1756
K. Asoh et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1753–1757
O
O
OHC
OH
Br
O
O
Br
CN
NC
NC
a
b
c
d
OH
F
MeO
F
F
Br
12
F
13
14
15
16
N
O
S
N
S O
N
F
N
N
NH
NH2
e
f
O
g
O
O
CN
CN
CN
NC
NC
NC
F
F
11f (99.5% ee)
17
18 (S : R = 20 : 1)
Scheme 2. Reagents and conditions: (a) i—4-fluoroboronic acid, Pd(OAc)2, PPh3, Na2CO3, MeOH, reflux, 14 h, ii—BBr3, CH2Cl2, 0 °C to rt, 87%; (b) i—paraformaldehyde, MgCl2,
Et3 N, THF, reflux, 2 h, ii—NBS, CHCl3, reflux, 2 h; (c) 4-cyanophenacyl bromide, K2CO3, CH3CN, reflux, 1 h, 78% from 13; (d) Zn(CN)2, Pd(PPh3)4, DMF, 100 °C, 4 h, 90%; (e) (S)-2-
methyl-2-propanesulfinamide, Ti(OEt)4, toluene, 70 °C, 2 h, 90%; (f) 5-bromo-N-methyl-imidazole, EtMgBr, CH2Cl2, THF, rt, 3 h, 80%; (g) i—4 M HCl in THF, rt, 1 h; ii—
crystallization from H2O/MeOH/EtOH, 70%.
with a carbonyl function such as a methylester (8q) amide (8r, 8v),
or CHO (8s), showed 3- to 8-fold higher enzyme inhibitory activity
(0.9–3.2 nM) than 8j (8.5 nM). X-ray crystal structure of FTase with
8w (Fig. 2) revealed a bridging water molecule forming a hydrogen
bonding between the oxygen atom of the carbonyl in the substitu-
ent R2 on the benzofuran and the amide back bone of the
Phe360B.10 Thus, the compounds with a carbonyl moiety have
strong activity against FTase (8a vs. 8w). Compounds 8t and 8x also
showed strong enzyme inhibitory activity (1.0 and 2.4 nM, respec-
tively, with FTase). These results indicate that not only a carbonyl
moiety but also a hydrogen bond accepting (HBA) group is essen-
tial to increase the enzyme inhibitory activity. 8x has strikingly
strong cellular activity (14.5 nM) even though it has no substituent
(R3 = H) on the B-ring.
lective addition of an imidazole moiety into chiral sulfineimine
17 to give 18. Although the addition of 1-methyl-2-TES-imidazole
treated with n-BuLi resulted in low diastereoselectivity (S:R = 2:1),
the Grignard condition, treated by EtMgBr, showed dramatically
improved selectivity (S:R = 20:1). This new synthetic process has
significant advantage for further scale-up synthesis because all
the intermediates can be purified by crystallization without using
column chromatography.12
In summary, we discovered a series of benzofuran compounds
showing potent FTase inhibitory activity. A cyano group at the
para-position of the A-ring (R1 position) showed excellent FTase
inhibition. Substituent of the functional group at the B-ring (R3 po-
sition) increased antiproliferative activity against human cancer
cell lines. Introduction of an HBA group at the R2 position resulted
in improved inhibitory activities both in enzyme and cellular as-
says. X-ray crystal structure of FTase with 8 k and 8w revealed
important hydrogen bonding of (1) the cyano group at the R1 posi-
tion of 8 k with Arg202 and (2) the carbonyl group at the R2 posi-
tion of 8w with a water molecule bound to the protein. Compound
11f, a clinical candidate, showed strong tumor regression in the
QG56 human NSCLC xenograft model with no noticeable body
weight loss. 11f also showed good pharmacokinetics profile in
mice. Furthermore, we established diastereoselective synthesis of
11f applicable to large scale synthesis.
The results from in vitro evaluation of compound 11 (R4 = NH2)
are shown in Table 2. With regard to the stereochemistry of the
chiral amino group, the nitro derivative with an (S)-configuration
(11a) showed 33 times higher enzyme inhibitory activity than
the corresponding (R)-isomer (11b). Furthermore, in the (S)-isomer
series, replacement of the nitro group (R2) with a cyano group and
the introduction of favorable substituents (R3) on the B-ring led us
to identify compounds (11d–f) showing potent inhibitory activities
against both FTase (0.7–1.2 nM) and QG56 cells (1.5–2.0 nM)
(Table 2).
The antitumor activity of compounds 11c–f was evaluated with
a human cancer xenograft model showing potent antitumor activ-
ity in vivo. The in vivo efficacy of 11f in human non-small cell lung
carcinoma (QG56) xenograft in mice is shown in Figure. 3.11 11f
was administered 5 times a week for 2 weeks orally at doses of
25–200 mg/kg per day. 11f showed strong tumor regression with
no noticeable body weight loss at the dose of 200 mg/kg.
After evaluation of several parameters including solubility, bio-
logical stability, and pharmacokinetics we identified 11f as a clin-
ical candidate. The pharmacokinetics data in mice of 11f is
presented in Table 3. 11f was administered intravenously at dose
of 25 mg/kg and orally at doses of 25 and 100 mg/kg. Oral bioavail-
ability of 11f was more than 50% and AUC was increased dose-
dependently. It has been confirmed that there was a good correla-
tion between AUC and tumor growth inhibition in vivo.
Acknowledgments
We are grateful to Mr. Toshihiko Fujii and Mr. Kiyoaki Sakata for
in vitro biological assays. We also thank Dr. Satoshi Sogabe for
analysis of the X-ray crystallographic data and Dr. Kazunao
Masubuchi and Ms. Miyuki Asai for synthesis of the compounds.
We thank Drs. Mamoru Suzuki and Noriyuki Igarashi and Prof. Nor-
iyoshi Sakabe at the Photon Factory for assistance with the beam
line station (BL-6B).
References and notes
1. (a) Clarke, S. Annu. Rev. Biochem. 1992, 61, 355; (b) Cox, A. D.; Der, C. J. Crit. Rev.
Oncog. 1992, 3, 365; (c) Gibbs, J. B. Cell 1991, 65, 1; (d) Newman, C. M. H.;
Magee, A. I. Biochim. Biophys. Acta 1993, 1155, 79; (e) Schafer, W. R.; Rine, J.
Annu. Rev. Genet. 1992, 26, 209.
2. Tamanoi, F.; Kato-Stankiewicz, J.; Jiang, C.; Machado, I.; Thapar, N. J. Cell.
Biochem. Suppl. 2001, 37, 64.
The chiral synthesis of 11f was established by modification of
the synthetic route shown in Scheme 1 and outlined in Scheme
2. The most dramatic improvement was made in the diastereose-