Synthesis of N,N′-disubstituted 3-aminobenzo[c] and [d]azepin-2-ones as potent and specific farnesyl transferase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2003.11.013

A structure-activity study was performed by synthesis on N,N′-disubstitution of 3-aminobenzo[c] and [d]azepin-2-one 2 and 3 to afford potent and specific farnesyl transferase inhibitors with low nM enzymatic and cellular activities.

Full text of DOI:10.1016/j.bmcl.2003.11.013

Bioorganic & Medicinal Chemistry Letters 14 (2004) 767–771
Synthesis of N,N0-disubstituted 3-aminobenzo[c] and [d]azepin-
2-ones as potent and specific farnesyl transferase inhibitors
Thierry Le Diguarher,a Jean-Claude Ortuno,a David Shanks,a Nicolas Guilbaud,b,y
Alain Pierre,b Eric Raimbaud,a Jean-Luc Fauchere,a,, John A. Hickman,b
Gordon C. Tuckerb and Patrick J. Casaraa,*
aDepartment of Medicinal Chemistry, Institut de Recherches Servier, 125 chemin de Ronde, 78290Croissy sur Seine, France
bDepartment of Experimental Oncology, Institut de Recherches Servier, 125 chemin de Ronde, 78290Croissy sur Seine, France
Received 20 August 2003; revised 4 November 2003; accepted 7 November 2003
This paper is dedicated to the memory of Professor Jean-Luc Fauchere, who left us prematurely after a long illness, on 30 May 2003
Abstract—A structure–activity study was performed by synthesis on N,N0-disubstitution of 3-aminobenzo[c] and [d]azepin-2-one 2
and 3 to afford potent and specific farnesyl transferase inhibitors with low nM enzymatic and cellular activities.
# 2003 Elsevier Ltd. All rights reserved.
1. Introduction
endowed with new putative interactions by comparison
Protein Farnesyl Transferase (FTase) is a zinc depen-
dent enzyme that catalyses the attachment of a farnesyl
lipid group to the sulfur atom of a cysteine residue of
numerous proteins involved in cell signalling, including
the oncogenic Ras protein.1 This lipid side chain is cri-
tical for the cell membrane anchoring of these proteins,2
and therefore FTase inhibition has been recognised as a
valuable antitumoral therapeutic approach.3 Recently,
we described caprolactam derivatives 1 as bioisostere of
the consensus terminal sequence CVIM of the Ras pro-
tein. Selected compounds were found to be very potent
FTase inhibitors with low nM IC50, and their activities
were validated in vitro in a cellular assay, and in vivo in
an antitumoral model.4 Conformational analysis of this
series of compounds is in agreement with the crystal
structures of the complex FTase/CVIM in an extended
conformation in the active site.6 The presence of an
hydrophobic cleft formed by aromatic residues in the
active site was confirmed by the van der Walls interac-
tions observed with the phenylalanine residue of CVFM
and the peptidomimetic L-739,750.5 These findings
prompted us to investigate benzo[c] and [d]azepinones, 2
and 3 respectively, as new cores for FTase inhibitors
with our previous caprolactam scaffold. These struc-
tures included the proximal N-p-cyanobenzyl substitu-
tion on the imidazole as a zinc-complexing function and
the SAR was focused on substituent R on the lactam
nitrogen atom to replace the lipophilic non-polar side
chains of the isoleucine and/or methionine (Fig. 1).
2. Chemical methods
2.1. Synthesis of benzo[c]azepin-2-ones 2
In order to validate this hypothesis and investigate the
N-lactam substitution, the chemical strategy was based
on two selective N-alkylations of the benzolactam 6.4
Since the S stereospecificity of the inhibition was
observed for 1, the S-phenyl alanine carboxamide 4 was
used as the starting material (route 1). Protection of the
primary amine as a methyl carbamate, allowed the for-
mation of the relatively stable acetoxymethylamide 5 in
quantitative yield.7 The ring closure of these compounds
was achieved smoothly at room temperature by acidic
catalysis to afford the desired lactam 6.8 In order to
compare this series with the most active type 1 com-
pounds, the various substitutions R were concentrated
on aryl, benzyl and cyclohexylmethyl derivatives. The
N-alkylation of 6 gave access to a large variety of sub-
stituted amines 7a–i. The N-arylation was achieved by
* Corresponding author. Tel.: +33-1-5572-2361; fax: +33-1-5572-
2470; e-mail: patrick.casara@fr.netgrs.com
y
Present address: Oncodesign, Parc Technologique Toison d’Or, 28
rue Louis de Broglie, 21000 Dijon, France.
0960-894X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2003.11.013

768
T. Le Diguarher et al. / Bioorg. Med. Chem. Lett. 14 (2004) 767–771
treatment of 6 with various triarylbismuth compounds
to give 7j, 7m, 7p and 7o.9 However, the N-arylation was
dependent on the formation of the triarylbismuth
reagent which is not an easy process and often not
compatible with all the desired substitutions. Therefore,
a more general approach was developed based on selec-
tive substitutions of a,a0-dibromo-o-xylene (10, route 2).
The mono alkylation of the bulky N-Boc-amino-
malonate 9 was easily controlled to give the adduct 11
without any cyclic side product obtained with the acet-
amide.10 This key intermediate allowed the introduction
of all kinds of primary amines by substitution of the
second benzyl bromide to afford compounds 12. The
lactam formation was achieved under basic conditions
to give, after decarboxylation and deprotection, the
amines 7j–o in their racemic form. All these amines 7a–
q were converted by reductive amination with the
aldehyde 8 to the desired products 2a–q (Scheme 1,
Table 1).11
tam afforded, after acidic deprotection, the amines 16a–
j which were converted to the corresponding products
3a–j by reductive amination with the aldehyde 8
(Scheme 2, Table 2).11
3. Biological evaluation
The inhibitory activities of these compounds were
examined in a binding assay against FTase and their
cellular activities were determined on a cell line trans-
fected with the v-H-Ras gene13 as described.4 The results
were compared to the clinical reference BMS-21466214
(Table 1).
In the benzo[c]azepinone series, the introduction of
selected benzyl substituents R afforded very potent
FTase inhibitors with almost the same potency in the
cell based assay (compounds 2a–i, Table 1). By com-
parison, the corresponding caprolactam series 1 showed
the same range of activity and even an identical SAR
with, at least an one order of magnitude lower potency
for para-substituted derivatives (e.g., 2e and 2f). Aryl
substitutions showed no major difference in potency
with the corresponding benzyl derivatives in this series
(e.g., 2a and 2j; 2d and 2l; 2g and 2p). Therefore, the
introduction of a fused phenyl ring in the structure 1 to
give benzo[c]azepinone 2, did not alter, nor improve
significantly the potency of these compounds.
2.2. Synthesis of benzo[d]azepin-2-ones 3
All attempts to obtain the corresponding benzo[d]aze-
pinone 16 in a chiral form from a S-phenylglycine by a
sequence similar to Scheme 1, failed. Therefore, the
ring closure was achieved on the other side of the
molecule by using the procedure described by Ben-
Ishai.12 The bis carbamate 14, easily obtained by heat-
ing in methanol phenethylamine and the glyoxalique
derivative 13, was stirred in methane sulfonic acid to
give the lactam 15 in the racemic form. By similar pro-
cedures, benzylic alkylation and arylation of the N-lac-
In the benzo[d]azepinone series, the introduction of
selected benzyl or aryl substituents R afforded less
Figure 1.
Scheme 1. Synthesis of compounds 2. Reagents and conditions: (i) (a) ClCO2Me, CH2Cl2 pyridine, 0 ꢀC, rt, 24 h; (b) (CHO)n, (Ac)2O, AcOH, rt, 24
h (5, 90%); (ii) TfOH, CH2Cl2, rt, 24 h (6, 77%); (iii) (a) NaH, RBr, Bu4NI, rt, 24 h; (b) HBr, AcOH, (S, 7a–i); (iii) BiPh3, pyridine, Cu(OAc)2,
DMF, CH2Cl2, rt, 24 h (S, 7j, 7l, 7p and 7q); (iv) EtONa, EtOH, rt, 12 h, (10, 88%); (v) RNH2, THF, NEt3, reflux; (vi) (a) KOH; (b) EDC; (c) LiI,
pyridine, reflux; (d) HCl, ether, (RS, 7j–o); (vii) (a) NaBH(OAc)3, CH2Cl2, 0 ꢀC, rt, 12 h; (b) HCl, ether, or fumaric acid, EtOH (2a–q).

T. Le Diguarher et al. / Bioorg. Med. Chem. Lett. 14 (2004) 767–771
769
Table 1. Structures and yields of compounds 7 and 2 and biological data of compounds 2
R
7 (conf.)
Yield (%)
2 (conf.)
Yield (%)
FTase
IC50 nMꢂSEM (n)
Ras cell
R
7 (conf.)
Yield (%)
2 (conf.)
Yield (%)
FTase
IC50 nMꢂSEM (n)
Ras cell
BMS-2146623 found
1.3
16
25
29
7j (RS) 32
2j (RS) 43
26ꢂ1(3)
36ꢂ3(3)
1
8ꢂ6(2)
17ꢂ5(3)
3ꢂ1(4)
10ꢂ1(15)
6ꢂ2(6)
7j (S) 73
7k (RS) 24
7l (RS) 60
7l (S) 38
2j (S) 53
2k (RS) 16
2l (RS) 34
2l (S) 41
19ꢂ4(3)
9ꢂ1(2)
4ꢂ0(2)
1ꢂ1(2)
15ꢂ2(2)
6ꢂ0(2)
13ꢂ2(2)
10ꢂ3(2)
15ꢂ4(3)
16ꢂ2(3)
39ꢂ10(3)
16ꢂ4(2)
7ꢂ1(2)
7a (S) 89
7b (S) 62
7c (S) 66
7d (S) 74
7e (S) 60
7f (S) 43
7g (S) 75
7h (S) 56
7i (S) 58
2a (S) 48
2b (S) 39
2c (S) 68
2d (S) 70
2e (S) 39
2f (S) 28
2g (S) 32.5
2h (S) 35
2i (S) 43
4ꢂ0(2)
10ꢂ4(3)
6ꢂ3(2)
6ꢂ1(6)
4ꢂ1(6)
7m (RS) 15
7n (RS) 44
7o (RS) 8
7p (S) 68
7q (S) 82
2m (RS) 47
2n (RS) 25
2o (RS) 46
2p (S) 24
5ꢂ1(3)
55ꢂ11(3)
142ꢂ66(9)
8ꢂ1(2)
115ꢂ36(2)
141ꢂ30(2)
16ꢂ2(3)
47ꢂ2(2)
26ꢂ0(2)
15ꢂ3(2)
46ꢂ9(2)
40ꢂ4(2)
7ꢂ4(2)
18ꢂ5(2)
47ꢂ5(2)
2q (S) 36
n, number of experiments.
Scheme 2. Synthesis of compounds 3a–j. Reagents and conditions: (i) Ph(CH2)2NH2, MeOH, reflux (15, 89%); (ii) MsOH, rt, 72 h (16, 78%); (iii)
(a) NaH, RBr, Bu4NI, rt, 24 h (17a–f, % Table 1); (iv) (a) Bi(R)3, pyridine, Cu(OAc)2, DMF, CH2Cl2, rt, 24 h; (b) HBr, AcOH, (17g–j,% Table 1).
(v) (a) 8, NaBH(OAc)3, CH2Cl2, 0 ꢀC, rt, 12 h; (b) HCl, ether, or fumaric acid, EtOH (3a–j).
potent FTase inhibitors with IC50 between 10 and 100
nM. A 5- to 10-fold lower potency was observed in the
cell based assay (compounds 3a–j, Table 2). By com-
parison with the corresponding benzo[c]azepinone ser-
ies, a significant loss of potency is observed, taking into
account that we are in presence of the racemate form,
either for the benzyl substitution (2b and 3a; 2c and 3b;
2d and 3d; 2g and 3f) or for the aryl substitution (2j and
3g; 2l and 3h; 2p and 3i). An exception to this general
observation could be made for the bis-m-chloroaryl
substitution present in 2q and 3j. These two compounds
are equipotent, and slightly more active in the cellular
assays. No simple explanation to this phenomena could
be provided, since cellular potency is dependent on many
different parameters such as membrane transport, protein
binding, chemical stability and so on. However, a com-
parative conformational analysis based on the FTase
X-ray crystal structure described recently with the cyclic
compound U49,15 showed that the total complex energy
is 5 kcal molꢁ1 lower for 2b than for 3a. These data

770
T. Le Diguarher et al. / Bioorg. Med. Chem. Lett. 14 (2004) 767–771
Table 2. Structures and yields of compounds 16 and 3 and biological data of compounds 3
R
16
Yield (%)
3
Yield (%)
FTase
IC50 nMꢂSEM (n)
Ras cell
R
16
Yield (%)
3
Yield (%)
FTase
IC50 nMꢂSEM (n)
Ras cell
16a 42
16b 38
16c 56
16d 65
16e 95
3a 67
3b 60
3c 29
3d 45
3e 34
54ꢂ9(2)
96ꢂ14(5)
92ꢂ31(2)
67ꢂ9(2)
16f 74
16g 66
16h 41
16i 34
16j 30
3f 26
3g 45
3h 33
3i 29
3j 49
26ꢂ6(2)
379ꢂ60(2)
95ꢂ32(3)
39ꢂ7(3)
23ꢂ12(2)
78ꢂ33(2)
23ꢂ1(2)
22ꢂ4(2)
54ꢂ9(2)
19ꢂ5(2)
62ꢂ16(3)
32ꢂ7(2)
183ꢂ61(2)
60ꢂ6(3)
86ꢂ41(2)
19ꢂ5(2)
n, number of experiments.
Table 3. Antitumoral evaluation and GGTase activity
#
%Control
100 mg/kg po
% Control
200 mg/kg po
GGTase
IC50 mM (n)
BMS-214662
48
22
31
2b (S)
2l (S)
3h (RS)
76
60
70
57
35
43
7.1ꢂ1(2)
25ꢂ2(2)
7.7ꢂ3(2)
explained by the relatively poor metabolic stability of
the compounds 2l and 3h (<10% in the presence of rat
and human hepatic microsomes) despite a complete
absorption (100% in the Caco2 cell permeability assay).
Nevertheless, these compounds, selective for FTase ver-
sus the geranyl–geranyl transferase enzyme (GGTase),
were selected for further pharmacological evaluation
and the synthesis of the active enantiomer of 3h is in
progress (Table 3).
Figure 2. Conformational analysis of compounds 2b and 3a. Com-
pounds 2b (white) and 3a (magenta) were docked manually in the
binding cavity and their conformations were fully optimised with Tri-
pos force field to obtain the best zinc chelation and protein interac-
tions.16
5. Conclusion
could be explained by a p stacking between the [c]phenyl
ring of 2b and Tyra166 and is consistent with the better
FTase inhibitory activities for all this series (Fig. 2).16
The studies described herein the SAR of N,N0-dis-
ubstituted 3-amino benzo[c] and [d]azepinones, 2 and 3,
as new series of FTase inhibitors. The [c] series, almost
comparable to the caprolactam derivative 1, afforded
very potent compounds with low nM IC50, whereas the
[d] series showed a significant decrease in potency by a
factor 5–10. The activities of the selected compounds
were validated in vitro in a cellular assay, and two
compounds, 2l and 3h, shown significant activities in
vivo in a tumour model.
4. Pharmalogical evaluation
As a preliminary evaluation of antitumour activity,
selected compounds, with low nM IC50 in the enzymatic
and cellular assays, were tested in an in vivo model. H-
Ras transfected rat fibroblasts were grafted sc into nude
mice, and compounds were administered at 100 and 200
mg/kg p.o. for 5 days, starting from day 6 after tumour
implantation.4 Among these compounds only 2l and 3h
showed a significant dose-dependent activity with a
marked reduction of the tumour volume (65 and 57%,
respectively, at 200 mg/kg) an effect less important than
for the reference BMS-214662. These results could be
Acknowledgements
The authors wish to thank Sophie Sciberras, Christel
Daumas, Stephanie Dupas and Annie Genton for their
skilful technical assistance, and Solange Huet for the
preparation of the manuscript. We would like also to

T. Le Diguarher et al. / Bioorg. Med. Chem. Lett. 14 (2004) 767–771
771
acknowledge the analytical department at the IdRS for
performing all the spectral analyses.
11. All new compounds 2 and 3 gave spectroscopic data (IR,
NMR), elemental analyses (CHN) and/or MS in agree-
ment with the assigned structures.
12. Ben-Ishai, D.; Sataty, I.; Peled, N.; Goldshare, R. Tetra-
hedron 1983, 43, 439.
˜
13. Martin, A.; Gomez-Munoz, A.; Waggoner, D. W.; Stone,
References and notes
J. C.; Brindley, D. J. Biol. Chem. 1993, 268, 23924.
14. Lamothe, M.; Perez, M. Idrugs 2000, 3, 1336, and refer-
ences cited herein..
1. Huang, C. C.; Casey, P. J.; Fierke, C. A. J. Biol. Chem.
1997, 272, 20.
2. Lerner, E. C.; Hamilton, A. D.; Sebti, S. M. Anti-Cancer
Drug Des. 1997, 12, 229.
3. (a) Prendergast, G. C.; Rane, N. Expert Opin. Invest.
Drugs 2001, 10, 2105. (b) Cox, A. D.; Der, C. J. Curr.
Opin. Pharm. 2002, 2, 388. (c) Haluska, P.; Dy, G. K.;
Adjei, A. A. Eur. J. Cancer 2002, 38, 1685, and references
cited herein.
4. Le Diguarher, T.; Ortuno, J. C.; Dorey, G.; Shanks, D.;
Guilbaud, N.; Pierre, A.; Fauchere, J. L.; Hickman, J. A.;
Tucker, G. C.; Casara, P. J. Bioorg. Med. Chem. 2003, 11,
3193.
5. (a) Long, S. B.; Hancock, P. J.; Kral, A. M.; Hellinga,
H. W.; Beese, L. S. P.NA.S. 2001, 98, 12948. (b) Strick-
land, C. L.; Windsor, W. T.; Syto, R.; Wang, L.; Bond,
R.; Wu, Z.; Schwartz, J.; Le, H. V.; Beese, L. S.; Weber,
P. C. Biochemistry 1998, 37, 16601.
6. Park, H. W.; Boduluri, S. R.; Moomaw, J. F.; Casey, P. J.;
Beese, L. S. Science 1997, 275, 1800.
7. Berkowitz, W. F.; John, T. V. J. Org. Chem. 1984, 49,
5269.
15. Bell, I. A.; Gallicchio, S. N.; Abrams, M.; Beese, L. S.;
Beshore, D. C.; Bhimnathwala, H.; Bogusky, M. J.;
Buser, C. A.; Culberson, J. C.; Davide, J.; Ellis-Hutch-
ings, M.; Fernandes, C.; Gibbs, J. B.; Graham, S. L.;
Hamilton, K. A.; Hartman, G. D.; Heimbrook, D. C.;
Homnick, C. F.; Huber, H. E.; Huff, J. R.; Kassahun, K.;
Koblan, K. S.; Kohl, N. E.; Lobell, R. B.; Lynch, J. J.;
Robinson, R.; Rodrigues, A. D.; Taylor, J. S.; Walsh,
E. S.; Williams, T. M.; Zartman, C. B. J. Med. Chem.
2002, 45, 2388.
16. All structures were modelled in SYBYL1 6.9. using stan-
dard bond lengths and angles. The whole protein, includ-
ing the zinc and the farnesyl di-phosphate (yellow), was
kept fixed in aggregate, except the side chains of some of
the hydrophobic amino-acid residues of the beta chain at
the active site in close contact with the ligands which were
allowed to change their conformations depending on the
ligands docked, that is TRP102, TRP106, TYR154,
TRP303, TYR361 and TYR365. All the hydrogen atoms
were added on the crystal structure pdb1LD8 of the
enzyme and their geometries were optimised using the
Tripos force field, using Kollman charges on all atom of
8. Flynn, G. A.; Burkholder, T. P.; Huber, E. W.; Bey, P.
Bioorg. Med. Chem. Lett. 1991, 1, 309.
9. Chan, D. M. Tetrahedron Lett. 1996, 37, 9013.
10. Kammermeir, B. O. T.; Lerch, U.; Sommer, C. Synthesis
1992, 1157.
the residues of the protein and Gasteiger–Huckel charges
¨
on the other non-protein molecules. All the water, U49
and acetate molecules were removed before docking.