4-Hydroxy-3-methyl-6-phenylbenzofuran-2-carboxylic acid ethyl ester derivatives as potent anti-tumor agents

Article abstract of DOI:10.1016/j.bmcl.2003.10.039

Based on the structure of 4-hydroxy-3-methyl-6-phenylbenzofuran-2- carboxylic acid ethyl ester (1), which exhibits selective cytotoxicity against a tumorigenic cell line, (2,4-dimethoxyphenyl)-(4-hydroxy-3-methyl-6- phenylbenzofuran-2-yl)-methanone (18m) was designed and synthesized as a biologically stable derivative containing no ester group. Although the potency of 18m was almost the same as our initial hit compound 1, 18m is expected to last longer in the human body as an anticancer agent.

Full text of DOI:10.1016/j.bmcl.2003.10.039

Bioorganic & Medicinal Chemistry Letters 14 (2004) 455–458
4-Hydroxy-3-methyl-6-phenylbenzofuran-2-carboxylic acid ethyl
ester derivatives as potent anti-tumor agents
Ichiro Hayakawa,a Rieko Shioya,a Toshinori Agatsuma,b Hidehiko Furukawa,b
Shunji Narutoc and Yuichi Suganoa,*
aExploratory Chemistry Research Laboratories, Sankyo Co. Ltd, Shinagawa-ku, Tokyo 140-8710, Japan
bBiomedical Research Laboratories, Sankyo Co. Ltd, Shinagawa-ku, Tokyo 140-8710, Japan
cResearch Information Department, Sankyo Co. Ltd, Shinagawa-ku, Tokyo 140-8710, Japan
Received 29 September 2003; revised 22 October 2003; accepted 23 October 2003
Abstract—Based on the structure of 4-hydroxy-3-methyl-6-phenylbenzofuran-2-carboxylic acid ethyl ester (1), which exhibits
selective cytotoxicity against a tumorigenic cell line, (2,4-dimethoxyphenyl)-(4-hydroxy-3-methyl-6-phenylbenzofuran-2-yl)-metha-
none (18m) was designed and synthesized as a biologically stable derivative containing no ester group. Although the potency of 18m
was almost the same as our initial hit compound 1, 18m is expected to last longer in the human body as an anticancer agent.
# 2003 Elsevier Ltd. All rights reserved.
In our project to develop new anticancer drugs as small
molecules, 4-hydroxy-3-methyl-6-phenyl-benzofuran-2-
carboxylic acid ethyl ester (1) was selected as one of the
most promising lead compounds,1 which showed selec-
tive and potent cytotoxicity against a tumorigenic cell
line, WI-38 VA-13 subline 2RA (VA13) (EC50=40 ng/
Figure 1. Initial lead compound 1 and corresponding carboxylic acid 2.
mL) compare to the normal parental cell line, WI-38
(EC50>4000 ng/mL). In general, the ethyl ester group
in 1 is readily hydrolyzed by esterases especially in the
serum and liver, and the resultant carboxylic acid (2)
completely loses its activity to kill tumor cells (Fig. 1).
Replacing the ester group by biologically stable func-
tional groups was considered to produce more potent
chemotherapeutic agents. Therefore, further studies were
conducted to synthesize such derivatives in search of
more potent compounds containing no ester group.
ques of combinatorial chemistry as solid phase chem-
istry and solution phase parallel synthesis.
1. Evaluation of biological activity
Human tumorigenic cell line, VA13 (CCL-75.1), was
purchased from ATCC, and was passaged in MEM-E
(11095-080, Gibco BRL) containing 100 U/mL of peni-
cillin and 100 mg/mL of streptomycin supplemented
with 10% fetal calf serum (Hyclone). Exponentially
growing VA-13 cells were collected and distributed into
96-well plates (3598, Corning-Coaster) at a density of
1.5–3.0Â103 cells/well. After cultivation for two or three
days, test compounds dissolved in DMSO were diluted
in fresh medium and then added to the plates. The final
concentrations of DMSO were always kept at less than
0.5%. The plates were incubated for 24 h. Then the
culture medium was removed and washed. Finally, 150
mL/well of fresh medium was added to the cells. After
Libraries with a vast number of derivatives were desired
for lead expansion and optimization, and combinatorial
chemistry, which is a powerful tool used to generate
huge libraries, was effective for this purpose. We wish to
report herein the discovery of highly potent derivatives
of 1 that have no ester group produced by the techni-
Keywords: Anticancer.
* Corresponding author. Tel.: +81-3-3492-3131; fax: +81-3-5436-
8578; e-mail: ysugan@shina.sankyo.co.jp
0960-894X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2003.10.039

456
I. Hayakawa et al. / Bioorg. Med. Chem. Lett. 14 (2004) 455–458
cultivation for three days, 50 mL/well of MEM-E con-
taining 1 mg/mL of 2,3-bis[2-methoxy-4-nitro-5-sulfo-
phenyl]-2H-tetrazolium-5-carboxanide (XTT: X-4251,
Sigma) and 25 mM of phenazine methosulfate (P-9625,
Sigma) was added and the plates were incubated for 2.5
h. OD450 was measured by SPECTRA MAX 250
(Molecular Devices) and cell viability was calculated
using the following formula:
Cell viability ð%Þ ¼ ðWC ꢀ BGÞ ꢁ 100=ðRF ꢀ BGÞ
WC: OD450 for the well containing cells treated with
compounds
RF:
OD450 for the well containing cells and no
compounds
BG:
OD450 for the well containing neither cells nor
compounds
Scheme 1. Conversion of ethyl ester to alcohols. (a) 1.5 equiv But-
Me2SiCl, imidazole in DMF, 25 ꢀC, 1 h (quant.); (b) 1.0 equiv LiAlH4
in THF, 25 ꢀC, 30 min (73%); (c) 3.0 equiv n-Bu4NF, AcOH in THF,
25 ꢀC, 14 h (quant.); (d) 10 equiv MeMgBr in THF, 25 ꢀC, 10 min
(93%); (e) 10 equiv MnO2 in DCM, 25 ꢀC, 1 h (82%); (f) 1.5 equiv
RMgBr in THF, 0 ꢀC, 15 min (8a: 87%, 8b: quant.).
A dose–response curve was drawn and the 50% effective
concentration (EC50) was determined as an indicator of
compound cytotoxicity.
2. Chemistry
Prior to the replacement of the ester group of 1 with
another biologically stable functional group, several
ester derivatives were prepared to investigate the possi-
bility of modifying this part of the molecule without
losing biological activity. Ester derivatives (3a–d) were
synthesized from ethyl ester 1 by employing solid-phase
synthesis with the same procedure as that for the fol-
lowing amide synthesis using 2-chloro-1,3-dimethylimi-
dazolinium chloride (DMC) as a dehydrating agent.2
Various types of esters were newly synthesized and tes-
ted. All the compounds had biological activity (Table
1). This result demonstrated a prospect for further
modifying the ester group.
Scheme 2. Conversion of ethyl ester to olefins. (a) 1.5 equiv Wittig
ꢀ
reagent in THF at 0 ꢀC to 2 5 C 20 min; (b) 0.05 equiv HCl in MeOH
at 55 ꢀC 1 h (10a: 99% 2steps).
Although most of the attempts to replace the ester group
of 1 with alcohols or olefins failed, as mentioned above,
amide derivatives were still expected to maintain biologi-
cal activity as both amides and esters contain a carbonyl
group. Although amides could be hydrolyzed biologi-
cally, they were still thought to be more stable than esters.
A modification of the ethyl ester moiety of 1 into an
alcohol or related functional group was initially
attempted. The phenolic hydroxy group of 1 was pro-
tected with tert-butyldimethylsilyl (TBDMS) group, and
then, the ethyl ester was reduced to an alcohol with
lithium aluminum hydride (LAH). After removal of the
protective group, alcohol 5 was obtained, and efforts to
convert the hydroxy group of 4 to an amino group were
unsuccessful due to its instability. Tertiary alcohol 6 was
prepared using the Grignard reagent, and secondary
alcohols 8 were prepared from aldehyde 7 with
Grignard reagents (Scheme 1).
A high-throughput solid-phase synthesis (in parallel
with Quest SynthesizerTM, Algonaut Technology) based
on a technique of combinatorial chemistry was devised
to generate amide derivatives of 1. As shown in Scheme
3, lead compound 1 was introduced on Wang resin
purchased from Advanced Chemtech (100–200 mesh).
After hydrolysis of the ethyl ester with potassium
trimethylsilanolate, various types of amines were intro-
duced using N-chlorosuccinimide and triphenylpho-
sphine,3 and the following acid cleavage from the
polymer support under acidic conditions yielded the
desired amides. This solid-phase synthesis conducted in
only two steps was still effective to prepare large-size
compound libraries.
For olefin synthesis, aldehyde 9 was prepared by a pro-
cedure similar to that for aldehyde 7 synthesis using the
methoxymethyl (MOM) group instead of TBDMS as a
protective group for the phenolic hydroxy group. By the
Wittig reaction, the aldehyde was converted to olefins
10a–d (Scheme 2).
Although various types of amides with aliphatic, aro-
matic, and heterocyclic amines were synthesized, only a
few derivatives (15a for example) that showed very
weak activity were discovered (Fig. 2).
For these alcohols and olefins, which lacked a carbonyl
group, cytotoxicity was no longer observed. This result
suggested that the carbonyl group of 1 is essential for
the biological activity.
During our attempt to replace the ester group of 1 with
a more stable functional group, ketones were finally

I. Hayakawa et al. / Bioorg. Med. Chem. Lett. 14 (2004) 455–458
Table 1. Ester derivatives and biological activity
457
Esters
ROH
Yielda (%)
16
Cytotoxicityb
EC50, (ng/mL)
3a
60
55
7
Scheme 4. Solution-phase synthesis of ketones. (a) 2equiv methoxy-
methyl chloride, NaH in DMF, 0 ꢀC, 1 h; (b) 4 equiv 1 N NaOH in
EtOH, 60 ꢀC, 1 h; (c) 1.5 equiv N,O-dimethylhydroxylamine hydro-
chloride, 1.5 equiv 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide
hydrochloride, 1 equiv 1-hydroxybenzotriazole hydrate, 2equiv tri-
ethylamine in DMF, 25 ꢀC 16 h (93% from 1); (d) 3.0 equiv RMgCl,
3b
3c
3d
10
2
2
0
3.0 equiv RLi, or 1–40 equiv RMgBr in THF; (e) HCl in THF–MeOH
(1:1).
30
21
shown in Table 2 (phenyl ketones) and in Table 3 (other
ketones), quite a few ketone derivatives showed biolo-
gical activity.
a Overall yields based on used solid support after purification by silica
gel chromatography.
b Selectivity against tumor cells was confirmed.
Although phenyl ketone 18a with no substituent on
the benzene ring showed relatively weak cytotoxicity,
its derivatives, which had a substituent at the ortho
position (18b, c, and d) were moderately active. The
substituent groups at the meta position (18e, f, and
g) were relatively less effective even with an ortho
(18o) or a para (18q and r) substituent, and deriva-
tives that had a para substituent (18i, j, k, and l) had
completely lost their activity except for those with a
methoxy group substituent (18h). The methoxy group
was considerd to be a desirable substituent for this series
of compounds, and, as expected, the phenyl ketone 18m
(EC50=40 ng/mL),5 which has two methoxy sub-
stituents at the ortho and para positions, was the most
potent compound.
Scheme 3. Solid-phase synthesis of amides. (a) 10 equiv methane-
sulfonyl chloride, 10 equiv NEt3 in DCM, 25 ꢀC, 18 h; (b) 3 equiv 1, 5
equiv K2CO3 in DMF, 60 ꢀC, 22 h; (c) 10 equiv Me3SiOK in THF,
25 ꢀC, 20 h; (d) 10 equiv RNH2, 10 equiv Ph3P, 10 equiv N-chloro-
succinimide, pyridine in DCM, 25 ꢀC, 10 min; (e) 20% TFA in DCM,
25 ꢀC, 2h.
Phenyl ketone 18m does not contain any functional
groups suspected to be biologically unstable as esters,
and its potency was appeared to be comparable to ori-
ginal ethyl ester 1. Although ethyl ester 1 was readily
decomposed in mouse liver microsomes, phenyl ketone
18m was confirmed to be very stable (Fig. 3).
Figure 2. Cytotoxicity of amide derivatives.
tested. Scheme 4 shows the solution phase synthetic
method employed for the ketone synthesis. N,O-dime-
thylhydroxylamide 16 was prepared from ethyl ester 1
via carboxylic acid using 1-(3-dimethylaminopropyl)-3-
ethylcarbodiimide (EDC) and 1-hydroxybenzotriazole
(HOBt), and, then, 16 was allowed to react with
Grignard reagents or alkyl (or aryl) lithium reagents to
be converted into corresponding ketones 17.4 The
methoxymethyl protective group was removed under
acidic conditions to generate the desired ketones 18. As
Figure 3. Stability test of ethyl ester 1 and ketone 18m in mouse liver
microsomes. The stability of each compound (20 mmol) was tested in
mouse (BDF1) liver microsomes (1 mg protein in 1 mL). The concen-
tration of remained compounds was determined by HPLC.

458
I. Hayakawa et al. / Bioorg. Med. Chem. Lett. 14 (2004) 455–458
Table 2. Cytotoxicity of substituted phenyl ketone derivatives
Compd
R1
R2
R3
R4
Yield (%)
Cytotoxicity EC50, ng/mLb
to 16
to 18
18aa
18b
18c
18d
18e
18f
18g
18h
18i
18j
18k
18l
18m
18n
18o
18p
18q
18r
H
OMe
CH2OMe
Ph
H
H
H
H
CF3
OPri
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
OMe
H
H
H
—
88
68
79
—
81
Quant.
Quant.
600
90
100
100
H
H
H
H
H
H
H
H
OMe
OMe
OMe
CF3
H
7298
>4000
350
90
H
H
Quant.
Quant.
Quant.
Quant.
Quant.
Quant.
79
Quant.
Quant.
90
97
83
89
29
81
45
51
73
51
89
89
93
CH2NMe2
H
H
H
H
H
H
H
H
H
OMe
OPh
Et
NMe2
CH2NMe2
OMe
CF3
H
280
>4000
>4000
>4000
>4000
40
260
90
400
450
OMe
OMe
F
71
Quant.
9247
OMe
F
H
400
a Synthesized from 8b, which was oxidized by MnO2 in DCM at 25 ꢀC for 4 h (47%) and deprotected using n-Bu4NF under standard conditions (23%).
b Selectivity against tumor cell was confirmed.
Table 3. Cytotoxicity of other ketone derivatives
In conclusion, phenyl ketone 18m was discovered from
the derivatives of a screening lead 1. As the structure of
18m was designed by replacing biologically unstable
ester group of 1 with phenyl ketone, it is predicted to be
a more promising compound as an anticancer drug.
Compd
R
Yield(%)
Cytotoxicity
EC50, ng/mLa
to 16
to 18
Acknowledgements
18A
18B
18C
18D
18E
18F
18G
18H
18I
18J
18K
18L
18M
18N
Me
Et
1-Pr
2-Pr
1-Bu
tert-Bu
cyclo-Hex
2-pyridyl
2-thiophenyl
2-thiazolyl
5-pyrimidyl
3-quinolyl
CH2Ph
88
88
Quant.
46
54
1273
34
57
88
62
59
53
34
85
54
48
25
70
80
270
590
410
The authors are grateful to Ms. Youko Oda for her
technical support in the cytotoxicity test and Ms.
Kumiko Koyama in the stability test in mouse liver
microsomes.
2000
410
79
45
88
91
450
600
2000
400
References and notes
62700
44
51
76
>4000
>4000
>4000
1. Grinev, A. N.; Lyubchanskaya, V. M.; Uretskaya, G. Y.;
Vlasova, Y. F. Khim. Geterotsikl. Soedin 1976, 7, 879.
2. Isobe, T.; Ishikawa, T. J. Org. Chem. 1999, 64, 6984.
3. Nishida, A.; Fuwa, M.; Naruto, S.; Sugano, Y.; Saito, H.;
Nakagawa, M. Tetrahedron Lett. 2000, 41, 4791.
CH2CH2Ph
Quant.
a Selectivity against tumor cell was confirmed.
4. Nahm, S.; Weinreb, S. M. Tetrahedron Lett. 1981, 22, 3815.
5. Compound 18m: colorless crystal. Mp 200–202 ꢀC, 1H
NMR (270 MHz, CDCl3) d 7.59–7.57 (m, 2H), 7.51 (d,
1H, J=8.3 Hz), 7.44 (t, 2H, J=7.3 Hz), 7.38–7.35 (m,
1H), 7.25 (s, 1H), 6.81 (s, 1H), 6.58 (dd, 1H, J=1.9 Hz,
8.7 Hz), 6.53 (d, 1H, J=2.2 Hz), 5.43 (s, 1H), 3.89 (s, 3H),
3.77 (s, 3H), 2.71 (s, 3H); MS (EI, m/z) 388 (M)+.
6. Compound 18a: colorless crystal. Mp 247–249 ꢀC (dec.),
1H NMR (270 MHz, CDCl3+CD3OD) d 761 (d, 2H,
J=7.1 Hz), 7.48–7.35 (m, 4H), 7.24 (s, 1H), 6.81 (s, 1H),
2.81 (s, 3H), 2.60 (s, 3H); MS (EI, m/z) 266 (M)+.
Introduction of a trifluoromethyl group instead of a
methoxy group at the para position (18n) weakened the
biological activity, and placing a methoxy group at the
meta position (18o) did not change the cytotoxicity of
18b. Methyl ketone 18A6 (EC50=80 ng/mL) was a
compound with potencies similar to 18m. Derivatives of
heterocycles (18H, J, and K) had almost same activity
as unsubstituted phenyl ketone 18a.