Pyridone-containing farnesyltransferase inhibitors: Synthesis and biological evaluation

Article abstract of DOI:10.1016/j.bmcl.2003.08.058

Farnesyltransferase inhibitors (FTIs) have been developed as potential anti-cancer agents due to their efficacy in blocking malignant growth in a variety of murine models of human tumors. To that end, we have developed a series of pyridone farnesyltransferase inhibitors with potent in vitro and cellular activity. The synthesis, SAR and biological properties of these compounds will be discussed.

Full text of DOI:10.1016/j.bmcl.2003.08.058

Bioorganic & Medicinal Chemistry Letters 13 (2003) 4001–4005
Pyridone-Containing Farnesyltransferase Inhibitors:
Synthesis and Biological Evaluation
Lisa A. Hasvold,* Weibo Wang, Stephen L. Gwaltney, II, Todd W. Rockway,
Lissa T. J. Nelson, Robert A. Mantei, Stephen A. Fakhoury, Gerard M. Sullivan, Qun Li,
Nan-Horng Lin, Le Wang, Haiying Zhang, Jerome Cohen, Wen-Zhen Gu,
Kennan Marsh, Joy Bauch, Saul Rosenberg and Hing L. Sham
Pharmaceutical Discovery, R-47B, Abbott Laboratories, 100 Abbott Park Road, Abbott Park, IL 60064-6101, USA
Received 3 June 2003; revised 18 August 2003; accepted 27 August 2003
Abstract—Farnesyltransferase inhibitors (FTIs) have been developed as potential anti-cancer agents due to their efficacy in block-
ing malignant growth in a variety of murine models of human tumors. To that end, we have developed a series of pyridone farne-
syltransferase inhibitors with potent in vitro and cellular activity. The synthesis, SAR and biological properties of these compounds
will be discussed.
# 2003 Elsevier Ltd. All rights reserved.
Ras oncogenes are present in a variety of human
tumors. The dependence of Ras proteins on post-trans-
lational farnesylation for cell membrane anchoring led
to the investigation of farnesyltransferase inhibitors
(FTIs) as potential anti-cancer agents.1 These agents
possess the ability to block or reverse malignant growth
while leaving non-cancerous cells virtually unaffected.
The mechanism through which they work is still under
debate, as it has been shown that inhibition of Ras iso-
prenylation is not necessary for their antineoplastic
effects.2
Early work on farnesyltransferase inhibitors in our
laboratories yielded ABT-839 (Fig. 1) which demon-
strated good in vitro activity (IC50=1.1 nM), selectivity
(FT/GGT-I=75,000) and cellular potency (EC50=16
nM). Unfortunately, the compound suffered from poor
oral bioavailability. Extensive SAR studies in our group
led to the discovery that a cyano group can replace the
methionine moiety of ABT-839. The tertiary amine
moiety was also replaced by a methylimidazole-second-
ary alcohol moiety. Significant improvements were seen
with these changes. An exemplary compound, A-315493
(Fig. 1), is a potent FTI with an IC50 of 0.73 nM and an
EC50 of 5.9 nM. It also has a much-improved pharma-
cokinetic profile with approximately 50% oral bioavail-
ability in both dog and monkey. It possesses only
In addition to farnesyltransferase, there are two other
protein-isoprenyltransferases, geranylgeranyltransferase
type I (GGT-I) and geranylgeranyltransferase type II
(GGT-II). FT and GGT-I act similarly transferring C-
15 and C-20 units, respectively; however, GGT-II acts
through a different mechanism. The majority of iso-
prenylated proteins are geranylgeranylated by GGT-I,
but all mutant Ras proteins in human cancers are mod-
ified exclusively by FT.2 Evidence that simultaneous
inhibition of both FT and GGT-I can lead to toxicity
prompted us to seek a selective inhibitor of FT over
GGT-I.3
*Corresponding author. Tel.: +1-847-937-7217; fax: +1-847-935-
5165; e-mail: lisa.hasvold@abbott.com
Figure 1. ABT-839 and A-315493.
0960-894X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2003.08.058

4002
L. A. Hasvold et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4001–4005
modest selectivity, however, at 26-fold for FT inhibition
over GGT-I inhibition.4
Coupling of intermediates 3 and 8 under Dean–Stark
conditions with p-TsOH in toluene provided com-
pounds 9 and 10.
Continued efforts toward the discovery of a potent and
selective FTI produced several structurally different
series of compounds including the pyridone series
shown in Figure 2.
The preparation of certain phenylboronic acids used to
synthesize the compounds of Table 1 is shown in
Scheme 2. Each was produced through anion formation
with butyllithium, followed by addition of B(Oi-Pr)3
and acidic workup.
The effects of substituent changes with respect to in
vitro activity (IC50),5 cellular potency or Ras processing
(EC50),5 and pharmacokinetic properties (specifically
oral bioavailability)6 of the pyridone series were studied
and the results are now presented.7
Table 1 illustrates the effects of various substituents on
ring A. All of the substituents provided compounds that
A representative synthesis of the pyridones is shown in
Scheme 1. Intermediate 3 was prepared by treatment of
2 with t-BuLi followed by addition of the appropriate 4-
cyanobenzaldehyde. Intermediate 8 was prepared by
initial bromination of 3-cyano-6-methyl-2-(1H)-pyr-
idinone (4) with NBS, followed by N-methylation with
K2CO3 and MeI. Subsequent Suzuki reaction with var-
ious phenylboronic acids, followed by benzylic bromi-
nation with NBS/AIBN and hydrolysis in the presence
of Celite1 completed the synthesis of intermediate 8.
Scheme 1. (i) (a) n-BuLi, THF, ꢀ78 ꢁC; (b) TESCl, ꢀ78 ꢁC to rt, 82%;
(ii) (a) t-BuLi, THF, ꢀ78 ꢁC; (b) 4-cyanobenzaldehyde, ꢀ78 ꢁC, 79%;
(iii) (a) NBS, DCE, reflux, 92%; (b) K2CO3, DMF, rt; MeI, 0 ꢁC,
76%; (iv) substituted phenylboronic acid, Pd(PPh3)4, 2 M N aCO3,
2
LiCl, toluene/EtOH, reflux, 84% for X=3-Cl; (v) NBS, AIBN, CCl4,
reflux; (vi) Celite1, 1,4-dioxane/H2O, reflux, 88% over two steps for
X=3-Cl; (vii) p-TsOH, toluene, reflux, 43% for X=3-Cl.
Figure 2. Structure of the pyridone series of FTIs.
Table 1. IC50, EC50 and oral bioavailability (F) for compounds 9a–u
Compd
X
FT IC50 (nM)
GGT-I IC50 (nM)
FT EC50 (nM)
F
Dog (%)
Mky (%)
9a
9b
9c
9d
9e
9f
9g
9h
9i
3-Cl
3,5-diCl
3,4-diCl
2,5-diCl
2,3-diCl
3-F
3,5-diF
3,4-diF
2,4-diF
2,5-diF
2,3-diF
3-Cl,5-F
3-Cl,4-F
2-F,3-Cl
3-Br
3-OCH3
3-OCH2CH3
3-OCF3
4-OCF3
3,4-OCH2O–
3,4-OCF2O–
0.44
0.46
0.49
1.6
6.1
2.1
0.84
1.5
5.5
2.7
6.3
0.7
1.1
1.8
0.51
0.41
0.67
0.42
0.73
0.42
0.69
2700
1700
2300
2900
0.5
6.8
86
1.2
na
11
7.3
10
32
52
76
54
38
61
90
83
na
>99
na
40
91
59
61
44
47
93
52
28
63
14
46
na
3
2500
na
42
25
na
na
46
na
na
28
na
3
0
3
36
34
na
34
>10,000
12,000
5500
4300
>100
110
na
9j
9k
9l
15,000
3400
1800
1900
1500
3400
3200
1500
1045
17
15
24
2.1
4.7
0.81
3.9
3.9
2.1
19
9m
9n
9o
9p
9q
9r
9s
9t
1400
3300
6200
9u
Dog, beagle; Mky, cynomolgus monkey.

L. A. Hasvold et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4001–4005
4003
Scheme 2. (i) (a) n-BuLi, THF, ꢀ78 ꢁC; (b) B(Oi-Pr)3, ꢀ78 ꢁC to rt;
(c) HCl; (ii) (a) t-BuLi, THF, ꢀ78 ꢁC; (b) B(Oi-Pr)3, ꢀ78 ꢁC to rt; (c) HCl.
Scheme 3. (i) Chlorophenylboronic acid, Pd(OAc)2, CsF, 2-dicyclo-
hexylphosphino-20-(N,N-dimethylamino)biphenyl, 1,4-dioxane, 91%
for 3-Cl; (ii) (a) NaBH4, CaCl2, EtOH/THF, 0 ꢁC to rt, 95% for 3-Cl;
(b) Dess–Martin periodinane, CH2Cl2, quant for 3-Cl.
Table 2. IC50 and EC50 for compounds 9a, 9b, 9s, 10a–m
reduced by in situ formation of Ca(BH4)2, followed by
Suzuki reaction with the appropriate chlorophenyl
boronic acid and oxidation with Dess–Martin period-
inane.
Table 2 illustrates the effects of changes to the cyano-
phenyl ring (B). The changes were tested using some of
the better substituents from Table 1 for ring A. All
substituents provided potent FTIs. Larger differences
were seen in the GGT-I assay and Ras processing
(EC50). Selectivity for FT inhibition over GGT-I inhi-
bition was generally reduced by the addition of a sub-
stituent to the cyanophenyl ring. The most marked
decreases were seen with the large chlorophenyl group
(10d, 10h, 10k–m) where selectivity was as low as 23-
fold. The substituent changes also resulted in dimin-
ished potency in the cellular assay.
Compd
Structure
(Y; X)
FT
IC50 (nM)
GGT-I
IC50 (nM)
FT
EC50 (nM)
9a
H; 3-Cl
Cl; 3-Cl
Br; 3-Cl
0.44
0.97
1.3
0.75
1.0
0.46
0.77
0.42
0.81
2.0
0.73
0.36
0.89
3.9
2700
890
610
>1000
94
1700
190
290
7500
120
1400
150
220
110
30
160
0.5
24
39
88
21
6.8
10a
10b
10c
10d
9b
10e
10f
10g
10h
9s
10i
10j
10k
10l
10m
I; 3-Cl
30-ClPh; 3-Cl
H; 3,5-diCl
Cl; 3,5-diCl
Br; 3,5-diCl
I; 3,5-diCl
16
>100
>100
na
30-ClPh; 3,5-diCl
H; 4-OCF3
Cl; 4-OCF3
Br; 4-OCF3
20-ClPh; 4-OCF3
30-ClPh; 4-OCF3
40-ClPh; 4-OCF3
3.9
15
10
We further examined replacement of the pyridone N-
methyl group with other substituents. Scheme 4 illus-
trates the synthesis of compound 22, the key inter-
mediate for the synthesis of 26, the N–H counterpart to
compound 9a. Compound 4 (Scheme 1) was brominated
with NBS. This was followed by Suzuki reaction with 3-
chlorophenylboronic acid and O-silylation with TBSCl.
Sequential benzylic bromination with NBS/AIBN, fol-
lowed by hydrolysis and concomitant deprotection in
the presence of Celite1 provided intermediate 22.
Intermediates 3 and 22 were then coupled as described
in Scheme 1 to provide 26.
>100
>100
>100
1.3
6.5
are potent against FT with more than half of them
possessing FT IC50 values less than 1 nM. In addition,
all substituents provided selectivities greater than 400-
fold for inhibition of FT over GGT-I. The 3-Cl (9a) and
3-OEt (9q) substitutions provided compounds with the
best cellular potency having EC50 values of 0.5 and 0.81
nM, respectively. The 3,5-diCl (9b) and 3,5-diF (9g)
compounds possess similar EC50 values of 6.8 and 7.3
nM, respectively, unlike the 2,5- and 3,4-dihalogenated
compounds (9d, 9j, 9c and 9h) which have significantly
different EC50 values. Of the 2,5-dihalogenated com-
pounds, the dichloro compound is more potent by
nearly 92-fold, while the reverse is seen for the 3,4-
dihalogenated compounds in which the difluoro com-
pound is more potent. Reasons for the differences seen
in the EC50 values of the dichloro and difluoro com-
pounds remain unclear. Although substituent changes
caused wide pharmacokinetic variability in both species,
oral bioavailability was consistently better in dog than
monkey. Several compounds show balanced properties
throughout FT IC50 and EC50, selectivity and oral
bioavailability (9a, 9b, 9g, 9r and 9s).
Scheme 4. (i) (a) NBS, DCE, reflux, 92%; (b) 3-chlorophenylboronic
acid, Pd(PPh3)4, 2M N aCO3, LiCl, toluene/EtOH, reflux, 73%;
2
(ii) TBSCl, imidazole, DMF, 58%; (iii) (a) NBS, AIBN, CCl4, reflux;
(b) Celite1, 1,4-dioxane/H2O, reflux, 33%.
We also investigated the effects of an added substituent
on the cyanophenyl ring (B). Synthesis of the chloro-
phenyl substituted 4-cyanobenzaldehydes (19) used in
the preparation of compounds 10d, 10h and 10k–m
(Table 2) is shown in Scheme 3. Known ester (17)4 was
Scheme 5. (i) allylbromide, NaH (60%), LiBr, DME/DMF, 0 ꢁC to rt
to 65 ꢁC, 57%; (ii) 10% Pd/C, EtOAc, H2 (1 atm.) rt, 70%;
(iii) CH2N2/ether, THF, Pd(OAc)2, 0 ꢁC, 55%.

4004
L. A. Hasvold et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4001–4005
Table 3. IC50, EC50 and oral bioavailability (F) for compounds 9a, 26–29
Compd
R
FT IC50 (nM)
GGT-I IC50 (nM)
FT EC50 (nM)
F
Dog (%)
Mky (%)
9a
26
27
28
29
CH3
H
CH3CH2
CH3CH2CH2
0.44
0.85
0.73
0.036
0.5
2700
820
3300
3000
2000
0.5
>100
2.9
32
na
na
60
na
14
na
na
na
na
12
4.8
Dog, beagle; Mky, cynomolgus monkey.
Table 4. IC50 and EC50 for compounds 9a and 32–35
Compd
Structure R; X
; 3-Cl
FT IC50 (nM)
0.44
GGT-I IC50 (nM)
2700
FT EC50 (nM)
0.5
9a
32
33
34
; 3-Cl
; 3-Cl
71
na
na
5.3
1.9
9.9
>900
68–75% Inh. @ 100 nM
; 3,5-diCl
; 4-OCF3
>10,000
>10,000
>100
35
na
Inh., inhibition.
Compound 27 was prepared as shown in Scheme 1 with
nitrogen alkylation being effected with ethyl iodide in
place of iodomethane. Compounds 28 and 29 were pre-
pared as shown in Schemes 1 and 5. Compound 4 was
alkylated according to the procedure described by Cur-
ran.8 The resulting compound (23) was either hydro-
genated to form intermediate 24 or cyclopropanated to
form intermediate 25. The remainder of the synthesis
follows Scheme 1.
Scheme 6. (i) (a) EtMgBr, CH2Cl2, rt; (b) 4-cyanobenzaldehyde,
CH2Cl2, ꢀ20 ꢁC to rt, quant; (ii) (a) 7, Ag2O, CH2Cl2, 28%; (b) 80%
aq AcOH, 11%.
It can be seen in Table 3 that FT inhibition, selectivity
and cellular potency were maintained, except in the case
of R=H (26). Compound 26 was potent against FT and
selective, but it showed weak cellular potency (EC50
>100 nM). Although it appears that a nitrogen sub-
stituent is required for cellular potency, groups of
various sizes can be tolerated.
Finally, we examined the effects of changing the
N-methylimidazole ring. Compound 32 was prepared as
shown in Scheme 6. 4-Iodo-1-trityl-1H-imidazole (30)
was treated with EtMgBr in CH2Cl2, followed by
4-cyanobenzaldehyde according to the procedure of

L. A. Hasvold et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4001–4005
4005
Turner et al.9 This intermediate (31) was then coupled
References and Notes
with intermediate 7 (Scheme 1) using Ag2O. Removal of
the trityl group with aqueous acetic acid provided com-
pound 32. Compounds 33–35 (Table 4) were prepared
as shown in Scheme 1 with replacement of thiazole or
1,2-dimethylimidazole for N-methylimidazole.
1. Prendergast, G. C.; Rane, N. Expert Opin. Investig. Drugs
2001, 10, 2105.
2. Prendergast, G. C. Curr. Opin. Cell Biol. 2000, 12, 166.
3. Lobell, R. B.; Omer, C. A.; Abrams, M. T.; Bhimnathwala,
H. G.; Brucker, M. J.; Buser, C. A.; Davide, J. P.; deSolms,
S. J.; Dinsmore, C. J.; Ellis-Hutchings, M. S.; Kral, A. M.;
Liu, D.; Lumma, W. C.; Machotka, S. V.; Rands, E.; Wil-
liams, T. M.; Graham, S. L.; Hartman, G. D.; Oliff, A. I.;
Heimbrook, D. C.; Kohl, N. E. Cancer Res. 2001, 61, 8758.
4. Wang, W.; Curtin, M. L.; Fakhoury, S. A.; Gwaltney, S.
L., II; Hasvold, L. A.; Hutchins, C. W.; Li, Q.; Lin, N.; Nel-
son, L. T. J.; O’Connor, S. J.; Sham, H. L.; Sullivan, G. M.;
Wang, G. T.; Wang, X. World Patent WO 01/81316; Chem
Abstr. 2001, 135, 344482.
5. Assay methods described in: Vogt, A.; Qian, Y.; Blas-
kovich, M. A.; Fossum, R. D.; Hamilton, A. D.; Sebti, S. M.
J. Biol. Chem. 1995, 270, 660.
6. Pharmacokinetic properties were determined by traditional
methods or through multiple dosing studies where three com-
pounds and a reference compound were administered at 1 mg/
kg in the appropriate species for a total dose of 4 mg/kg.
Plasma concentrations were determined by HPLC-MS.
7. All compounds reported have physical data consistent with
the assigned structures.
As seen in Table 4, compound 32 exemplifies the
importance of the nitrogen methyl group. A 140-fold
loss of potency was seen compared to compound 9a.
Addition of a second methyl group to the imidazole ring
(33) also showed a loss of potency, but to a lesser extent.
Replacement of N-methylimidazole with thiazole pro-
vided varying losses of potency as seen in examples 34
and 35. Although selectivity was maintained in these
examples, a critical loss was suffered in the Ras proces-
sing of example 34 which has an EC50 greater than 100
nM. Overall, these changes were not well tolerated and
among these limited examples, the N-methylimidazole
ring remains the best choice.
We have demonstrated that, with few exceptions, a
variety of substituent changes were well tolerated in the
in vitro assay. Although selectivity and cellular potency
varied widely with the substituent changes, we have
shown the pyridone series of FTIs to be potent and
selective with good PK profiles.
8. Liu, H.; Ko, S.-B.; Josien, H.; Curran, D. P. Tetrahedron
Lett. 1995, 36, 8917.
9. Turner, R. M.; Lindell, S. D.; Ley, S. V. J. Org. Chem.
1991, 56, 5739.