 (d,
J = 8.7 Hz, 1H, aromatic), 8.65 (d, J = 8.7 Hz, 1H, aromatic).
HRMS for (Mþ þ H): calcd 357.0126, found: 357.0120.

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 5 2307
1
1
Yield 89%; mp 211-213 ꢀC. H NMR (300 MHz, CDCl3,
ppm): δ 1.39 (t, J = 7.2 Hz, 3H, CH2CH3), 2.28 (d, J = 0.9 Hz,
3H, CH3), 3.11 (q, J = 7.5 Hz, 2H, CH2CH3), 7.18 (s, 1H, OH),
7.33 (d, J = 1.2 Hz, 1H, OCH), 7.40 (d, J = 6.9 Hz, 1H, aroma-
tic), 7.52 (dd, J = 7.2, 8.4 Hz, 1H, aromatic), 7.71 (d, J = 9.0 Hz,
1H, aromatic), 7.86 (d, J = 9.0 Hz, 1H, aromatic), 8.40 (d, J =
8.7 Hz, 1H, aromatic). HRMS for (Mþ þ H): calcd 294.1130,
found: 294.1118.
Yield 65%; mp 193-195 ꢀC. H NMR (300 MHz, CDCl3,
ppm): δ 2.55 (s, 3H, CH3), 7.35-7.38 (m, 1H, aromatic),
7.57-7.63 (m, 2H, aromatic), 7.71 (d, J = 9.0 Hz, 1H, aroma-
tic), 7.82-7.85 (m, 1H, aromatic), 7.91 (s, 1H, aromatic), 8.00
(d, J = 9.3 Hz, 1H, aromatic), 8.30 (d, J = 7.8 Hz, 1H, aroma-
tic), 8.53-8.56 (m, 1H, aromatic). HRMS for (Mþ þ H): calcd
261.0916, found: 261.0909.
Methodology of MTT Assay.26 The MTT assay was used to
access the in vitro anticancer activity of 2 against two normal
breast cancer cell lines 184A1 and MCF10A (CRL-10317)
purchased from ATCC (Rockville, MD) and using SK-BR-3
as a positive control. Cells were seeded into 96-well plates at a
density of 5000 cells per well in the recommended growth
medium. The drug was dissolved in DMSO. The drug was added
into wells after overnight incubation. After 72 h of incubation at
37 ꢀC in 5% CO2, 20 μL of MTT [3-(4,5-dimethylthiazol-2-yl)-
2,5 -diphenyl tetrazolium bromide] reagent was added to each
well and incubated for 2 h. The amount of formazon product
was measured at an OD of 570 nM using a plate-reader.
Antitumor Activity in Vivo. Male (for PC-3) and female (for
MDA-MB-231, and ZR-75-1) SCID mice (NTUH Animal
Facility) were 5 weeks old and had a body weight (BW) range of
20-24 g on D1 of the study. The animals were fed ad libitum water
(reverseosmosis,1ppmCl) andPicoLabRodentDiet20Modified
and Irradiated Lab Diet consisting of 20.0% crude protein, 9.9%
crude fat, and 4.7% crude fiber. The mice were housed at National
Taiwan University Laboratory Animal Center, NTUMC, on a
12 h light cycle at 21-23 ꢀC and 60-85% humidity. Nude-
athymic mice were maintained in accordance withthe Institutional
Animal Care and Use Committee procedures and guidelines. All
human cancer cells were maintained in RPMI 1640 medium
containing 100 units/mL penicillin G sodium, 100 μg/mL strepto-
mycin sulfate, 0.25 μg/mL amphotericin B, and 25 μg/mL genta-
micin. The medium was supplemented with 10% heat-inactivated
fetal bovine serum and 2 mM glutamine. The cells were cultured in
tissue culture flasks in a humidified incubator at 37 ꢀC, in an
atmosphere of 5% CO2 and 95% air. All human cancer cells used
for implantation were harvested during log phase growth and
resuspended in phosphate-buffered saline at 5 ꢀ 107 cells/mL.
Each mouse was injected sc in the right flank with 1 ꢀ 107 cells
(0.2 mL cell suspension). Tumors were monitored twice weekly
and then daily as their volumes approached 80-150 mm3. Tumor
size, in mm3, was calculated from:
Synthesis of Neo-tanshinlactone Analogues 54-55. Com-
pound 3 or 2 (0.2 mmol) was dissolved in acetone at 40 ꢀC under
argon. Pd/C (81 mg, 10%), triethylamine (0.33 mL, 2.40 mmol),
and formic acid (0.075 mL, 2.00 mmol) were added to the solution.
After stirring overnight, TLC showed some substrate remained
unreacted. The solution was filtered through celite and solvent
removed in vacuo to yield a dark oil. The residue was dissolved in
DCM before washing with saturated aqueous sodium bicarbonate
(5 mL), aqueous citric acid (5 mL, 10% v/v), water (5 mL), and
brine (5 mL) and then dried (MgSO4). Removal of solvent under
reduced pressure yielded a white solid, which was purified by
column chromatography, eluting with EtOAc-hexane (1:4).
1-Methyl-1H-benzo[h]furo[3,2-c]chromen-11(2H)-one (54).
1
Recovered yield 30%; mp 143-145 ꢀC. H NMR (300 MHz,
CDCl3, ppm): δ 1.46 (d, J = 6.9 Hz, 3H, CHCH3), 3.68-3.77
(m, 1H, CHCH3), 4.47 (q, J = 6.3 Hz, 1H, CH2), 5.00 (t, J = 6.6
Hz, 1H, CH2), 7.62-7.71 (m, 4H, aromatic), 7.87-7.90 (m, 1H,
aromatic), 8.59-8.62 (m, 1H, aromatic). HRMS for (Mþ þ H):
calcd 253.0865, found: 253.0856.
6-Ethyl-1-methyl-1H-benzo[h]furo[3,2-c]chromen-11(2H)-one
(55). Recovered yield 56%; mp 83-85 ꢀC. 1H NMR (300 MHz,
CDCl3, ppm): δ 1.39 (t, J=7.8 Hz, 3H, CH2CH3), 1.46 (d, J=
7.5 Hz, 3H, CHCH3), 3.13 (q, J=7.5 Hz, 2H, CH2CH3), 3.66-
3.78 (m, 1H, CHCH3), 4.47 (q, J=6.0 Hz, 1H, CH2), 5.00 (t,
J=6.6 Hz, 1H, CH2), 7.50-7.60 (m, 2H, aromatic), 7.64 (d,
J=9.3 Hz, 1H, aromatic), 7.91 (d, J=9.3 Hz, 1H, aromatic), 8.49
(d, J = 8.7 Hz, 1H, aromatic). HRMS for (Mþ þ H): calcd
281.1178, found: 281.1163.
Naphthalen-1-yl 2-Bromo-4-methylbenzoate (57). Thionyl
chloride (0.17 mL, 2.40 mmol) was added to 2-bromo-4-methyl-
benzoic acid (430 mg, 2 mmol) in DCM (3 mL) and DMF (0.1
mL), and the mixture was refluxed under nitrogen atmosphere
for 1 h. After cooling to rt, it was concentrated in vacuo to give
the title compound as a pale-yellow solid, which was used
directly in the next step.
Naphthalen-1-ol (288 mg, 2.00 mmol) was dissolved in THF
(5 mL), then DMAP (5 mg) and ethyldiisopropylamine (0.36
mL, 2.05 mmol) were added, and the mixture was cooled to 0 ꢀC
for 10 min. Freshly prepared 2-bromo-4-methylbenzoyl chlor-
ide in dry THF (10 mL) was added to the mixture via cannula,
and the resulting mixture was stirred at 25 ꢀC for 2 h, diluted
with diethyl ether (150 mL), and quenched by the addition of
water (15 mL). The organic layer was washed with HCl and
NaHCO3 and then dried (Na2SO4) and concentrated in vacuo.
The residue was purified with flash chromatography, eluting
with hexane:EtOAc = 10:1, to give 57.
tumor volume ¼ w2 ꢀ l=2
where w = width and l = length in mm of the tumor. Tumor
weight can be estimated with the assumption that 1 mg is
equivalent to 1 mm3 of tumor volume. Treatment efficacy was
determined from tumor growth delay (TGD), which is defined
as the increase in the mean TTE (the time to end point) for a
treatment group compared to the control group:
TGD ¼ T - C
expressed in days, or as a percentage of the mean TTE of the
control group:
Yield 96%. 1H NMR (300 MHz, CDCl3, ppm): δ 2.41 (s, 3H,
CH3), 7.27 (dd, J = 0.9, 8.1 Hz, 1H, aromatic), 7.40 (dd, J = 1.2,
7.5 Hz, 1H, aromatic), 7.48-7.52 (m, 3H, aromatic), 7.60 (d,
J = 0.9 Hz, 1H, aromatic), 7.77 (d, J = 8.1 Hz, 1H, aromatic),
7.87-7.90 (m, 1H, aromatic), 7.96-7.99 (m, 1H, aromatic), 8.13
(d, J = 8.1 Hz, 1H, aromatic).
T - C
C
%TGD ¼
ꢀ 100
where T = mean TTE for a treatment group, C = mean TTE for
control group 1.
9-Methyl-6H-dibenzo[c,h]chromen-6-one (58). A mixture of 57
(68 mg, 0.2 mmol), PdCl(OAc)2 (4.5 mg, 0.02 mmol), PPh3 (10.5
mg, 0.04 mmol), and NaOAc (32.8 mg, 0.4 mmol) was dissolved
in dry dimethylacetamide (10 mL), and the solution was de-
gassed and then heated to 150 ꢀC for 3 h. On cooling to rt, the
solution was diluted with diethyl ether (50 mL) and washed with
HCl, and the ethereal extracts were dried over Na2SO4. The
solution was filtered, the filtrate condensed in vacuo, and the
resulting oil purified by flash chromatography (hexane:EtOAc
4:1) to give the title compound as a white solid.
Acknowledgment. This work was supported by NIH grant
CA-17625 from the National Cancer Institute, awarded to
K. H. Lee.
Supporting Information Available: HPLC analysis and statis-
tical analysis for final compounds and preliminary results of
kinases inhibition assay for 2. This material is available free of
charge via the Internet at http://pubs.acs.org.

2308 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 5
Dong et al.
References
Evaluation of Novel 2-(Furan-2-yl)naphthalen-1-ol Derivatives as
Potent and Selective Antibreast Cancer Agents. J. Med. Chem.
2009, 52, 3586–3590.
(1) Balunas, M. J.; Kinghorn, A. D. Drug discovery from medicinal
plants. Life Sci. 2005, 78, 431–441.
(2) Saklani, A.; Kutty, S. K. Plant-derived compounds in clinical trials.
(15) Hatano, M.; Suzuki, S.; Ishihara, K. Highly efficient alkylation to
ketones and aldimines with Grignard reagents catalyzed by zinc(II)
chloride. J. Am. Chem. Soc. 2006, 128, 9998–9999.
(16) Risitano, F.; Grassi, G.; Foti, F.; Bilardo, C. A convenient synthe-
sis of furo[3,2-c]coumarins by a tandem alkylation/intramolecular
aldolization reaction. Tetrahedron Lett. 2001, 42, 3503–3505.
(17) Sha, C.-K.; Lee, R.-S.; Wang, Y. Synthesis and Diels-Alder
reactions of furo[2,3-c]pyrroles and benzofuro[2,3-c]pyrroles. Tet-
rahedron 1995, 51, 193–202.
(18) Boeckman, R. K., Jr.; Ge, P.; Reed, J. E. New Heterocyclic
Precursors for Thermal Generation of Reactive, Electron-Rich
1,2-Diaza-1,3-butadienes. Org. Lett. 2001, 3, 3647–3650.
(19) Yokoyama, M.; Menjo, Y.; Ubukata, M.; Irie, M.; Watanabe, M.;
Togo, H. Transformation of alkyl N-(vinyloxy)benzimidates to
alkyloxazoles. Mechanism and extension. Bull. Chem. Soc. Jpn.
1994, 67, 2219–2226.
Drug Discovery Today 2008, 13, 161–171.
(3) Rishton, G. M. Natural products as a robust source of new drugs
and drug leads: past successes and present day issues. Am. J.
Cardiol. 2008, 101, 43D–49D.
(4) Vuorelaa, P.; Leinonenb, M.; Saikkuc, P.; Tammelaa, P.; Rauhad,
J. P.; Wennberge, T.; Vuorela, H. Natural products in the process of
finding new drug candidates. Curr. Med. Chem. 2004, 11, 1375–1389.
(5) Tan, G.; Gyllenhaal, C.; Soejarto, D. D. Biodiversity as a source of
anticancer drugs. Curr. Drug Targets 2006, 7, 265–277.
(6) Parkin, D. M. Global cancer statistics in the year 2000. Lancet
Oncol. 2001, 2, 533–543.
(7) Coley, H. M. Mechanisms and strategies to overcome chemother-
apy resistance in metastatic breast cancer. Cancer Treat. Rev. 2008,
34, 378–390.
(8) American Cancer SocietyBreast Cancer Facts
& Figures
(20) Row, E. C.; Brown, S. A.; Stachulski, A. V.; Lennard, M. S.
Synthesis of 8-geranyloxypsoralen analogues and their evaluation
as inhibitors of CYP3A4. Bioorg. Med. Chem. 2006, 14, 3865–3871.
(21) Qabaja, G.; Jones, G. B. Annulation Strategies for Benzo-
[b]fluorene Synthesis: Efficient Routes to the Kinafluorenone
and WS-5995 Antibiotics. J. Org. Chem. 2000, 65, 7187–7194.
(22) Harayama, T.; Yasuda, H. A concise synthesis of arnottin I via
internal biaryl coupling reaction using palladium reagent. Hetero-
cycles 1997, 46, 61–64.
(23) Prat, A.; Baselga, J. The role of hormonal therapy in the manage-
ment of hormonal-receptor-positive breast cancer with co-expres-
sion of HER2. Nat. Clin. Pract. Oncol. 2008, 5, 531–542.
(24) Bowman, R. E.; Campbell, A.; Tanner, E. M. Reaction between
diphenylamine and malonic esters. J. Chem. Soc. 1959, 444–447.
(25) Abass, M.; Mostafa, B. B. Synthesis and evaluation of molluscici-
dal and larvicidal activities of some novel enaminones derived from
4-hydroxyquinolinones: Part IX. Bioorg. Med. Chem. 2005, 13,
6133–6144.
2007-2008; American Cancer Society, Inc.: Atlanta, GA, 2007.
(9) Baumann, C. K.; Castiglione-Gertsch, M. Estrogen receptor modu-
lators and down regulators: optimal use in postmenopausal women
with breast cancer. Drugs 2007, 67, 2335–2353.
(10) Amar, S.; Roy, V.; Perez, E. A. Treatment of metastatic breast
cancer: looking towards the future. Breast Cancer Res. Treat. 2008,
114, 413–422.
(11) Wang, X.; Morris-Natschke, S. L.; Lee, K. H. New developments
in the chemistry and biology of the bioactive constituents of
Tanshen. Med. Res. Rev. 2007, 27, 133–148.
(12) Wang, X.; Bastow, K. F.; Sun, C. M.; Lin, Y. L.; Yu, H. J.; Don,
M. J.; Wu, T. S.; Nakamura, S.; Lee, K. H. Antitumor Agents. 239.
Isolation, structure elucidation, total synthesis, and anti-breast
cancer activity of neo-tanshinlactone from Salvia miltiorrhiza.
J. Med. Chem. 2004, 47, 5816–5819.
(13) Wang, X.; Nakagawa-Goto, K.; Bastow, K. F.; Don, M. J.; Lin,
Y. L.; Wu, T. S.; Lee, K. H. Antitumor agents. 254. Synthesis and
biological evaluation of novel neo-tanshinlactone analogues as potent
anti-breast cancer agents. J. Med. Chem. 2006, 49, 5631–5634.
(14) Dong, Y.; Shi, Q.; Liu, Y.-N.; Wang, X.; Bastow, K. F.; Lee,
K.-H. Antitumor Agents. 266. Design, Synthesis, and Biological
(26) Hansen, M. B.; Nielsen, S. E.; Berg, K. Re-examination and further
development of a precise and rapid dye method for measuring cell
growth/cell kill. J. Immunol. Methods 1989, 119, 203–210.