5-Imidazolyl-quinolinones, -quinazolinones and -benzo-azepinones as farnesyltransferase inhibitors

Article abstract of DOI:10.1016/S0960-894X(03)00180-X

The evaluation of structure-activity relationships associated with the modification of the R115777 quinolinone ring moiety displaying potent in vitro inhibiting activity is described.

Full text of DOI:10.1016/S0960-894X(03)00180-X

Bioorganic & Medicinal Chemistry Letters 13 (2003) 1543–1547
5-Imidazolyl-quinolinones, -quinazolinones and -benzo-azepinones
as Farnesyltransferase Inhibitors
Patrick Angibaud,a,* Xavier Bourdrez,a Ann Devine,b David W. End,b
Eddy Freyne,c Yannick Ligny,a Philippe Muller,a Geert Mannens,d Isabelle Pilatte,a
Virginie Poncelet,a Stacy Skrzat,b Gerda Smets,c Jacky Van Dun,c
Pieter Van Remoortere,e Marc Veneta,y and Walter Woutersc
aMedicinal Chemistry Department, Johnson & Johnson Pharmaceutical Research & Development,
Campus de Maigremont BP615, 27106 Val de Reuil, France
bOncology Drug Discovery, Johnson & Johnson Pharmaceutical Research & Development, L.L.C. Welsh and McKean Roads,
Spring-House, PA 19477-0776, USA
cOncology Discovery Research, Johnson & Johnson Pharmaceutical Research & Development, Turnhoutseweg 30 B-2340, Belgium
dPreclinical Pharmacokinetics, Johnson & Johnson Pharmaceutical Research & Development, Turnhoutseweg 30 B-2340, Belgium
eDrug Evaluation, Johnson & Johnson Pharmaceutical Research & Development, Turnhoutseweg 30 B-2340, Belgium
Received 17 October 2002; revised 16 January 2003; accepted 17 February 2003
Abstract—The evaluation of structure–activity relationships associated with the modification of the R115777 quinolinone ring
moiety displaying potent in vitro inhibiting activity is described.
# 2003 Elsevier Science Ltd. All rights reserved.
Oncogenic mutations of the ras gene are frequently
found in approximately 30% of human cancers includ-
ing a substantial proportion of pancreatic and colon
adenocarcinomas.1,2 By preventing attachment of the
Ras protein to the cell membrane, inhibitors of the
enzyme farnesyl protein transferase (FPT) prevent a key
step in the post-translational processing of Ras and the
subsequent propagation of the cell growth promoting
signals.3À10 Hence it was hypothesized that farnesyl
protein transferase inhibitors (FTI’s) could be very use-
ful as anticancer agents for tumors in which Ras con-
tributes to transformation. With the advent of potent
FTI’s it has become clear that the antitumor activity of
the class is quite complex involving other farnesylated
proteins such as Rho B, centromere associated proteins
or modulation of transcriptional events.11À14
extensive structure–activity relationship(SAR) studies
for this class of compounds. One identified in vivo
metabolite17 of R115777 is the N-demethylated quinoli-
none 2 (Fig. 1) which displayed a 10-fold lower in vitro
potency on cells than 1.
In attempts to block or attenuate this metabolization,
1,8-annelated quinolinones were developed.18 Further-
more the attempts for solubility, polarity and pharmaco-
kinetic profile improvements of R115777 led us to
several other modifications of the quinolinone backbone.
We report here features that are consistent with main-
taining high enzyme activity and cellular potency and
show their impact on in vivo results in murine models.
The pyrrolo[3,2,1-ij]quinolin-4-one 14, 18 and the
2,3-dihydro-1H,5H-benzo[ij]quinolizin-5-one 15, 19
were prepared as described in Scheme 1. 2,3-Dihydro-
1H-indole 3 and tetrahydroquinoline 4 were acylated
and cyclized into tetrahydro-quinolinone and-quinolizi-
none 8 and 9. Then Friedel–Craft acylation of 8 and 9
gave preferentially isomers 10 and 11. Oxidation of 10
and 11 was followed by addition of the lithio derivative
of 1-methylimidazole onto the ketone to provide 14 and
The discovery and progression to phase III clinical
studies of R115777 (1)15,16 have spawned in our group
*Corresponding author. Tel.: +33-2-3261-7457; fax: +33-2-3261-
7298; e-mail: pangibau@prdfr.jnj.com
yCurrent address: 10, Square de Bourgogne, F 76240 Le Mesnil-
Esnard, France.
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00180-X

1544
P. Angibaud et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1543–1547
Figure 1. Structure and FPT inhibitory potency of R115777 1 and its
desmethyl metabolite 2 (racemic form).
Scheme 2. Reagents and conditions: (a) AlCl3, CH2Cl2; (b) CH3I,
K2CO3, CH3CN; (c) ethyleneglycol, APTS, Toluene, 50%, 3 steps; (d)
3-chlorobenzyl cyanide, NaOH, MeOH, 27%; (e) H2, Pt/C, HCl conc.,
THF, 84%; (f) Ac2O, toluene; (g) tBuOK, DME, 78%; (h) (i) 1-
methylimidazole, ClSiEt3, n-BuLi, THF, À70 ꢀC; (ii) n-BuLi, À70 ꢀC,
88%; (i) BBr3, CH2Cl2; (j) Br–(CH2)n–Br, K2CO3, DMF, 31%; (k)
SOCl2, RT; (l) NH3/iPrOH, THF, 10 ꢀC, 50%.
Scheme 1. Reagents and conditions: (a) Et3N, CH2Cl2, rt, 73–98%;
(b) PPA, 140 ꢀC; (c) 4-chlorobenzoic acid, PPA, 140 ꢀC, 20–45%; (d)
Br2, PhBr, 160 ꢀC, 22–43%; (e) (i) 1-methylimidazole, ClSiEt3, n-BuLi,
THF, À70 ꢀC; (ii) n-BuLi, À70 ꢀC, 30–80%; (f) SOCl2, 40 ꢀC; (g)
NH4OH, THF, 5 ꢀC, 23–38%.
3-chlorobenzyl cyanide in presence of a base provided
the benzisoxazole 23 which was converted into the
corresponding ortho-aminobenzophenone 24 by hydro-
genation under acidic conditions. Acylation and base-
catalyzed cyclization into quinolinone 26 were followed
by condensation of the lithio derivative of 1-methylimi-
dazole onto the ketone to provide 27. Demethylation of
27 by BBr3 gave the quinolinone 28 which was con-
verted by intramolecular alkylation into quinolin-4-one
29 and benzoxazin-5-one 30.
15. Substitution of the hydroxy groupby a chlorine and
subsequent condensation of ammonia gave finally com-
pounds 18 and 19.
The same scheme could not be applied to the synthesis
of oxazolo[5,4,3-ij]quinolin-4-one 29 or pyrido[1,2,3-de]-
1,4-benzoxazin-5-one 32 as Friedel–Craft acylation of
intermediates of type 8,9 did not give the correct isomer.
Therefore we stepped back to the formation of the 8-
methoxyquinolinone intermediate 26 taking advantage
of the experience on quinolinone chemistry gained with
the R115777 project (Scheme 2).19À21
Benzisoxazole chemistry was also used as starting point
to synthesize the quinazolinone 41, a strict analogue of
R115777. 3322 was converted into ortho-amino benzo-
phenone 34 by hydrogenation and the amino moiety
acylated by trichloroacetyl chloride to afford 35. Cycli-
zation of 35 into a quinazolinone was achieved by
heating it in the presence of ammonium acetate in a
polar solvent. Subsequent cleavage of the ketal ring fol-
lowed by N-alkylation provided the ketone 38. Com-
pound 41 was then obtained in three steps following the
chemistry presented in Scheme 3.
Acylation of the 1-methoxy-2-nitrobenzene by the
4-chlorobenzoyl chloride gave the (4-chlorophenyl)(3-
hydroxy-4-nitrophenyl)methanone 20 which was then
alkylated into 21 and ketalized. Reaction of 22 with

P. Angibaud et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1543–1547
1545
Scheme 4. Reagents and conditions: (a) HCl 3N; (b) (i) 1-methylimi-
dazole, ClSiEt3, n-BuLi, THF, À70 ꢀC; (ii) n-BuLi, À70 ꢀC, 58%;
(c) TiCl3/H2O THF, 5 ꢀC, 69%; (d) NaBH4, MeOH; (e) BrC(O)CH2Br,
K2CO3, CH2Cl2; (f) tBuOK, iPrOH, 65%; (g) HOC(O)CH2SH, HCl
6 N, 98%; (h) toluene, MgSO4, 30%; (i) CH3I, NaH, DMF, 13–19%;
(j) SOCl2, 5 ꢀC; (k) NH3/iPrOH, THF, 27–45%.
Scheme 3. Reagents and conditions: (a) H2, Pd/C, MeOH; (b)
ClC(O)CCl3, Et3N, CH2Cl2; (c) AcONH4, HMPT, 100 ꢀC, 76%, three
steps; (d) HCl 3 N, MeOH, 95%; (e) CH3I, NaH, DMF, 65%; (f) (i) 1-
methylimidazole, ClSiEt3, n-BuLi, THF, À70 ꢀC; (ii) n-BuLi, À70 ꢀC,
35%; (g) SOCl2, 40 ꢀC; (h) NH3/iPrOH, THF, 10 ꢀC, 50%.
As some benzodiazepines23 have shown in the past some
FPT inhibitory effect the ring enlargement of the qui-
nolinone into benzoxa-, benzothia-, or benzo-diazepi-
nones was naturally envisioned. To prepare the
1,5-dihydro-1-methyl-4,1-benzoxazepin-2(3H)-one 50a
or benzothiazepin-2(3H)-one 50b we first condensed the
1-methyl-5-imidazolyl moiety onto the 5-(4-chloro-
benzoyl)-benzisoxazole 42, obtained from 33 by cleav-
ing the ketal moiety in acidic media, to provide 43. The
derivative 43 was converted into the corresponding
ortho-aminobenzophenone 44 as already described and
the keto function was reduced into the key alcohol 45.
Bromoacetyl bromide or thioacetic acid were then con-
densed on the amino and alcohol functions of 45 to
provide the heterocycles 47a and 47b. N-methylation of
the amide nitrogen was followed by the chlorination
and amination steps already depicted to afford 50a and
50b (Scheme 4).24
Scheme 5. Reagents and conditions: (a) ethylglycinate, pyridine, mol.
sieves, 88%; (b) NaH, CH3I, DMF, 43%; (c) (i) 1-methylimidazole,
ClSiEt3, n-BuLi, THF, À70 ꢀC; (ii) n-BuLi, À70 ꢀC, 38%; (d) SOCl2,
5 ꢀC; (e) NH3/iPrOH, THF, 5 ꢀC, 48%.
Benzodiazepinone 56 was prepared in a similar way
from [2 - amino - 5- (4 -chlorobenzoyl)phenyl](3 -chloro-
phenyl)-methanone 5125 except that the imidazole was
introduced after heterocycle formation as depicted in
Scheme 5.
A survey of 1,8-annelated analogues of R115777 (Table
1) revealed that X=C or O had only a minor impact on
enzyme activity, whereas X=O significantly decreased
cellular activity compared to R115777. As 18 showed the
best enzyme and cellular potency, it was separated into
its enantiomers 18a and 18b.27 Stereochemistry clearly
has an influence on in vitro potency and 18b was tested in
murine models (Table 2). Relative to R115777, metabo-
lism was slightly reduced as foreseen. But 18b showed a
decrease in activity at a dose of 25 mg/kg probably due to
several factors including a lesser solubility and a reduc-
tion of the free circulating fraction in plasma.
The structure–activity relationships for these com-
pounds are presented in Tables 1–3.
We turned our attention towards more drastic modifi-
cations of the R115777 skeleton, still keeping intact the

1546
P. Angibaud et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1543–1547
Table 1. FPT inhibition results for compounds 14–32
and therefore they were not considered as valuable can-
didates for further studies.
Compd
X
n
R
FPT (enz)
IC50, nMa
FPT (cells)
IC50, nMb
GGPT
IC50, mMc
Modification of the R115777 backbone has led us to
derivatives showing similar but not superior in vitro and
in vivo FPT inhibiting potency. In vitro selectivity of
R115777 for FPTase versus GGPTase I has also been
conserved within the tested compounds. Further chemi-
cal efforts in these series are being focused on improving
activity, solubility and pharmacokinetic parameters.
1
2
—
—
CH2
CH2 1 NH2
CH2 OH
CH2 2 NH2
O
O
O
— NH2
— NH2
0.8
3.5
1
0.5
4
1.7
6
6
1.6
1.7
23
9
4
54
9
45
10
nt
nt
1.5
nt
14
18
15
19
29
30
32
18ad
18bd
1
OH
2
nt
1
2
OH
OH
51% inh@10 mM
22% inh@10 mM
38% inh@10 mM
nt
62
>500
nt
2 NH2
CH2 1 NH2 31% inh@0.1 mM
CH2 1 NH2 0.5
References and Notes
3.3
0.6
nt, not tested.
1. Rodenhuis, S. Semin. Cancer Biol. 1992, 3, 241.
2. Bos, J. L. Cancer Res. 1989, 49, 4682.
3. Barbacid, M. Ann. Rev. Biochem. 1987, 56, 779.
4. Kato, K.; Cox, A. D.; Hisaka, M. M.; Graham, S. M.;
Buss, J. E.; Der, C. J. Proc. Natl. Acad. Sci. U.S.A. 1992, 89,
6403.
5. Der, C. J.; Cox, A. D. Cancer Cells 1991, 3, 331.
6. Seabra, M. C.; Reiss, Y.; Casey, P. L.; Brown, M. S.;
Goldstein, J. L. Cell 1991, 65, 429.
aThe concentration required to reduce the FPTase-catalyzed incor-
poration of [3H]-farnesylpyrophosphate into a biotinylated laminB
peptide by 50%.
bSee ref 26.
cThe concentration required to reduce the Geranylgeranylprotein
transferase-I (GGPTase-I) catalyzed incorporation of [3H]-geranylger-
anylpyrophosphate into a biotinYRASNRSCAIL substrate by 50%.
dee>98%.27
7. Rowinsky, E. K.; Windle, J. J.; Von Hoff, D. D. J. Clin.
Oncol. 1999, 17, 3631.
8. Rowinsky, E. K.; Patnaik, A. Emerg. Drugs 2000, 5, 161.
9. End, D. W. Invest. New drugs 1999, 17, 241.
10. Johnston, S. Lancet Oncol. 2001, 2, 18.
11. Lebowitz, P. F.; Casey, P. J.; Prendergast, G. C.; Thissen,
J. A. J. Biol. Chem. 1997, 272, 15591.
12. Ashar, H. R.; James, L.; Gray, K.; Carr, D.; Black, S.;
Armstrong, L.; Bishop, W. R.; Kirschmeier, P. J. Biol. Chem.
2000, 275, 30451.
13. Adnane, J.; Bizouarn, F. A.; Chen, Z.; Ohkanda, J.;
Hamilton, A. D.; Munoz-Antonia, T.; Sebti, S. M. Oncogene
2000, 19, 5525.
14. Feldkamp, M. M.; Lau, N.; Rak, J.; Kerbel, R. S.; Guha,
A. Int. J. Cancer 1999, 81, 118.
Table 2. Comparison of in vivo results for compounds R115777 and
18b
Compd
Metabolizationa
(%)
In vivob
% inhib.
Protein bindingc
% free
R115777
18b
66
74
37
28
0.66
0.25
aPercentage of parent drug remaining after 120 min incubation with
human liver microsomes.28
bIn vivo screening at 25 mg/kg in mice injected with T24 NIH 3T3
cells, percentage of tumor weight inhibition.29
cPercentage of free circulating fraction of drug in plasma.
15. End, D. W.; Smets, G.; Todd, A. V.; Applegate, T. L.;
Fuery, C. J.; Angibaud, P.; Venet, M.; Sanz, G.; Poignet, H.;
Skrzat, S.; Devine, A.; Wouters, W.; Bowden, C. Cancer Res.
2001, 61, 131.
Table 3. FPT inhibition results for compounds 41, 50a, 50b and 56
Compd
FPT (enz)
IC50, nMa
FPT (cells)
IC50, nMb
GGPT
IC50, mMc
16. Norman, P. Curr. Opin. Investing. Drugs 2002, 3, 313.
17. End, D. W. Unpublished results.
18. Venet, M. G.; Angibaud, P. R.; Ligny, Y. A. E.; Poncelet,
V. S.; Sanz, G. C. W.O. Patent 98/40383, 1998.
19. Venet, M. G.; Angibaud, P. R.; Muller, P.; Sanz, G. C.
W.O. patent 97/21701, 1997.
41
6.5
48% inh@0.1 mM
29
64
nt
nt
nt
nt
nt
nt
50a
50b
56
22% inh@0.1 mM
32% inh@10 mM
a,b,cSee corresponding footnotes in Table 1. nt, not tested.
20. Filliers, W. F. M.; Broeckx, R. L. M.; Leurs, S. M. H.
W.O. patent 02/72574, 2002.
21. Venet, M. Abstract of papers, 222nd National Meeting of
the American Chemical Society, Chicago, IL, 26–30 August
2001; American Chemical Society: Washington, DC, 2001;
69BUZP.
basic features previously identified as crucial, namely,
the C5-linked imidazole, the two chloro-phenyl moieties
and the N-methyl–amide bond.30
22. 3-(3-Chlorophenyl)-5-[2-(4-chlorophenyl)-1,3-dioxolan-2-yl]-
2,1-benzisoxazole 33 was prepared from 4-chloro-40-nitro-
benzophenone exactly as described in Scheme 2 steps c and d
(86% yield). Angibaud, P. R.; Venet, M. G.; Freyne, E. J. E.
W.O. Patent 98/49157, 1998.
23. James, G. L.; Goldstein, J. L.; Brown, M. S.; Rawson,
T. E.; Somers, T. C.; McDowell, R. S.; Crowley, C. W.; Lucas,
B. K.; Levinson, A. D.; Marsters, J. C., Jr. Science 1993, 260,
1937.
Quinazoline analogue 41 displayed a reduced activity in
vitro (Table 3) but owing to its lower calculated logP
value (3.4 compared to 4.2 in 1), we considered that the
introduction of another nitrogen atom in R115777
backbone could result in improved solubility. However,
no clear in vivo activity could be detected at a dose of
25 mg/kg.
24. Angibaud, P. R.; Venet, M. G.; Poncelet, V. S. W.O
Patent 02/42296, 2002.
25. Angibaud, P. R.; Venet, M. G.; Bourdrez, X. M. W.O.
Patent 00/39082, 2000.
26. Cellular assays for H-ras transforming function: T24
Quinoline ring enlargement into benzo-azepinone close
analogues of R115777 was our next attempt of modifi-
cation. However the three compounds synthesized
showed inferior in vitro enzymatic profile than R115777

P. Angibaud et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1543–1547
1547
F1H-ras transfected NIH 3T3 cells were inoculated into six-
well cluster dishes at an initial density of 200,000 cells per well
in 3 mL of complete growth medium. Test compounds were
added at concentrations ranging from 0.5 to 500 nM in dime-
thylsulfoxide. Cells were allowed to proliferate to high satura-
tion densities beyond confluence for 4 days. Cell numbers were
quantified by detaching cells in 1 mL of trypsin/EDTA and
counting cell suspensions on a Coulter particle counter. IC50
(cells): The concentration required to decrease the cell number
by 50% as compared to a control (untreated) sample.
27. (12.4 g) was separated and purified by chiral column
chromatography over Chiracel OD1 (eluent: 100% CH3OH).
Two pure fraction groups were collected and the solvent was
evaporated. The first fraction was crystallized from 2-propa-
nol/DIPE. The precipitate was filtered off and dried, yielding
4.4 g of (À)-8-[amino(4-chlorophenyl)(1-methyl-1H-imidazol-
5-yl)methyl]-6-(3-chlorophenyl)-1,2-dihydro-4H-pyrrolo[3,2,1-
ij]quinolin-4-one 18a. The solvent of the second fraction group
was evaporated. The second fraction was crystallized from
2-propanol/ DIPE. The precipitate was filtered off and dried,
yielding: 4.1 g of (+)-8-[amino(4-chlorophenyl)(1-methyl-1H-
imidazol-5-yl)methyl]-6-(3-chlorophenyl)-1,2-dihydro-4H-pyr-
rolo[3,2,1-ij]quinolin-4-one 18b.
and samples were stored at <18 ꢀC until analysis. Samples
were analysed for unchanged compounds by HPLC-UV. The
compounds were also incubated with boiled 12,000 g fractions
to check the compound stability and recovery.
29. Tumor studies in nude mice: Tumor cell lines maintained
as monolayer cultures were detached by tripsynisation. Tumor
cell suspensions were pooled after trypsin was inactivated by
the addition of serum-containing medium. Cells were collected
by centrifugation and washed once in HBSS. Cell suspensions
were adjusted to a final concentration of 1 ꢁ 106 cells/0.1mL of
HBSS. Mice were inoculated with a single s.c. injection of
0.10 mL of tumor cell suspension in the inguinal region of the
thigh. Mice were housed five per cage, with 15 mice randomly
assigned to treatment groups. Three days after tumor inocula-
tion, treatment with R115777 1 or 18b was initiated. Compounds
were administered once daily by oral gavage in a 20% b-cyclo-
dextrin vehicle as a volume of 0.10 mL of solution/10 g of body
weight. Control groups received the same dosage/volume of the
20% b-cyclodextrin vehicle. Body weight and tumor size as
determined by caliper measurements were monitored weekly. At
the end of the study mice were sacrificed by CO2 asphyxiation.
Tumors were excised, weighted, and fixed immediately in 4%
paraformaldehyde. ANOVA, mean values for treatment groups,
and SE for in vivo parameters were calculated on a VAX com-
puter using IMSL subroutines compiled by the Science Informa-
tion Department of the R.W. Johnson Pharmaceutical Research
Institute. A value of p<0.05 was considered significant.
28. Metabolism: Compounds were incubated at 30 mM with
human subcellular liver fractions (12,000 g supernatants) dur-
ing 120 min at a final microsomal protein concentration of
1 mg/mL in a presence of an NADPH-generating system.
Reactions were stopped by freezing the incubates in dry ice,
30. Meyer, C. Unpublished results.