Novel and selective imidazole-containing biphenyl inhibitors of protein farnesyltransferase

Article abstract of DOI:10.1016/S0960-894X(03)00096-9

A series of imidazole-containing biphenyls was prepared and evaluated in vitro for inhibition of FTase and cellular Ras processing. Several of these analogues, such as 21, are potent inhibitors of FTase (<1 nM), FTase/GGTase selective (>300-fold) and cellularly active (≤80 nM). An X-ray crystal structure of inhibitor 21 bound to rat farnesyltransferase is also presented.

Full text of DOI:10.1016/S0960-894X(03)00096-9

Bioorganic & Medicinal Chemistry Letters 13 (2003) 1367–1371
Novel and Selective Imidazole-Containing Biphenyl Inhibitors of
Protein Farnesyltransferase
Michael L. Curtin,* Alan S. Florjancic, Jerome Cohen, Wen-Zhen Gu, David J. Frost,
Steven W. Muchmore and Hing L. Sham
Departments of Cancer Research and Advanced Technology, Global Pharmaceutical Research and Development, Abbott Laboratories,
100 Abbott ParkRoad, Abbott Park, IL 60064, USA
Received 1 November 2002; accepted 31 December 2002
Abstract—A series of imidazole-containing biphenyls was prepared and evaluated in vitro for inhibition of FTase and cellular Ras
processing. Several of these analogues, such as 21, are potent inhibitors of FTase (<1 nM), FTase/GGTase selective (>300-fold)
and cellularly active (ꢀ80 nM). An X-ray crystal structure of inhibitor 21 bound to rat farnesyltransferase is also presented.
# 2003 Elsevier Science Ltd. All rights reserved.
Ras proteins play an integral role in the transmission of
signals from extracellular stimuli to the nucleus and the
ras gene has been identified as one of the most fre-
quently mutated genes in human cancers.1,2 Ras pro-
chains;10 this work culminated in the discovery of the
clinical candidate ABT-839 (3, FTase IC50 1.0 nM).
teins must undergo
a series of posttranslational
modifications to provide the active species, the first
occurring through the transfer of a farnesyl group from
farnesylpyrophosphate (FPP) to the cysteine near the
C-terminus (CAAX box) of the protein.3 This transfer is
catalyzed by the enzyme protein farnesyltransferase
(FTase). Thus, the inhibition of FTase represents a
potential target for the development of small molecule
anticancer agents.4
A number of structurally distinct families of farnesyl-
transferase inhibitors (FTIs) have been disclosed in the
literature.5 Diverse examples from two of these classes
include imidazole-containing R115777 (1, FTase IC50
0.57 nM)6 and tricyclic inhibitor SCH66336 (2, FTase
IC50 7.8 nM),7 both of which are reported to be in Phase
III clinical trials for cancer.8 Another structural class
encompasses the RAS C-terminal tetrapeptide mimics in
which the central two amino acids have been replaced
with a biphenyl core which is flanked by cysteine and
methionine.9 Our laboratories have published exten-
sively on a modification of this motif in which the
cysteine is replaced with aryl, alkyl or heterocyclic side
Because many of these amino acid-containing biphenyl
compounds possessed modest bioavailability and a
short half-life, potent analogues lacking the methionine
moiety were desired. Despite the observation that the
methionine residue is crucial for efficient binding of
these compounds with FTase, a series of potent inhibi-
tors arising from the lead structure 4 (FTase IC50
*Corresponding author. Tel.: +1-847-938-0463; fax +1-847-935-
5165; e-mail: mike.curtin@abbott.com
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00096-9

1368
M. L. Curtin et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1367–1371
65 nM) has recently been discovered.11 While that work
was on-going, a separate series of compounds was
examined in which the biphenyl/imidazole hydroxy-
methylene-tether was replaced by an aminomethylene
linker with alkyl and alkylaryl substituents on the tether
nitrogen (e.g., 9, Scheme 1). Previous work on amino
acid-containing biphenyl inhibitors with this imidazole
tether12 suggested that this modification would be per-
mitted although its effect on the enzyme selectivity or
cellular activity of these compounds was not clear. The
SAR of this tether-modified series against FTase and
cellular RAS processing is described here.
give compound 10; deprotection of the imidazole pro-
vided inhibitor 11.
The ability of these compounds to inhibit FTase in vitro
was examined along with their selectivity against the
related prenyltransferase, geranylgeranyltransferase I
(GGTase), which is believed to be essential for normal
cellular processes (Table 1).19 While the parent imi-
dazolylmethyl compound 8 was only weakly active against
FTase (IC50 1 mM), substitution of the biphenyl/imi-
dazole-tether nitrogen improved potency significantly.
As it had been reported that addition of 4-substituted
benzyl groups to other non-peptidic, imidazole-containing
FTIs enhanced potency,20 initial work focused on benzyl
and cyanobenzyl substitution. As expected, N-benzyl deri-
vative 9 (FTase IC50 11 nM) was much more potent than 8,
but equipotent with the 4-cyanobenzyl analogue 12. Inter-
estingly, the 3-cyano compound 13 (FTase IC50 0.97 nM)
was 10-fold more potent than the 4-cyano inhibitor. Both
12 and 13 were also quite selective for inhibition of FTase
versus GGTase (100- to 1000-fold).
The preparation of the tether-modified compounds
(Scheme 1) began with methyl 3-chloro-4-cyanobenzo-
ate, prepared in two steps from 4-amino-2-chloro-
benzonitrile,13 which was coupled with 2-methylphenyl
boronic acid in good yield;14 saponification of the
biphenyl ester with LiOHat room temperature gave
carboxylic acid 5. This compound was subjected to
Curtius reaction conditions in t-BuOHand the resulting
N-(tert-butoxycarbonyl)aniline was isolated and depro-
tected to afford aniline 6. Using reductive amination
conditions,15 aniline 6 was coupled with aldehyde 7,
prepared by regioselective formylation of 1-methyl-2-
triethylsilylimidazole16 with N-formylmorpholine and
tert-butyllithium, to give imidazole 8 after in situ silyl
deprotection. This compound was then N-alkylated
with a variety of alkyl halides using KOtBu to afford
inhibitors such as 9.17 Substitution of 2-methylboronic
acid with 1-naphthylboronic acid in the first step of the
synthesis gave naphthylbiphenyl analogues such as 21 in
comparable yields.
The presence and substitution of the imidazole was
important for enzymatic activity. N-Desmethyl analo-
gue 11 (FTase IC50 84 nM) was a weaker inhibitor than
9, while replacement of the imidazole of 9 with a 3-pyr-
Synthesis of the imidazole N-desmethyl analogues were
prepared as shown in Scheme 2. Aniline 6 was coupled
with
1-(triphenylmethyl)imidazole-4-carboxaldehyde,
prepared in two steps from 4-(hydroxymethyl)-
imidazole,18 under reductive amination conditions to
Scheme 2. Reagents and conditions: (a) 1-(triphenylmethyl)imidazole-
4-carboxaldehyde (2 equiv), NaBH(OAc)3 (2.7 equiv), AcOH(5 equiv),
DCE, 74%; (b) benzyl chloride, KOtBu, THF, 0 ꢁC, 92%; (c) Et3SiH,
TFA, CH2Cl2, 87%.
Table 1. FTase and GGTase inhibition of o-tolyl analogues
Scheme 1. Reagents and conditions: (a) 2-methylphenylboronic acid,
CsF, cat. Pd(OAc)2, cat. 2-(dimethylamino)-20-dicyclohexylphos-
phinobiphenyl, dioxane, 80%; (b) LiOH, 1:1 THF/MeOH, 91%; (c)
DPPA, Et3N, t-BuOH; (d) TFA, CH2Cl2, 56% two steps; (e) 7
(2 equiv), NaBH(OAc)3 (2.7 equiv), AcOH(5 equiv), DCE, 68%; (f)
benzyl chloride, KOtBu, THF, 0 ꢁC, 70% (g) t-BuLi, 4-formyl-
morpholine, THF, ꢂ78 ꢁC, >95%.
Compd
R
FTase IC50 (nM)19 GGTase IC50 (nM)19
ND
8
9
12
13
H1000
PhCH2
(4-CNPh)CH2
(3-CNPh)CH2
11
10
ND
1000
1200
0.97

M. L. Curtin et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1367–1371
Table 2. FTase and GGTase inhibition of naphthyl analogues
1369
Table 3. Enzymatic and cellular data for selected inhibitors
Compd
FTase IC50 (nM)19
RAS EC50 (mM)21
1
2
0.57
7.8
0.002
0.16
12
13
20
21
22
23
10
0.97
1.5
0.39
0.45
0.37
29% inh. @ 1
0.07
64% inh. @ 1
0.05
45% inh. @ 1
0.08
Compd
R
FTase IC50 (nM)19 GGTase IC50 (nM)19
ND
14
15
16
17
18
19
20
21
22
H31
Methyl
n-Hexyl
PhCH2
PhCO
PhSO2
38
1.9
1.5
14
18
1.5
ND
180
ND
310
1900
1000
150
against FTase than the corresponding o-tolyl com-
pounds by 3- to 30-fold. The unsubstituted and tether
N-methylated analogues (14 and 15, respectively) were
the weakest in this series while nitrogen substitution with
n-hexyl or benzyl gave very potent inhibitors (FTase IC50
1.9 nM and 1.5 nM, respectively). The N-benzoyl (18)
and N-benzenesulfonyl (19) compounds had inter-
mediate enzymatic potency against FTase. Consistent
with the o-tolyl series SAR, the 4-cyanobenzyl analogue
20 was equipotent with the benzyl-substituted inhibitor
17 while the 3-cyano derivative 21 exhibited the best
potency of the series (FTase IC50 0.39 nM). The use of
3-chlorobenzyl (22) or (3-cyanopyrid-4-yl)methyl (23)
gave inhibitors equipotent with 21. The imidazole
N-desmethyl analogue of 17 was somewhat less potent
(FTase IC50 11 nM, compound not shown).
(4-CNPh)CH2
(3-CNPh)CH2
(3-ClPh)CH2
0.39
0.45
230
23
0.37
650
idyl gave a significant loss of potency (FTase IC50
450 nM, compound not shown).
In light of the observation that replacement of the
o-tolyl of 4 with 1-naphthyl gave a>20-fold enhance-
ment of FTase inhibitory activity,11 most of the tether
substitution work was completed on this biphenyl core
(e.g., 14 in Table 2). Regardless of tether substitution,
the naphthyl analogues were consistently more potent
Concerning GGTase inhibitory activity, all compounds
in the naphthyl series had significant FTase/GGTase
Figure 1. Inhibitor 21 bound to rat farnesyltransferase and a-hydroxyfarnesylphosphonic acid (HFP).

1370
M. L. Curtin et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1367–1371
selectivity (ꢃ90-fold) with inhibitor 23 being the most
Acknowledgements
selective (1600-fold).
The authors wish to thank Weibo Wang, Stephen
Fakhoury and Gerry Sullivan for the preparation of
starting materials used in this work. Robert Warner,
Peter Kovar and Jang Lee are acknowledged for the
pharmacokinetic studies.
The cellular potency of selected compounds was eval-
uated in a RAS processing assay and a portion of that
data, including the reference inhibitors 1 and 2, is shown
in Table 3.21 While there was a general correlation
between FTase enzymatic activity and inhibition of
RAS processing (e.g., 12 vs 13), there were a number of
examples of equipotent FTase inhibitors with sig-
nificantly different cellular activity (e.g., 21 or 23 vs 22).
The most potent inhibitor in this assay (21) compared
favorably to clinical candidate 2 in both enzymatic and
cellular potency.
References and Notes
1. Barbacid, M. Annu. Rev. Biochem. 1987, 56, 779.
2. Bos, J. L. Cancer Res. 1990, 50, 1352.
3. Newman, C. M.; Magee, A. I. Biochim. Biophys. Acta 1993,
1155, 79.
4. Prendergast, G. C.; Rane, N. Expert Opin. Investig. Drugs
2001, 10, 2105.
5. Dinsmore, C. J. Curr. Opin. Oncol. Endoc. Metab. Invest.
Drugs 2000, 2, 26.
However, pharmacokinetic studies in rat of several of
these inhibitors, including compound 21, revealed
short half-lives (ꢀ1 h) and low oral bioavailability
(<10%).
6. The given IC50 was determined with bovine brain
FTase using the SPA assay, as in ref 10b. Crul, M.; de
Klerk, G. J.; Swart, M.; van’t Veer, L. J.; de Jong, D.; Boer-
rigter, L.; Palmer, P. A.; Bol, C. J.; Tan, H.; de Gast, G. C.;
Beijnen, J. H.; Schellens, J. H. M. J. Clin. Oncol. 2002, 20,
2726.
7. The given IC50 was determined with bovine brain FTase
using the SPA assay, as in ref 10b. Adjei, A. A.; Erlichman,
C.; Davis, J. N.; Cutler, D. L.; Sloan, J. A.; Marks, S.;
Hanson, L. J.; Svingen, P. A.; Atherton, P.; Bishop, W. R.;
Kirschmeier, P.; Kaufmann, S. H. Cancer Res. 2000, 60, 1871.
8. (a) Karp, J. E.; Kaufmann, S. H.; Adjei, A. A.; Lancet,
J. E.; Wright, J. J.; End, D. W. Curr. Opin. Oncol. 2001, 13,
470. (b) Hahn, S. M.; Bernhard, E.; McKenna, W. G. Semin.
Oncol. 2001, 28, 86.
An X-ray crystal structure was acquired for inhibitor
21 bound to rat farnesyltransferase along with
a-hydroxyfarnesylphosphonic acid (HFP), an FPP ana-
logue (Fig. 1).22,23 It was suspected that the imidazole-
containing biphenyl inhibitors (e.g., 4 and 9) maintain
binding affinity to the FTase active site by replacing the
important methionine-enzyme interaction of inhibitor 3
with an interaction between the imidazole and the
putative active site Zn2+ ion. The data from the X-ray
structure supported this hypothesis and also revealed
several other key inhibitor/enzyme interations which are
consistent with the SAR of this series. In possibly the
most important interaction, the distal imidazole nitro-
gen of 21 coordinates with the active site zinc (2.36 A)
through displacement of a water molecule, an interac-
tion not observed with the methionine-containing
biphenyl inhibitors such as 3 but similar to the zinc-
imidazole interaction of FTase-bound 1.22b The
naphthyl moiety sits in a hydrophobic pocket formed
by Trp102b and Trp106b, with a naphthyl/phenyl
dihedral angle of 60.1ꢁ. The tether N-benzyl ring fits
on a shelf created by Tyr361b with the m-cyano
substituent resting in a niche formed by the displace-
ment of a water molecule and making a hydrogen
bond with the enzyme backbone (3.22 A). The crystal
structure also revealed several interactions between 21
and the bound HFP as well as a hydrogen bond
between the biphenyl cyano group and the side-chain
of Tyr166b (3.47 A).
9. Qian, Y.; Vogt, A.; Sebti, S. M.; Hamilton, A. D. J. Med.
Chem. 1996, 39, 217.
10. (a) Augeri, D. J.; O’Connor, S. J.; Janowick, D.; Szcze-
pankiewicz, B.; Sullivan, Gerry; Larsen, J.; Kalvin, D.; Cohen,
J.; Devine, E.; Zhang, H.; Cherian, S.; Saeed, B.; Ng, S.;
Rosenberg, S. J. Med. Chem. 1998, 41, 4288. (b) O’Connor,
S. J.; Barr, K. J.; Wang, L.; Sorensen, B. K.; Tasker, A. S.;
Sham, H.; Ng, S.; Cohen, J.; Devine, E.; Cherian, S.; Saeed,
B.; Zhang, H.; Lee, J. Y.; Warner, R.; Tahir, S.; Kovar, P.;
Ewing, P.; Alder, J.; Mitten, M.; Leal, J.; Marsh, K.; Bauch,
J.; Hoffman, D. J.; Sebti, S. M.; Rosenberg, S. H. J. Med.
Chem. 1999, 42, 3701. (c) Henry, K. J., Jr.; Wasicak, J.; Tas-
ker, A. S.; Cohen, J.; Ewing, P.; Mitten, M.; Larsen, J. J.;
Kalvin, D. M.; Swenson, R.; Ng, S.; Saeed, B.; Cherian, S.;
Sham, H.; Rosenberg, S. H. J. Med. Chem. 1999, 42, 4844. (d)
Shen, W.; Fakhoury, S.; Donner, G.; Henry, K.; Lee, J.;
Zhang, H.; Cohen, J.; Warner, R.; Saeed, B.; Cherian, S.;
Tahir, S.; Kovar, P.; Bauch, J.; Ng, S.; Marsh, K.; Sham, H.;
Rosenberg, S. Bioorg. Med. Chem. Lett. 1999, 9, 703. (e)
Augeri, D. J.; Janowick, D.; Kalvin, D.; Sullivan, G.; Larsen,
J.; Dickman, D.; Ding, H.; Cohen, J.; Lee, J.; Warner, R.;
Kovar, P.; Cherian, S.; Saeed, B.; Zhang, H.; Tahir, S.; Ng, S.;
Sham, H.; Rosenberg, S. H. Bioorg. Med. Chem. Lett. 1998, 9,
1069.
In summary, a series of potent and novel imidazole-
containing biphenyl inhibitors of FTase has been dis-
covered which has significant FTase/GGTase selectivity
and cellular Ras processing inhibitory activity. An
X-ray crystal structure of the most potent member of
the series with rat farnesyltransferase revealed several
key inhibitor/enzyme interactions including an imida-
zole/zinc coordination and two cyano/enzyme hydrogen
bonds. Analysis of several of these compounds in rat
indicated an unacceptable pharmacokinetic profile.
Further modification of the imidazole-containing
biphenyl FTase inhibitors will be reported in due
course.
11. Wang, W. Unpublished data.
12. Vasudevan, A.; Qian, Y.; Vogt, A.; Blaskovich, M. A.;
Ohkanda, J.; Sebti, S. M.; Hamilton, A. D. J. Med. Chem.
1999, 42, 1333.
13. For a similar conversion, aniline to methyl benzoate via
diazotization/iodination followed by palladium catalyzed carbo-
methoxylation, see: Diaz, P.; Michel, S.; Stella, L.; Charpen-
tier, B. Bioorg. Med. Chem. Lett. 1997, 7, 2289.
14. Wolfe, J. P.; Singer, R. A.; Yang, B. H.; Buchwald, S. L.
J. Am. Chem. Soc. 1999, 121, 9550.

M. L. Curtin et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1367–1371
1371
15. Abdel-Magid, A. F.; Carson, K. G.; Harris, B. D.; Mar-
yanoff, C. A.; Shah, R. D. J. Org. Chem. 1996, 61, 3849.
once and the reliability of the assay was ꢄ50–100%. See ref
10b for further details.
´
16. Molina, P.; Fresneda, P. M.; Garcıa-Zafra, S. Tet. Lett.
22. (a) Crystals of rat farnesyltransferase were exposed to
compound stabilization solutions for approximately four
hours before data collection Diffraction data was collected on
a rotating anode source, and a Mar-135 CCD camera system.
Data reduction using the HKL2000 program suite yielded a
data set which was 98% complete to 2.8 A, with merging
R-values of 4.8% overall. Electron density maps were calcu-
lated from models which had been rigid body refined against
the data. Models were constructed to fit the observed electron
density, and were refined using the XPLOR program package.
Final models have R work and R-free values of 24.9 and
32.1% respectively, with no water molecules. Models of this
complex have been submitted to the Protein Data Bank
(accession 1NL4). (b) Muchmore, S. Unpublished data.
23. For the X-ray crystal structure of 2 with FTase/FPP,
see: Strickland, C. L.; Weber, P. C.; Windsor, W. T.; Wu, Z.;
Le, H. V.; Albanese, M. M.; Alvarez, C. S.; Cesarz, D.; del
Rosario, J.; Deskus, J.; Mallams, A. K.; Njoroge, F. G.;
Piwinski, J. J.; Remiszewski, S.; Rossman, R. R.; Tavera,
A. G.; Vibulbhan, B.; Doll, R. J.; Girijavallabhan, V. M.;
Ganguly, A. K. J. Med. Chem. 1999, 42, 2125.
1996, 37, 9353.
17. All compounds were characterized by 1HNMR, mass
spectroscopy and elemental analysis.
18. Hunt, J. T.; Lee, V. G.; Leftheris, K.; Seizinger, B.; Car-
boni, J.; Mabus, J.; Ricca, C.; Yan, N.; Manne, V. J. Med.
Chem. 1996, 39, 353.
19. In vitro IC50 data were determined against FTase and
GGTase1 (purified from bovine brain) using the scintillation
proximity assay (SPA). The compoundswereassayed once and the
reliability of the assay wasꢄ50%. See ref 10b for further details.
20. (a) Ciccarone, T. M.; MacTough, S. C.; Williams, T. M.;
Dinsmore, C. J.; O’Neill, T. J.; Shah, D.; Culberson, J. C.;
Koblan, K. S.; Kohl, N. E.; Gibbs, J. B.; Oliff, A. I.; Graham,
S. L.; Hartman, G. D. Bioorg. Med. Chem. Lett. 1999, 9, 1991.
(b) Dinsmore, C. J.; Williams, T. M.; O’Neill, T. J.; Liu, D.;
Rands, E.; Culberson, J. C.; Lobell, R. B.; Koblan, K. S.;
Kohl, N. E.; Gibbs, J. B. Bioorg. Med. Chem. Lett. 1999, 9,
3301.
21. Subconfluent NIH3T3 ras-transformed cells were used for
the Ha-Ras processing assay. The compounds were assayed