A library synthesis of 4-hydroxy-3-methyl-6-phenylbenzofuran-2-carboxylic acid ethyl ester derivatives as anti-tumor agents

Article abstract of DOI:10.1016/j.bmcl.2004.06.067

A compound library for structure optimization was synthesized employing solution-phase parallel synthesis to find highly potent derivatives. As a result of a hit-to-lead program using a technique of solution-phase parallel synthesis, a highly potent (2,4-dimethoxyphenyl)-[6-(3-fluorophenyl)-4-hydroxy- 3-methylbenzofuran-2-yl]methanone (15b) was synthesized as an optimized derivative of 4-hydroxy-3-methyl-6-phenylbenzofuran-2-carboxylic acid ethyl ester (1), which was discovered as a screening hit from small-molecule libraries and exhibited selective cytotoxicity against a tumorigenic cell line.

Full text of DOI:10.1016/j.bmcl.2004.06.067

Bioorganic & Medicinal Chemistry Letters 14 (2004) 4383–4387
A library synthesis of 4-hydroxy-3-methyl-6-phenylbenzofuran-
2-carboxylic acid ethyl ester derivatives as anti-tumor agents
Ichiro Hayakawa,a Rieko Shioya,a Toshinori Agatsuma,b Hidehiko Furukawa,b
Shunji Narutoc and Yuichi Suganoa,
*
aLead Discovery Research Laboratories, Sankyo Co. Ltd, Shinagawa-ku, Tokyo 140-8710, Japan
bBiomedical Research Laboratories, Sankyo Co. Ltd, Shinagawa-ku, Tokyo 140-8710, Japan
cR&D Information Technology Department, Sankyo Co. Ltd, Shinagawa-ku, Tokyo 140-8710, Japan
Received 2 June 2004; revised 17 June 2004; accepted 21 June 2004
Abstract—As a result of a hit-to-lead program using a technique of solution-phase parallel synthesis, a highly potent (2,4-dimeth-
oxyphenyl)-[6-(3-fluorophenyl)-4-hydroxy-3-methylbenzofuran-2-yl]methanone (15b) was synthesized as an optimized derivative of
4-hydroxy-3-methyl-6-phenylbenzofuran-2-carboxylic acid ethyl ester (1), which was discovered as a screening hit from small-mol-
ecule libraries and exhibited selective cytotoxicity against a tumorigenic cell line.
Ó 2004 Elsevier Ltd. All rights reserved.
1.Introduction
tigated by the systematic modification of substructure C
employing techniques of combinatorial chemistry. The
study resulted in discovering a highly potent compound,
derivative 2, which was metabolically very stable and
lasted much longer than the original ethyl ester (1) in
mouse liver microsomes.1 In this communication, we re-
port the synthetic modification of substructure A, which
was still left untouched, and the biological activity of the
synthesized derivatives (Fig. 1).
As described in our preceding reports, a potential hit
compound 4-hydroxy-3-methyl-6-phenylbenzofuran-2-
carboxylic acid ethyl ester (1) was discovered from the
screening of small-molecule libraries in search for novel
anti-tumor drugs. The hit compound 1 exhibited potent
and selective cytotoxicity against a tumorigenic cell line,
WI-38 VA-13 subline 2RA (VA-13) (EC50 =40ng/mL),
and no cytotoxicity was observed against the normal cell
line, WI-38 (EC50 >4000ng/mL).1–3
2.Chemistry
Substructures B and C of compound 1 were already op-
timized to improve the biological activity and the meta-
bolic tolerance. As substructure B, which consists of the
central benzofuran ring and its substituents, was sug-
gested to be responsible for the constitutive expression
of biological activity, modification of substructure B
was considered a challenge.3 The biologically labile ethyl
ester of substructure C was revealed to be replaceable by
phenyl ketones, which are generally tolerant against
hydration and metabolic decomposition.1 The relation-
ship between substructure C and cytotoxicity was inves-
The synthetic procedure of compound 1 is shown
in Scheme 1. Starting from a cyclohexane-1,3-dione
OH
Me
O
O
O
A
B
C
1: EC50 = 40 ng/mL
Keyword: Anti-cancer.
*
Figure 1. Structure of
screening.
a hit compound of an anti-tumor drug
Corresponding author. Tel.: +81-3-3492-3131; fax: +81-3-5436-
8578; e-mail: ysugan@sankyo.co.jp
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.06.067

4384
I. Hayakawa et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4383–4387
O
O
Me
O
O
O
O
a, b
+
O
O
Cl
Ph
Ph
O
3
5
4
c, d
OH
Me
O
O
O
Ph
1
Scheme 1. Reagents and conditions: (a) 1.5equiv NaH (55% disper-
sion in mineral oil) in THF, 25°C, 14.0h (64%); (b) in AcOH, reflux,
1.0h (82%); (c) 1.05equiv LiN(Pri)2 in THF, À78°C, 1.0h, and then
1.7equiv Me3SiCl, À78 to 25°C, 1.0h; (d) 0.5equiv Pd(OAc)2
and 0.5equiv p-benzoquinone in MeCN, 25°C, 1.5h (42%, two steps
from 5).
Scheme 2. Using a simple reaction vessel, cyclohexane-1,3-dione was
dissolved in aqueous 2M potassium hydroxide (1.0equiv), and then, a
methanol solution of chloroketone (4) (1.0equiv) was added. After the
condensation–cyclization reaction was completed, the dehydration by
adding acetic acid (large excess) and heating for 2h was carried out to
generate a cyclohexenone 7.
with a phenyl group as a substituent at the fifth
position (3), 3-methyl-4-oxo-6-phenyl-4,5,6,7-tetra-
hydrobenzofuran (5) was prepared, and the subsequent
oxidation with palladium acetate converted the cyclo-
hexenone 5 to phenol 1.
For the construction of a combinatorial library focusing
on compounds with a modified substructure C of com-
pound 1, a synthetic procedure employing solid-phase
chemistry had been developed.1 Although solid-phase
chemistry produced numerous outputs for library syn-
thesis, a solution-phase parallel synthesis was chosen
for the library construction. Since the reaction steps
were relatively short and the substructure to be modified
was one substituent, it was difficult to take advantage of
solid-phase chemistry. In addition, solid-phase synthe-
sis, in general, requires extra reaction steps for the intro-
duction of the starting molecule onto the solid support
and for the cleavage and removal of the synthesized
compound (Fig. 2).
The moisture labile sodium hydride of the first step was
replaced with an aqueous solution of potassium hydrox-
ide. The subsequent dehydration was carried out by sim-
ply adding acetic acid to the reaction mixture without
further work after the preceding condensation reaction
to obtain cyclohexenone 7 in good yield (Scheme 2).
The following oxidation of cyclohexenone 5 using palla-
dium acetate via trimethylsilyl enol ether was simplified
by utilizing bromine under basic conditions (Scheme 3).
As the developed procedure required no moisture-labile
reaction steps, our original Ôreaction stationÕ, which is
arrays of reaction vessels placed in a heat block com-
bined with an orbital shaker, was available for the lib-
rary synthesis. This optimized procedure for a library
The original synthetic procedure of compound 6
employed air and moisture sensitive reagents such as
sodium hydride under strictly controlled reaction condi-
tions, and this was the major issue to resolve for the par-
allel solution-phase synthesis. For the library synthesis,
simple shape reaction vessels that were compact enough
to carry out a large number of reactions in a limited
workspace were desired. Labile reagents, which require
space-consuming apparatuses and complex handling,
were to be replaced.
OH
Me
O
OMe
O
Scheme 3. Cyclohexenone 7 was dissolved in a chloroform solution of
bromine (1.1equiv). The solution was mixed with a dimethylforma-
mide solution of lithium bromide (12equiv) and lithium carbonate
(12equiv).
OMe
2: EC50 = 40 ng/mL
Figure 2. The structure of a potent and metabolically stable derivative.

I. Hayakawa et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4383–4387
4385
synthesis did not demand any highly sophisticated
library synthesizers, which have capability of operating
complicated reaction procedures.
ents placed at the ortho-position also weakened the
activity (9a,c through 9f,C, and 9D), the effect of intro-
ducing substituents at the meta-position was relatively
smaller. Most of the meta-substituted derivatives (9h
through 9m, and 9o) showed almost the same or a
slightly weaker cytotoxicity than the unsubstituted com-
pound 1.
3.Structure–activity relationship
Derivatives of compound 1, which had various substitu-
ents on the benzene ring at the sixth position of benzo-
furan, were synthesized, and their cytotoxicity were
evaluated. As shown in Table 1, most of the compounds
that had a substituent at the para-position of the ben-
zene ring had completely lost their biological activity
(9s through 9B) except when the substituent was a hyd-
roxy group (9q) or fluoride (9r). While bulky substitu-
The synthesis of derivatives that have heterocycles or
aliphatic substituents instead of the benzene ring was
problematic because of the instability of the intermedi-
ates. A few examples, however, shown in Table 2, indi-
cated that derivatives with heterocycles had moderate
cytotoxicity and those with aliphatic substituents had
lost their activity.
Table 1. Cytotoxicity of ester derivatives possessing a substituted benzene ring
OH
Me
O
R1
R2
R3
O
O
R4
9
Compds
R1
R2
R3
R4
Yield
Cytotoxicity EC50 (ng/mL)c
To 7a
To 9b
1
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Cl
Cl
58
46d
14
33
26
11
22
––
––
57
2
91
60
50
72350
19
27
28
70f
99g
49
5
40
310
48
9a
9b
9c
9d
9e
9f
HO
F
H
H
H
H
Cl
Br
MeO
BnO
H
H
H
H
H
390
380
H
H
H
H
>4000
520
90
9g
9h
9i
HO
NH2
NO2
F
H
H
H
H
H
75
9j
H
H
40
39
17
80
9k
9l
H
Cl
H
2
5
H
Br
Me
CF3
MeO
BnO
H
H
2
2
2
4
9m
9n
9o
9p
9q
9r
H
H
13
38
20
44
46
53
45
34
47
72e
76e
42e
58
16e
38
28
38
38
62
63
18
63
60
52370
54
61
56
9
80
700
H
H
H
H
80
850
75
H
H
H
HO
F
H
H
9s
H
H
Cl
Br
Me
MeO
EtO
PniO
BnO
>4000
>4000
>4000
>4000
>4000
>4000
>4000
>4000
>4000
>4000
>4000
>4000
9t
H
H
9u
9v
9w
9x
9y
9z
9A
9B
9C
9D
H
H
H
H
H
H
76
54
38
67
53
12
55
40
H
H
H
H
H
H
PhCH2CH2O
Cl
Cl
H
H
MeO
H
MeO
H
F
Cl
H
H
a Yields from diketone 6 after purification by chromatography.
b Yields from 7 after purification by chromatography.
c Selectivity against tumor cells was confirmed.
d Yield was calculated after the removal of benzyl protecting group for the hydroxy group.
e Yield from 7 corresponding to 9e.
f Synthesized from 9p.
g Synthesized from 9i.

4386
I. Hayakawa et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4383–4387
Table 2. Cytotoxicity of ester derivatives possessing an aromatic
heteroring or saturated substituent
responding ketone by the same procedure, and its cyto-
toxicity was evaluated.
OH
Me
The results of the cytotoxicity test of ketone derivatives
and related compounds are shown in Table 3. The orig-
inal hit compound 1 was converted to dimethoxyphenyl
ketone 2 without losing its biological activity.1 The bio-
logical activity of dimethoxyphenyl ketone 15b corre-
sponding to ethyl ester 9j was slightly weaker than the
original ester 9j but was still more potent than com-
pound 2. Considering the gastrointestinal absorption,
3-methoxypyridin-2-yl ketone was recognized as another
promising structure for its effect of reducing the MlogP
(Moriguchi logP) value and providing remarkable cyto-
toxicity to compound 15a.4 Integrated compound 15c
with fluorophenyl group and dimethoxyphenyl ketone
had almost the same potency as the dimethoxyphenyl
ketone 15b.
O
O
10
O
R
Compds
R
Yield
To 7a To 10b
Cytotoxicity EC50
(ng/mL)c
10a
10b
10c
10d
2-Furanyl 44
2-Thiophenyl 37
2d
5d
180
160
cyclo-Hex
Me3CCH2
47
35
31
6
>4000
>4000
a Yields from diketone 6 after purification by chromatography.
b Yields from 7 after purification by chromatography.
c Selectivity against tumor cell was confirmed.
d Heterocycles were unstable under oxidative condition.
Finally the oxygen atom in benzofuran was replaced with
nitrogen to confirm the biological activity of this core
structure. Indole derivative 14 was synthesized as shown
in Scheme 4 starting from the same substituted cyclohex-
ane-1,3-dione 3 used to synthesize benzofuran deriva-
tives. Phenyl ketone 15d was synthesized via the
corresponding ester using the same procedure as that
mentioned above. Although the biological activity of
the indole derivatives was, unfortunately, relatively weak,
We have already reported that the ethyl ester group of
compound 1 was readily hydrolyzed by esterases espe-
cially in the serum and liver, and the resultant carboxylic
acid had completely lost its activity to kill tumor cells.1
Substituted phenyl ketone derivatives corresponding to
compound 1 were synthesized and dimethoxyphenyl ke-
tone 2 was found to be a metabolically stable derivative.
The most potent ethyl ester 9j was converted to the cor-
Table 3. Cytotoxicity against VA13 cell assay for ketone derivatives and related compounds
OH
Me
O
15
X
R2
R1
Compds
X
O
R1
R2
MlogPa
Cytotoxicity EC50 (ng/mL)b
40
1
OEt
2.6
2
O
O
2.6
2.2
40
28
O
O
N
15a6
O
F
F
9j
O
O
OEt
2.7
2.7
17
30
15b7
O
O
N
F
15c8
O
2.3
27
O
149
NH
NH
OEt
2.6
2.7
470
120
F
15d10
O
O
a Selectivity against tumor cells was confirmed.
b Moriguchi logP.

I. Hayakawa et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4383–4387
4387
Acknowledgements
O
O
O
Me
Me
O
O
O
O
a, b
c
The authors are grateful to Ms. Youko Oda for her
technical support in the cytotoxicity test.
N
S
N
Ph
O
Ph
Ph
H
O
O
3
11
References and notes
12
O2N
Me
1. Hayakawa, I.; Shioya, R.; Agatsuma, T.; Furukawa, H.;
Naruto, S.; Sugano, Y. Bioorg. Med. Chem. Lett. 2004, 14,
455.
d, e
OH
OH
Me
O
O
O
f
2. US Patent Application US60/306116, July 17 2001.
3. Hayakawa, I.; Shioya, R.; Agatsuma, T.; Furukawa, H.;
Sugano, Y. Bioorg. Med. Chem. Lett. 2004, 14, 3411.
4. Lipinski, C. A. JSSX Symposium, Tokyo, April 2003.
5. Kan, T.; Fukuyama, T. Chem. Commun. 2004, 353.
6. Compound 15a: two steps from 1 (46%). Yellow powder.
N
N
O
Ph
Ph
H
O
S
O
14
13
O2N
1
Mp 243°C, H NMR (270MHz, CDCl3): d 8.35 (d, 1H,
Scheme 4. Synthesis of indole derivatives. Reagents and conditions: (a)
1.0equiv ethyl acetylacetate, 1.1equiv NaNO2 in AcOH, 10°C, 3.0h;
(b) 2.2equiv zinc powder in AcOH, 60°C, 3.0h, 20% from 3; (c)
1.2equiv ButOK and then 2.5equiv p-NO2PhSO2Cl in THF, 0°C,
1.0h; (d) 1.0equiv Br2 in CHCl3, 2 5°C, 2.0h, 48% from 11; (e)
4.0equiv LiBr, 4.0equiv Li2CO3 in DMF, 130°C, 1.0h; (f) 1.2equiv
PhSH, 3.0equiv K2CO3 in DMF, 25°C, 1.0h, 65% from 13 (see
Ref. 5).
J=3.0Hz), 7.91 (s, 1H), 7.49–7.26 (m, 7H), 6.91 (s, 1H),
6.69 (s, 1H), 3.90 (s, 3H), 2.53 (s, 3H); MS (APCI, m/z) 360
(M+1)+.
7. Compound 15b: two steps from 9j (59%). Yellow powder.
Mp 194°C, 1H NMR (270MHz, acetone-d6): d 7.89 (s, 1H),
7.39–7.27 (m, 4H), 7.15 (d, 1H, J=1.4Hz), 7.05–6.97 (m,
1H), 6.88 (d, 1H, J=1.4Hz), 6.59–6.53 (m, 2H), 3.79 (s,
3H), 3.65 (s, 3H), 2.55 (s, 3H); MS (APCI, m/z) 407 (M+1)+.
8. Compound 15c: two steps from 9j (49%). Yellow powder.
Mp 226°C, 1H NMR (270MHz, acetone-d6): d 8.13 (d,
1H, J=4.6Hz), 7.89 (s, 1H), 7.43 (d, 1H, J=4.6Hz), 7.40
(d, 1H, J=4.6Hz), 7.38–7.29 (m, 4H), 7.12 (d, 1H,
J=1.4Hz), 7.16–6.98 (m, 1H), 6.90 (d, 1H, J=1.4Hz),
3.75 (s, 3H), 2.49 (s, 3H); MS (APCI, m/z) 378 (M+1)+.
9. Compound 14: yellow powder. Mp 160°C, 1H NMR
(270MHz, CDCl3): d 8.69 (s, 1H), 7.61–7.31 (m, 5H), 7.11
(d, 1H, J=1.4Hz), 6.64 (d, 1H, J=1.1Hz), 4.42(q, 2H,
J=7.1Hz), 2.84 (s, 3H), 1.43 (t, 3H, J=7.0Hz); MS
(APCI, m/z) 296 (M+1)+.
their cytotoxicity was maintained. The indole derivatives
will be applied to a higher level of biological tests as com-
pounds different from benzofuran derivatives.
In conclusion, using a technique of solution-phase
library synthesis, various substituents were introduced
on the phenyl ring at the sixth position of the benzofu-
ran core structure of compound 1. By further introduc-
ing substituent at the ester moiety of 1, highly potent
derivatives 15b and 15c were discovered in addition to
phenyl ketones 2 and 15a.
10. Compound 15d: two steps from 14 (36%). Yellow powder.
1H NMR (270MHz, CDCl3): d 8.93 (s, 1H), 7.39–7.26 (m,
5H), 7.10 (s, 1H), 7.06–7.03 (m, 1H), 6.59 (s, 1H), 3.88 (s,
3H), 3.79 (s, 3H), 2.38 (s, 3H); MS (APCI, m/z) 406 (M+1)+.