The synthesis and biological evaluation of a series of potent dual inhibitors of farnesyl and geranyl-Geranyl protein transferases.

Article abstract of DOI:10.1016/S0960-894X(02)00308-6

We have prepared a series of potent, dual inhibitors of the prenyl transferases farnesyl protein transferase (FPTase) and geranyl-geranyl protein transferase I (GGPTase). The compounds were shown to possess potent activity against both enzymes in cell cul

Full text of DOI:10.1016/S0960-894X(02)00308-6

Bioorganic & Medicinal Chemistry Letters 12 (2002) 2027–2030
The Synthesis and Biological Evaluation of a Series of
Potent Dual Inhibitors of Farnesyl and Geranyl-Geranyl
Protein Transferases
Thomas J. Tucker,a,* Marc T. Abrams,b Carolyn A. Buser,b Joseph P. Davide,b
Michelle Ellis-Hutchings,b Christine Fernandes,b Jackson B. Gibbs,b Samuel L. Graham,a
George D. Hartman,a Hans E. Huber,b Dongming Liu,b Robert B. Lobell,b William
C. Lumma,a Ronald G. Robinson,b John T. Siskoa and Anthony M. Smitha
aDepartment of Medicinal Chemistry, Merck Research Laboratories, PO Box 4, Sumneytown Pike, West Point, PA 19486, USA
bDepartment of Cancer Biology, Merck Research Laboratories, PO Box 4, Sumneytown Pike, West Point, PA 19486, USA
Received 16 January 2002; accepted 12 March 2002
Abstract—We have prepared a series of potent, dual inhibitors of the prenyl transferases farnesyl protein transferase (FPTase) and
geranyl-geranyl protein transferase I (GGPTase). The compounds were shown to possess potent activity against both enzymes in
cell culture. Mechanistic analysis has shown that the compounds are CAAX competitive for FPTase inhibition but geranyl-geranyl
pyrophosphate (GGPP) competitive for GGPTase inhibiton. # 2002 Elsevier Science Ltd. All rights reserved.
Ras proteins are a family of GTPases which are pro-
cessed by a series of intracellular enzymes. Ras func-
tions primarily as a molecular on–off switch, cycling
between an inactive GDP-bound form and an active
GTP-bound form.1 The biologically inactive pro-Ras
peptide becomes activated by anchorage to the inner
surface of cell membranes through a series of post-
translational C-terminal modifications.1 Ras proteins
share a C-terminal CAAX motif, and the first and most
important post-translational modification covalently
adds a farnesyl group from farnesyl diphosphate onto
the cysteine residue of the CAAX sequence catalyzed by
farnesyl protein transferase (FPTase).2 The fully mod-
ified and anchored Ras proteins initiate a cascade of
phosphorylations, leading to a nuclear signal for cell
division. In the case of mutant Ras proteins, the signal
is always activated and is believed to be responsible for
the oncogenic response.1 Inhibition of FPTase has thus
been accepted as a viable approach to treating tumors in
which mutant Ras proteins are involved.
(GGPTase), resulting in geranyl-geranylated Ras pro-
teins that are functionally equivalent to the corre-
sponding farnesylated proteins.3 This suggests that
inhibition of FPTase in and of itself may not be com-
pletely effective as a means of blocking the signaling
process, as cross prenylation by GGPTase could restore
the function of the prenylated proteins involved in the
signaling cascade. Therefore, when attempting to block
mutant Ras protein function pharmacologically, it may
be critical to inhibit both FPTase and GGPTase activ-
ity. Sebti and co-workers4 have shown that inhibition of
FPTase is more effective than inhibition of GGPTase at
inhibiting tumor growth in mouse xenograft models;
however, these studies were conducted with less than
optimal inhibitors at extremely high exposures. A
critical need remains for highly potent dual FPTase-
GGPTase inhibitors that could be used experimentally
to further elucidate the roles of these enzymes in the
control of cellular proliferation. This paper describes
our efforts toward the design and synthesis of potent,
dual inhibitors of FPTase and GGPTase.
Upon inhibition of FPTase, N-Ras and K-Ras act as
substrates for geranyl-geranyl protein transferase I
In the preclinical evaluation of the Merck FPTase inhi-
bitor 1a (Fig. 1), it was noted that the compound also
inhibited GGPTase activity through a mechanism that
was GGPP competitive rather than CAAX competitive
as was seen with FPTase.5 This suggested that the
*Corresponding author. Fax: +1-215-652-7310; e-mail: tom_tucker@
merck.com
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00308-6

2028
T. J. Tucker et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2027–2030
molecule might occupy part of the binding region used
by GGPP in the GGPTase active site. A previous
manuscript6 has described a similar series of compounds
derived from 1a that possess dual inhibitor activity.
These compounds (Fig. 1, 1b) used large lipophilic
groups appended to the cyanobenzyl moiety of 1a to
provide a series that possessed activity against both
prenylation enzymes. The data presented in this pre-
vious manuscript showed that a large arylthio or aryl-
oxy group adjacent to the cyano group provided
compounds with potent activity against both enzymes.
The results obtained with this series of compounds as
well as with 1a indicate that both the large lipophilic
group appended to the cyanobenzyl moiety and the
m-chlorophenyl ring on the opposite end of the molecule
play a key role in binding to GGPTase.
Boc-piperazine was effected utilizing titanium isoprop-
oxide catalysis.10 Equivalent amounts of the aldehyde,
amine, and titanium isopropoxide were premixed and
stirred for 1 h, with a trace of THF added to provide
efficient stirring. The resulting slurry was diluted with
dry ethanol, and the mixture treated with 0.7 equiv of
NaBH3CN to give the desired products 4 in good yield.
These reductive amination conditions provided con-
sistently good results with this series of compounds.
Compound 4 was deprotected upon treatment with
TFA to provide the free amine 5. Reaction of 5 with the
appropriate p-nitrophenyl carbonate (prepared by reac-
tion of the appropriate alcohol with p-nitrophenyl
chloroformate) provided the target compounds 6–23 in
good yield.8
Compounds 6–23 were evaluated for their ability to
inhibit both FPTase and GGPTase in vitro. Com-
pounds were tested as FPTase inhibitors using purified
recombinant enzyme to catalyze the reaction between
[1-3H]FPP and a recombinant protein substrate con-
taining the K-ras C-terminus.11 The IC50 value reported
is the concentration of inhibitor required to reduce
radiolabel incorporation by 50%, as compared to unin-
hibited controls.11 Inhibition of GGPTase was eval-
uated in a scintillation proximity assay, by carrying out
the enzymatic reaction between [1-3H]GGPP and a
biotinylated peptide that represents the C-terminus of
the K4B-Ras protein in the presence of 5 mM ATP and
varying concentrations of inhibitor.12 IC50 values
were reported as described above for FPTase. Enzyme
inhibition data for the compounds are detailed in
Table 1; all reported values are the average of at least
two determinations.
Based on the discussion above, as well as previous
modeling studies with various compounds in this series,7
we theorized that the lipophilic m-chlorophenyl moiety
of 1a and 1b was occupying part of the lipophilic GGPP
binding region in the GGPTase active site. We therefore
began a synthetic program designed to investigate the
placement of larger, more lipophilic groups off of the
piperazinone ring rather than on the cyanobenzyl moi-
ety as previously reported.6 For synthetic ease, we used
the simple piperazine ring system, and replaced the
m-chlorophenyl ring with a series of aryl carbamate
moieties. Since compound 2 (Fig. 1), the prototype
compound for this series, exhibited reasonable potency
versus both enzymes, we followed this compound with a
series of substituted aryl carbamates.
Inhibitors 7–23 were synthesized as shown below in
Scheme 1.8 Reductive amination of the aldehyde 39 with
In general, most of the compounds evaluated in this
study were found to be potent inhibitors of both
enzymes. Addition of a small to moderate-sized lipo-
philic 2-substituent to the aryl ring of the carbamate
moiety provided compounds that demonstrated a
marked enhancement in both FPTase and GGPTase
inhibitory potency (6 and 8–13). Introduction of a more
polar methanesulfonyl moiety to the 2-position of the
aryl ring (14) tempered the activity against both
enzymes, although 14 remained a more potent inhibitor
of GGPTase than the unsubstituted aryl compound 2.
Compound 12 was an extremely potent, sub-nanomolar
inhibitor of both enzymes. Movement of the lipophilic
2-substituent to the 3-position (7 and 18) gave com-
pounds that exhibited a loss in inhibitory potency
towards both enzymes as compared to the correspond-
ing 2-substituted analogues. The 2,3-disubstituted
Figure 1. Structure and inhibitory potency for 1 and 2.
Scheme 1. Reagents and conditions: (a) Ti(OPr)4, NaBH3CN, Boc-piperazine/THF–EtOH; (b) TFA/CH2Cl2; (c) appropriate benzyl-(p-nitrophenyl)
carbonate, DIEA/DMF, 80 ꢀC.

T. J. Tucker et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2027–2030
2029
analogues 15–17 retained good inhibitory potency ver-
sus both enzymes; however, the activity was somewhat
less than that observed for the 2-monosubstituted deri-
vatives. The 3,4-methylenedioxy derivative 19 demon-
strated a marked loss in potency versus both enzymes
when compared to the 2,3-disubstuted 17, suggesting
that a 4-substituent is detrimental to activity versus
both enzymes.
Several of the more potent inhibitors of GGPTase
(Table 1, 9, 12, 13, 22, and 23) were evaluated for their
GGPTase inhibitory activity in cells, as measured by the
inhibition of geranyl-geranylation of the C-terminal
CAAX sequence of Rap 1a in PSN-1 cells (Table 1).15
Reported values are the average of at least two deter-
minations. All four compounds exhibited potent Rap 1a
activity in PSN-1 cells, with the potency correlating
directly with the observed GGPTase potency in the
enzyme assay.
Addition of a methyl group to the 2-position of the
imidazole ring (Table 1, 20–23) led to profound activity
changes compared to the previous compounds. While
these compounds remained potent inhibitors of FPTase,
a marked decrease in GGPTase inhibitory potency was
observed for all of these analogues. The structural basis
for the decreased GGPTase inhibitory activity exhibited
by these 2-methyl imidazole analogues is unclear at
present.
Further biochemical evaluation16 of a number of potent
inhibitors (9, 12, 13, 15, 16, 20, and 23) demonstrated
that all of the potent GGPTase inhibitors were GGPP
competitive and non-competitive with CAAX peptide.17
In contrast, evaluation of the mechanism of FPTase
inhibition indicated that these compounds were in gen-
eral non-competitive with respect to FPP, and competi-
tive with CAAX peptide. All of the GGPP competitive
compounds displayed a marked dependence on phos-
phate anions to achieve maximal potency, with ATP
among the most effective anions.5 The combination of
inhibitor and phosphate was found to compete effec-
tively with GGPP binding and to lower IC50 values by
as much as 200-fold when compared to inhibitor alone.5
These experimental findings suggest that this group of
compounds is likely to occupy at least part of the prenyl
binding region of the GGPTase active site. Phosphates
are believed to competitively bind in the pyrophosphate
site of the GGPTase binding site, thereby displacing
GGPP and facilitating binding of the inhibitors. This
group of compounds also displayed slow-binding
All of the compounds that were shown to be potent
inhibitors of FPTase were also evaluated for FPTase
inhibitory activity in cells by a cell-based radiotracer
assay for farnesyl transferase inhibition (CRAFTI).13 In
this assay conducted in cultured H-ras transformed
Rat1 cells, the concentration of compound required to
displace 50% of a highly potent radiolabeled FPTase
inhibitor14 from FPTase was determined and reported
as an IC50 value. Reported values are the average of at
least two determinations. All of the potent in vitro
inhibitors of FPTase also possessed potent FPTase
inhibitory activity in our cell-based CRAFTI assay
(Table 1).
Table 1. FPTase and GGPTase inhibitory and cell culture activity for 2 and 6–23
Compda
R
Ar
FPTase
IC50 (nM)b
GGPTase
IC50 (nM)b
CRAFTI
IC50 (nM)b
Rap 1a
EC50 (nM)b
2
6
7
8
HPh
H2-MePh
H3-MePh
H2-(OMe)Ph
H2-(OEt)Ph
H2-Cl Ph
62.5
6.4
34
60.6
9.1
76
21.8
62.9
166
ND
ND
ND
n=2)
71
3.1
1.6
6.8
0.84
0.32
0.68
4, 27 (
ND
9
4.52
7.95
10
11
12
13
14
15
16
17
18
19
20
21
22
23
ND
2)PHh2-(MeOCH
H2-(OCF3)Ph
2)PHh 2-(F3CCH
H2-(MeSO2)Ph
6.1
2.5
0.32
4.82
9.1
ND
50
0.36
0.16
2.2
18
0.85
99
18
ND
ND
ND
ND
ND
H2-(OMe)-Pyridin-3-yl
H2-(OEt)-Pyridin-3-yl
H2,3-Methylenedioxy Ph
H3-(OCF3)Ph
18
5
48
8.7
19.8
13
8.67
10.9
15.5
13.4
74
12.1
ND
H3,4-Methylenedioxy Ph
102
121
ND
ND
Me
Me
Me
Me
Ph
2-(OEt)Ph
2-(MeOCH2)Ph
2-(OCF3)Ph
19
1.5
17
0.33
789
80
178
53
ND
1.13
3.17
0.34
ND
ND
680
660
aAll compounds tested in biological assays were fully characterized and exhibited satisfactory analytical (NMR, HPLC purity >98%, HRMS) data.
bDetails of assay conditions are provided in the text and references cited therein. All cited values are the average of at least two determinations.

2030
T. J. Tucker et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2027–2030
characteristics on GGPTase, demonstrating a 20-fold
enhancement in GGPTase inhibitory potency after a
30-min preincubation of GGPTase and inhibitor (prior
to addition of GGPP).5 This procedure was necessary
to obtain accurate GGPTase inhibition values for
compounds in this class. The reaction mechanism of
many prenyl-protein transferases utilizes a carbocation-
pyrophosphate intermediate. The combination of slow-
binding kinetics and phosphate dependence suggests
that these inhibitors may be acting as transition state
analogues, where the phosphate may be a mimic of the
pyrophosphate leaving group.5
Fernandes, C.; Gibbs, J. B.; Graham, S. L.; Hartman, G. D.;
Heimbrook, D. C.; Homnick, C. F.; Huff, J. R.; Kassahun,
K.; Koblan, K. S.; Kohl, N. E.; Lobell, R. B.; Lynch, J. J., Jr.;
Miller, P. A.; Omer, C. A.; Rodrigues, A. D.; Walsh, E. S.;
Williams, T. M. J. Med. Chem. 2001, 44, 2933.
8. Satisfactory analytical data (NMR, HPLC purity >98%,
HRMS) were obtained for all final compounds that were
tested in biological assays. Full experimental details for the
synthesis of these compounds are presented in the following
patent: Hartman, G. D.; Lumma, W. C.; Sisko, J. T.; Smith,
A. M.; Tucker, T. J.; Stokker, G. E. PCT International
Application WO 00/51611, 2000.
9. Williams, T. M.; Bergman, J. M.; Brashear, K.; Breslin,
M. J.; Dinsmore, C. J.; Hutchinson, J. H.; MacTough, S. C.;
Stump, C. A.; Wei, D. D.; Zartman, C. B.; Bogusky, M. J.;
Culberson, J. C.; Buser-Doepner, C.; Davide, J.; Greenburg,
I. B.; Hamilton, K. A.; Koblan, K. S.; Kohl, N. E.; Liu, D.;
Lobell, R. B.; Mosser, S. D.; O’Neill, T. J.; Rands, E.; Scha-
ber, M. D.; Wilson, F.; Senderak, E.; Motzel, S. L.; Gibbs,
J. B.; Graham, S. L.; Heimbrook, D. C.; Hartman, G. D.;
Oliff, A. I.; Huff, J. R. J. Med. Chem. 1999, 42, 3779.
10. The reductive amination procedure was adapted from a
procedure described in the following paper: Mattson, R. J.;
Pham, K. M.; Leuck, D. J.; Cowen, K. A. J. Org. Chem. 1990,
55, 2552.
In summary, we have discovered a series of highly
potent, dual FPTase/GGPTase inhibitors. Representa-
tive compounds were shown to be potent inhibitors of
both enzymes in cell culture, and were shown to be
GGPP competitive for GGPTase inhibition but CAAX
competitive for FPTase inhibition. These compounds
represent a new class of highly potent dual prenylation
inhibitors, and may be important experimental tools for
elucidating the role of prenylation inhibitors as novel
chemotherapeutic agents.
11. Experimental details of the FPTase assay procedure are
presented in the patent application cited in ref 8.
12. Experimental details of the GGPTase assay procedure are
presented in refs 5 and 8.
References and Notes
1. Rowinsky, E. K.; Windle, J. J.; Von Hoff, D. D. J. Clin.
Oncol. 1999, 17, 3631.
2. Graham, S. L. Exp. Opin. Ther. Pat. 1995, 5, 1269.
3. Whyte, D. B.; Kirschmeier, P.; Hockenberry, T. N. J. Biol.
Chem. 1997, 272, 14459.
13. Experimental details of the cell based radiotracer assay for
farnesyl transferase (CRAFTI) are presented in the patent
application cited in ref 8.
14. The radiolabeled FPTase inhibitor used in this assay is
125I]-4-{[5-(2S)-4-(3-iodophenyl)-2-[2-(methylsulfonyl)ethyl]-5-
[
4. Sun, J.; Qian, Y.; Hamilton, A. D.; Sebti, S. M. Oncogene
1998, 18, 1467.
oxopiperazin-1-yl]methyl}benzonitrile. This compound has
approximately 50,000 high-affinity binding sites (apparent Kd
ꢁ1 nM) in Rat1 cells.
5. Huber, H. E.; Robinson, R. G.; Watkins, A.; Nahas, D. D.;
Abrams, M. T.; Buser, C. A.; Lobell, R. B.; Patrick, D.;
Anthony, N. J.; Dinsmore, C. J.; Graham, S. L.; Hartman,
G. D.; Lumma, W. C.; Williams, T. M.; Heimbrook, D. C. J.
Biol. Chem. 2001, 276, 24457.
15. Experimental details of the Rap 1a cell-based assay are
provided in refs 5 and 8.
16. Details of the experimental techniques used in the enzyme
kinetics analysis are detailed in ref 5.
6. Bergman, J. M.; Abrams, M.; Davide, J.; Greenburg, I. B.;
Robinson, R. G.; Buser, C. A.; Huber, H. E.; Koblan, K. S.;
Kohl, N. E.; Lobell, R. B.; Graham, S. L.; Hartman, G. D.;
Williams, T. M.; Dinsmore, C. J. Bioorg. Med. Chem. Lett.
2001, 1411, 1411.
17. The slope of a plot of log(IC50) versus log[GGPP] in the
presence of 5 mM ATP was used to determine the mechanism
of GGPTase inhibition. Slope values were determined for the
following compounds: 9, slope=1.05; 12, slope=0.84; 13,
slope=0.99; 15, slope=0.84; 16, slope=0.90; 20, slope=0.63;
23, slope=0.67. All of the slope values are >0.50 and are
strongly suggestive of a GGPP competitive inhibitor.
7. Bell, I. M.; Gallicchio, S. N.; Abrams, M.; Beshore, D. C.;
Buser, C. A.; Culberson, J. C.; Davide, J.; Ellis-Hutchings, M.;