Non-peptidic, non-prenylic bisubstrate farnesyltransferase inhibitors. Part 3: Structural requirements of the central moiety for farnesyltransferase inhibitory activity

Article abstract of DOI:10.1016/S0968-0896(00)00173-5

Recently, we have described non-peptidic, non-prenylic bisubstrate analogues as a novel type of farnesyltransferase inhibitor composed of a farnesyl-mimetic, a linker and an AAX-peptidomimetic substructure. With this study, we showed that the amide function connecting the farnesyl-mimetic and the linking substructures of our inhibitors is crucial for their activity. We suggest that the amide is bound to the essential zinc ion in the farnesyltransferases active center. We identified succinic and glutaric acid, respectively, in addition to the initially used β-alanyl moiety as suitable linking structures. Glycine can also be used in this function provided the distance between the α-amide group and the center of the peptidomimetic substructure is enlarged by introduction of an additional methylene unit into the peptidomimetic substructure. Copyright (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0968-0896(00)00173-5

Bioorganic & Medicinal Chemistry 8 (2000) 2399±2406
Non-Peptidic, Non-Prenylic Bisubstrate Farnesyltransferase
Inhibitors. Part 3: Structural Requirements of the Central Moiety
for Farnesyltransferase Inhibitory Activity
Martin Schlitzer,a,* Markus Bohma and Isabel Sattlerb
aInstitut fur Pharmazeutische Chemie, Philipps-Universitat Marburg, Marbacher Weg 6, D-35032 Marburg, Germany
È
È
bHans-Knoll-Institut fur Natursto€-Forschung e. V., Beutenbergstraûe 11, D-07745 Jena, Germany
È
È
Received 14 June 1999; accepted 13 May 2000
AbstractÐRecently, we have described non-peptidic, non-prenylic bisubstrate analogues as a novel type of farnesyltransferase
inhibitor composed of a farnesyl-mimetic, a linker and an AAX-peptidomimetic substructure. With this study, we showed that the
amide function connecting the farnesyl-mimetic and the linking substructures of our inhibitors is crucial for their activity. We
suggest that the amide is bound to the essential zinc ion in the farnesyltransferases active center. We identi®ed succinic and glutaric
acid, respectively, in addition to the initially used b-alanyl moiety as suitable linking structures. Glycine can also be used in this
function provided the distance between the a-amide group and the center of the peptidomimetic substructure is enlarged by intro-
duction of an additional methylene unit into the peptidomimetic substructure. # 2000 Elsevier Science Ltd. All rights reserved.
Introduction
the second substrate, farnesylpyrophosphate, are much
less common. A third approach for inhibitor design is
bisubstrate analogues: molecules that contain structural
elements of both substrates, the CAAX-tetrapeptide
and the farnesylpyrophosphate.2
Ras proteins serve as molecular switches in the signal
transduction cascade which controls cell di€erentiation
and proliferation. Point mutations in the ras oncogene
yield ras proteins locked in their GTP-bound active
state. In approximately 30% of all human cancers,
including 90% of pancreatic, 50% of colon, and 50% of
thyroid tumors, these oncogenic ras proteins are found.
To perform both their normal and their oncogenic activ-
ity, ras proteins must be farnesylated. The C-terminal
CAAX amino acid sequence (C: cysteine, A: aliphatic
amino acid, X: methionine or serine) of the ras protein
is recognized by the enzyme farnesyltransferase (FTase)
and farnesylated at the cysteine thiol function. In this
reaction farnesylpyrophosphate (FPP) is serving as the
prenyl group donor.1
We have recently reported the development of a novel
type of bisubstrate analogue inhibitor of the farnesyl-
transferase.3 These novel compounds (e.g., 1) are con-
veniently composed of three modular building blocks: (1)
a farnesyl-mimetic, (2) a linker and (3) an AAX- or AA-
peptidomimetic, thus representing novel non-peptidic,
non-prenylic bisubstrate analogue farnesyltransferase
inhibitors. In contrast to other bisubstrates, our com-
pounds lack peptidic substructures which are sensitive
towards hydrolytic or enzymatic degradation and they
lack the farnesyl residue which is susceptible towards
oxidation due to its allylic positions.
Inhibition of the farnesylation reaction reverses the trans-
formation caused by oncogenic ras. Therefore, inhibition
of the farnesyltransferase was identi®ed as a promising
target in cancer therapy. Most known farnesyltransferase
inhibitors are peptidomimetic analogues of the CAAX
recognition sequence. Stable non-substrate analogues of
Recent crystallographic studies provide considerable
insight into the structure of a ternary FTase±FPP±CAAX
complex.4 The CAAX tetrapeptide has been shown to
bind in an extended conformation with its thiol sulfur
coordinated to the zinc ion located in the FTase active
site4 as has been postulated5 (Fig. 1(a)). We imagine that
our bisubstrate inhibitors should occupy considerable
portions of the peptide as well as the farnesyl binding
regions in the FTase active center (Fig. 1(b)). We sug-
gest that our bisubstrate analogue inhibitors are inserted
*Corresponding author. Tel.: +49-6421-28-5825; fax: +49-6421-28-
7052; e-mail: schlitze@mailer.uni-marburg.de
0968-0896/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(00)00173-5

2400
M. Schlitzer et al. / Bioorg. Med. Chem. 8 (2000) 2399±2406
into the farnesyltransferase, in a way that the carbonyl
oxygen of the amide group which connects the farnesyl
surrogate to the linking structure is coordinated to the
zinc ion in the active center. This interaction should be
crucial for the inhibitory activity of our inhibitors.
Therefore, the length and the nature of the linking
structure which governs the position of this presumably
essential carbonyl group should be crucial for the inhi-
bitory activity of our bisubstrate inhibitors. In the pre-
sent study we chose compound 1 as a lead and replaced
the b-alanyl linker by various moieties (Figure 2, Table
1). These variations were accompanied by minor varia-
tions of the overall dimensions of the molecules. In a
previous study,6 we have shown that minor variations in
the length of the farnesyl-mimetic do not change the
inhibitory activity signi®cantly so that di€erences in the
inhibitory activity can be assigned to the variations of
the linking moiety.
acylated with N-Boc-glycine and N-Boc-g-aminobutyric
acid, respectively, using mixed anhydride activation.
The resulting amides 12 and 13 were deprotected with
HCl in dioxane and acylated with palmitic acid chlor-
ide. Heating of 11 with succinic and glutaric anhydride,
respectively, in toluene/dioxane yielded the acids 14 and
15, which were coupled with palmityl amine to the target
compounds 5 and 6, using the PyBOP methodology.
Compound 16 was obtained from 4-(tert-butyloxycar-
bonylaminomethyl)benzoic acid8 and methionine methyl
ester hydrochloride. Compound 16 was deprotected and
acylated with stearoyl chloride yielding 7. Compounds 8
and 9 were obtained from deprotected 16 using the same
sequence as described for the synthesis of 3.
Farnesyltransferase inhibition assay
The inhibitory activity of the inhibitors was determined
using the ¯uorescence enhancement assay as described
Chemistry
The synthesis of the target compounds 2±10 is outlined
in Scheme 1. Compound 2 was obtained from 4-amino-
benzoylmethionine methyl ester 117 and arachidic acid,
which was activated with iso-butyl chloroformate as a
mixed anhydride. Compound 10 was obtained in the
same way using palmitoyl-N-methyl-b-alanine instead
of arachidic acid. For the synthesis of 3 and 4, 11 was
Figure 1. Schematic model of the FTase active center with: (a) the
natural substrates; (b) compound 1 in the suggested binding mode.
Table 1. Farnesyltransferase inhibition of compounds 1±10
Compound
IC50 Æ SD (mM)
1
2
3
4
5
6
7
8
9
10
3.7Æ0.7
>100
12.0Æ1.1
13.1Æ1.0
5.7Æ0.5
5.4Æ0.1
90Æ5
7.5Æ0.3
7.3Æ0.3
>100
Figure 2. Structures of the lead compound 1 and the novel inhibitors
2±10.

M. Schlitzer et al. / Bioorg. Med. Chem. 8 (2000) 2399±2406
2401
Scheme 1. (I) (a) Arachidic acid or N-palmitoyl-N-methyl-b-alanine, i-BuOCOCl, NMM, DMF, 15 ꢀC, 5 min, (b) add 11, DMF to (a) 15 ꢀC!rt,
overnight; (II) (a) N-Boc-glycine or N-Boc-g-aminobutyric acid, i-BuOCOCl, NMM, DMF, 15 ꢀC, 5 min, (b) add 11, DMF to (a), 15 ꢀC!rt,
overnight; (III) 4 N HCl/dioxane, rt, 2 h; (IV) palmitoyl chloride, NMM, CH2Cl2, 0 ꢀC!rt, overnight; (V) succinic anhydride or glutaric anhydride,
toluene/dioxane, 80 ꢀC, 3 h; (VI) hexadecylamine, PyBOP, DIPEA, DMF, rt, overnight; (VII) 4 N HCl/dioxane, rt, 2 h; (VIII) stearoyl chloride,
NMM, CH2Cl2, 0 ꢀC!rt, overnight; (IX) 4 N HCl/dioxane, rt, 2 h; (X) (a) N-Boc-glycine or N-Boc-b-alanine, i-BuOCOCl, NMM, DMF, 15 ꢀC,
5 min, (b) add product from preceding step, NMM, DMF to (a), 15 ꢀC!rt, overnight; (XI) 4 N HCl/dioxane, rt, 2 h; (XII) palmitoyl chloride,
NMM, CH2Cl2, 0 ꢀC!rt, overnight.
by Pompliano.9 The assay employed yeast farnesyl-
transferase (FTase) fused to glutathione S-transferase at
the N-terminus of the b-subunit.10 Farnesylpyrophosphate
and the dansylated pentapeptide Ds-GlyCysValLeuSer
were used as substrates. Upon farnesylation of the
cysteine thiol the dansyl residue is placed in a lipophilic
environment which results in an enhancement of ¯uor-
escence at 505 nm which is used to monitor the enzyme
reaction.
For the initial design of 1 as well as for other bisubstrate
analogue inhibitors, we chose b-alanine as the linking
moiety because the distance between its carbonyl carbon
and its nitrogen resembles the distance between the car-
bonyl carbon and the thiol sulfur in cysteine.3,6 After
showing the importance of the amide group, we next
changed its position by replacing the b-alanyl moiety by
glycine (3) and g-aminobutyric acid (4), respectively,
thereby reducing or increasing the distance between the
amide group and the AAM-peptidomimetic substructure
by one methylene unit. Both variations led to a threefold
reduction in inhibitory activity.
Results and Discussion
In our inhibitor binding model (Fig. 1(b)), the amide
oxygen should exert a crucial interaction with the zinc
ion located in the active center of the FTase. Therefore,
a molecule in which this amide function is replaced by a
structure unable to act as a p-donor like an ethylene
group should show a considerable decrease in farnesyl-
transferase inhibitory activity. To study this hypothesis,
we prepared compound 2 (Fig. 2), which has the same
overall dimensions as our lead structure 1 but lacks the
amide function in question. As shown in Table 1, 2 is at
least two orders of magnitude less active against farnesyl-
transferase than 1. This strongly supports our picture of
how our molecules are bound to the enzyme's active center.
We next inverted the amide structure by replacing the
amino acid linkers by dicarboxylic acids succinic (5) and
glutaric acid (6), respectively. In both cases, this resulted in
an approximately 2-fold reduction in inhibitory activity,
compared to the lead compound 1, but an improvement in
activity if compared to compounds 3 and 4.
Finally, we introduced a methylene unit into the AAM-
peptidomimetic substructure by replacing the 4-amino-
benzoyl moiety by an 4-aminomethylbenzoyl structure.
The leftmost amide group in the N-stearoylamino-
methylbenzoyl derivative 7 is obviously still too close to
the center of the AAM-peptidomimetic substructure to

2402
M. Schlitzer et al. / Bioorg. Med. Chem. 8 (2000) 2399±2406
produce a relevant inhibitory activity against farnesyl-
transferase although the activity of this compound is
somewhat higher than that of 2. Reintroduction of gly-
cine and b-alanine as linkers (compounds 8 and 9)
improved the inhibitory activity signi®cantly; both
compounds being almost as potent inhibitors as our
lead structure 1. In addition to the secondary amides
investigated so far we have prepared the tertiary amide
derivative 10. In comparison to the lead structure 1,
compound 10 shows a distinct decrease in inhibitory
activity which may be due to a dislodgement of the car-
bonyl oxygen from the enzyme-bound zinc because of
steric reasons.
Recently, we reported6 the docking of compound 1 into
the active site using the coordinates of the published4
crystal structure of a ternary complex of farnesyl-
transferase, a farnesylpyrophosphate analogue and N-
Ac-Cys-Val-Ile-selenoMetOH. Flexible docking was
performed using the program FlexX.11 The C-terminal
methionine was used as starting fragment and placed at
the same position as the seleno-methionine of N-Ac-
Cys-Val-Ile-selenoMetOH in the crystal structure. Then,
the docking program placed the remainder of the mole-
cule fragment by fragment into the active site searching
for favorable interactions between the ligand and the
amino acid side chains while avoiding steric overlaps.
Figure 3. (a) Interactions between 1 and amino acid side chains in the active site; (b) Interactions between 6 and amino acid side chains in the active site.

M. Schlitzer et al. / Bioorg. Med. Chem. 8 (2000) 2399±2406
2403
Figure 3(a) shows interactions between 1 and amino acid
side chains in the active site. One key interaction calcu-
lated by FlexX is the interaction between the carbonyl
oxygen of the palmitoyl residue and the enzyme-bound
zinc, as we suggested above. In the present study we per-
formed additional docking runs with compounds 2 and 6.
Not surprisingly, no interaction between 2 and the enzyme-
bound zinc could be found by FlexX. This easily explains
the drop in activity of at least two orders of magnitude.
Interactions calculated between 6 and the amino acid side
chains of the farnesyltransferase are basically the same as
for 1 (Fig. 3(b)). Hydrophobic interactions are found
between the methionine side chain of 6 and Tyr 131a. The
C-terminal carbonyl group of 6 is hydrogen-bonded to
Gln 167a and to the guanidino group of Arg 202b,
which also forms a second hydrogen bond to the car-
bonyl of the central aminobenzoyl moiety. Another
hydrogen bond is formed between the hydroxy group of
Tyr 361b and the amide nitrogen in the 4-position of the
aminobenzoyl moiety. The docking program calculates
again an interaction between the terminal carbonyl group
of the glutaryl linker and the zinc. The central part of the
alkyl chain is placed close to the side chains of Lys 164a,
Lys 294b and Arg 291b, which form the positively charged
cleft binding the pyrophosphate portion of FPP.4 Inter-
actions are found between the terminal carbons of the
alkyl chain hydrophobic and Tyr 166a, Tyr 200a and
Tyr 251b, which form a part of the prenyl binding cleft.4
Micromass. IR spectra were recorded on a Nicolet 510P
FT IR-spectrometer. Microanalyses were obtained from
a CH analyzer according to Dr. Salzer from Labormatic
and from a Hewlett-Packard CHN-analyzer type 185.
Melting points were obtained with a Leitz microscope and
are uncorrected. Column chromatagraphy was carried
out using silica gel 60 (0.062±0.200 mm) from Merck.
General procedures for amide bond formation
Procedure A: mixed anhydride method. The appropriate
acid was dissolved in a sucient amount of dry DMF in
a ¯ame dried ¯ask under an atmosphere of Ar. After
addition of N-methylmorpholine (NMM) (0.25 mL per
mmol acid) the solution was cooled to 15 ꢀC and iso-
butyl chloroformate (0.13 mL per mmol acid) was added.
A solution of the amine component (1 equiv) in dry
DMF was added after 5 min. When the amine compo-
nent was employed as a hydrochloride, additional NMM
(0.25 mL per mmol) was added. The mixture was
allowed to warm up to room temperature overnight and
then poured into brine (400±800 mL). In case a solid
precipitate was formed, this was collected by suction
and thoroughly washed with water. Otherwise, the aqu-
eous mixture was extracted with ethyl acetate (3Â100
mL) and the combined organic extracts were washed
successively with 2 N citric acid, satd NaHCO3-soln and
brine and dried with MgSO4. The residue obtained after
removal of the solvent was puri®ed by recrystallization
or ¯ash chromatography.
In summary, structure±activity relationships and dock-
ing experiments provide strong evidence for the initial
hypothesis that the amide structure connecting the far-
nesyl mimetic and the linking substructures is crucial for
the inhibitory activity of our novel type of bisubstrate
analogue farnesyltransferase inhibitors. Variations of
the position and inversion of the amide group showed
distinct in¯uence on the inhibitory activity. We assign
the observed di€erence in inhibitory activity to the var-
iations of the amide group position, since in a previous
study on the structure±activity relationships of our
bisubstrate inhibitors we have shown that variations of
the chain length of the farnesyl mimetic to the extent
done here do not produce signi®cant e€ects on the
inhibitory activity.6 This study left our initially chosen
linker b-alanine as the best linking structure but shows
that in addition to that moiety succinic and glutaric acids
can be used for that position. A tolerable reduction in
activity is also observed upon the replacement of b-ala-
nine by glycine if an additional methylene unit is intro-
duced in the AAX-peptidomimetic substructure. As they
open a broader selection of building blocks our linking
substructure can be chosen from, the reported results
will be applied for our ongoing research on these novel
bisubstrate analogue farnesyltransferase inhibitors.
Procedure B: PyBOP method. One equiv of the appro-
priate acid, hexadecylamine and PyBOP (520 mg per
mmol acid), respectively, were dissolved in dry DMF in
a ¯ame dried ¯ask. After addition of diisopropylethyl-
amine (0.65 mL per mmol acid) the mixture was stirred
at room temperature overnight. Then, the mixture was
poured into brine (400±800 mL). The aqueous mixture
was extracted with ethyl acetate (3Â100 mL) and the
combined organic extracts were washed successively
with 2 N citric acid, satd NaHCO3-soln and brine and
dried with MgSO4. The residue obtained after removal
of the solvent was puri®ed by ¯ash chromatography.
Procedure C: acid chloride method. To a solution of the
appropriate amine hydrochloride in a sucient amount
of dry CH2Cl2 and N-methylmorpholine (0.24 mL per
mmol amine), 1 equiv of the appropriate acid chloride
was added at 0 ꢀC. Stirring was continued overnight.
The reaction mixture was diluted with CH2Cl2 and
washed successively with 2 N citric acid, satd NaHCO3-
soln and brine and dried with MgSO4. The residue
obtained after removal of the solvent was puri®ed by
¯ash chromatography.
General procedure for the cleavage of N-Boc protective
groups
Experimental
1H and 13C NMR spectra were recorded on a Jeol
JMN-GX-400 and a Jeol JMN-LA-500 spectrometer.
Mass spectra were obtained with a Vacuum Generators
VG 7070 H using a Vector 1 data acquisition system
from Teknivent or an AutoSpec mass spectrometer from
The N-Boc derivatives were dissolved in 4 N HCl in
dioxane (10 mL per mmol) and stirred for 2 h at room
temperature. After addition of diethyl ether, the volatiles
were removed in vacuo. The residue was used without
further puri®cation.

2404
M. Schlitzer et al. / Bioorg. Med. Chem. 8 (2000) 2399±2406
N-(4-Arachidoylaminobenzoyl)methionine methyl ester (2).
From arachidic acid (313 mg, 1 mmol) and 11 (282 mg,
1 mmol) using procedure A. Puri®cation: ¯ash chroma-
tography (ethyl acetate:hexane, 3:2). Yield: 385 mg
(67%); mp 52 ꢀC; IR (KBr) n 3440, 2920, 2850, 1740,
3.55 (s, 3H), 4.48 (m, 1H), 7.56±7.58 (m, 2H), 7.73±7.75
(m, 2H), 8.48 (m, 1H), 10.00 (s, 1H); 13C NMR (100
MHz, DMSO-d6) d 15.4, 16.1, 23.5, 26.6, 26.8, 30.2,
30.3, 30.4, 30.45, 30.5, 31.4, 31.7, 32.7, 35.4, 37.0, 53.1,
53.4, 119.6, 128.8, 129.6, 143.6, 167.6, 172.8, 173.6,
174.0. MS (EI) m/z 605 (20, M+), 545 (38), 531 (100),
443 (96). Anal. calcd for C33H55N3O5S: C, 65.45; H,
9.09; N, 6.94; found: C, 65.50; H, 8.67; N, 6.84.
1
1710, 1625 cm 1. H NMR (400 MHz, DMSO-d6) d
0.85 (m, 3H), 1.18±1.25 (m, 32H), 1.46 (m, 2H), 1.98±
2.06 (m, 5H), 2.54 (m, 2H), 3.64 (s, 3H), 4.52 (m, 1H),
5.61 (s, 1H), 6.56 (m, 2H), 7.61 (m, 2H), 8.20 (m, 1H);
13C NMR (100 MHz, DMSO-d6) d 15.5, 16.2, 22.6,
29.4, 29.5, 30.8, 30.9, 32.2, 37.0, 52.2, 52.3, 113.0, 120.1,
129.6, 153.3, 167.6, 168.0, 168.3. MS (EI): m/z 339 (9),
312 (18), 100 (27), 87 (100). Anal. calcd for C33H56
N2O4S: C, 68.71; H, 9.78; N, 4.86; S, 5.56; found: C,
68.83; H, 10.09; N, 4.70; S, 5.88.
3-{{N-{4-[N-(1-Methoxycarbonyl-3-methylthiopropyl)-
carbamoyl]phenyl}carbamoyl}}-propionic acid (14). Com-
pound 11 (564 mg, 2 mmol) was dissolved in a hot
mixture of toluene (50 mL) and dioxane (10 mL). Then,
succinic anhydride (200 mg, 2 mmol) dissolved in diox-
ane (10 mL) was added and the mixture was kept at 80 ꢀC
for 3 h. Most of the solvent was removed in vacuo and
pentane was added carefully to the remaining solution.
The solution was kept at room temperature until crys-
tallization occurred. Yield: 600 mg (78%). 1H NMR
(400 MHz, DMSO-d6) d 2.02±2.06 (m, 5H), 2.48±2.62
(m, 6H), 3.64 (s, 3H), 4.56 (m, 1H), 7.66 (m, 2H), 7.83
(m, 2H), 8.57 (d, J=7 Hz, 1H), 10.16 (s, 1H).
N-[4-(tert-Butyloxycarbonylaminoacetylamino)benzoyl]-
methionine methyl ester (12). From N-Boc-glycine (875
mg, 5 mmol) and 11 (1.41 g, 5 mmol) using procedure
1
A. Yield: 1.76 g (99%). H NMR (400 MHz, CDCl3) d
1.44 (s, 9H), 2.08 (s, 3H), 2.12±2.15 (m, 1H), 2.21±2.29 (m,
1H), 2.58 (m, 2H), 3.76 (s, 3H), 3.82±3.95 (m, 2H), 4.90
(m, 1H), 7.39 (m, 1H), 7.50 (m, 1H), 7.67±7.74 (m, 2H).
N-{4-[(3-Hexadecylaminocarbonylpropanoyl)amino]benz-
oyl}methionine methyl ester (5). From 14 (382 mg, 1
mmol) and hexadecylamine (241 mg, 1 mmol) using
procedure B. Puri®cation: ¯ash chromatography (1.
ethyl acetate:hexane, 3:2; 2. ethyl acetate). Yield 310 mg
(51%); mp 163 ꢀC. IR (KBr) n 3315, 2920, 2850, 1750,
1665, 1640, 1610 cm 1. 1H NMR (400 MHz, DMSO-d6)
d 0.77 (t, J=7 Hz, 1.15 (s, 26H), 1.28 (m, 2H), 1.96±1.99
(m, 5H), 2.31 (m, 2H), 2.46±2.49 (m, 4H), 2.94 (m, 2H),
3.57 (s, 3H), 4.48 (m, 1H), 7.58 (m, 2H), 7.69 (m, 1H),
7.75 (m, 2H), 8.49 (m, 1H), 10.05 (s, 1H); 13C NMR
(100 MHz, DMSO-d6) d 13.8, 14.5, 21.9, 26.3, 28.5,
28.6, 28.8, 28.9, 29.0, 30.1, 31.1, 31.7, 38.4, 51.5, 51.8,
117.9, 127.7, 128.2, 142.1, 166.0, 170.7, 170.8, 172.4. MS
(EI) m/z 605 (0.1, M+), 323 (35), 120 (64), 100 (100).
Anal. calcd for C33H55N3O5S: C, 65.45; H, 9.09; N,
6.94; found: C, 65.32; H, 9.20; N, 6.54.
N - [4 - (Hexadecanoylaminoacetylamino)benzoyl]methio-
nine methyl ester (3). Compound 12 (660 mg, 1.5
mmol) was deprotected as described and coupled with
palmitoyl chloride (0.46 mL, 1.5 mmol) using procedure
C. Puri®cation: ¯ash chromatography (ethyl acetate:
hexane, 3:2). Yield: 800 mg (92%); mp 97 ꢀC. IR (KBr) n
3400, 3315, 2920, 2850, 1745, 1695, 1640, 1605 cm 1. 1H
NMR (400 MHz, DMSO-d6) d 0.85 (t, J=7 Hz, 3H),
1.21±1.24 (m, 24H), 1.51 (m, 2H), 2.06±2.08 (m, 5H),
2.15 (t, J=7 Hz, 2H), 2.52±2.60 (m, 2H), 3.65 (s, 3H),
3.88 (d, J=5 Hz, 2H), 4.58 (m, 1H), 7.65±7.68 (m, 2H),
7.83±7.86 (m, 2H), 8.07 (m, 1H), 8.59 (d, J=7 Hz, 1H),
10.14 (s, 1H); 13C NMR (DMSO-d6, 100 MHz):
d=14.5, 15.2, 22.6, 25.7, 29.2, 29.4, 29.5, 29.6, 30.5,
30.8, 31.8, 35.7, 43.2, 52.2, 52.5, 118.8, 128.7, 129.0,
142.3, 166.7, 168.9, 173.0, 173.3. MS (EI): m/z 577 (6,
M+), 415 (40), 208 (25), 146 (28), 120 (100). Anal. calcd
for C31H51N3O5S: C, 64.44; H, 8.89; N, 7.27; S, 5.55;
found: C, 64.79; H, 8.91; N, 7.15; S, 5.83.
4-{{N-{4-[N-(1-Methoxycarbonyl-3-methylthiopropyl)-
carbamoyl]phenyl}carbamoyl}}butyric acid (15). Com-
pound 15 was prepared as described for 14 from 11 (564
mg, 2 mmol) and glutaric anhydride (228 mg, 2 mmol).
Yield 749 mg (95%). 1H NMR (400 MHz, DMSO-d6) d
1.84 (m, 2H), 2.02±2.06 (m, 5H), 2.28 (m, 2H), 2.38 (m,
2H), 2.52±2.62 (m, 2H), 3.65 (s, 3H), 4.58 (m, 1H), 7.67
(m, 2H), 7.83 (m, 2H), 8.57 (m, 1H), 10.10 (s, 1H).
N - {4 - [(4 - tert - Butyloxycarbonylamino)butanoylamino]-
benzoyl}methionine methyl ester (13). From N-Boc-g-
aminobutyric acid (508 mg, 2.5 mmol) and 11 (705 mg,
1
2.5 mmol) using procedure A. Yield: 900 mg (75%). H
NMR (400 MHz, CDCl3) d 1.43 (s, 9H), 1.80 (m, 2H),
2.08±2.12 (m, 5H), 2.38 (m, 2H), 2.56 (m, 2H), 3.14 (m,
2H), 3.75 (s, 3H), 4.87 (m, 1H), 6.61 (m, 1H), 7.60±7.76
(m, 3H).
N-{4-[(4-Hexadecylaminocarbonylbutanoyl)amino]benz-
oyl}methionine methyl ester (6). From 15 (396 mg, 1
mmol) and hexadecylamine (241 mg, 1 mmol) using pro-
cedure B. Puri®cation: ¯ash chromatography (1. ethyl
acetate:hexane, 3:2; 2. ethyl acetate). Yield 338 mg
(55%); mp 134 ꢀC. IR (KBr) n 3305, 2920, 2850, 1740,
1665, 1640, 1610 cm 1. 1H NMR (400 MHz, DMSO-d6)
d 0.76 (t, J=7 Hz, 3H), 1.12±1.18 (m, 26H), 1.28 (m, 2H),
1.72 (m, 2H), 1.96±1.98 (m, 5H), 2.03 (m, 2H), 2.24 (m,
2H), 2.46 (m, 2H), 2.93 (m, 2H), 3.56 (s, 3H), 4.48 (m, 1H),
7.58 (m, 2H), 7.60 (m, 1H), 7.74 (m, 2H), 8.48 (d, J=7 Hz,
1H), 9.99 (s, 1H); 13C NMR (100 MHz, DMSO-d6) d
13.8, 14.5, 21.0, 21.9, 26.3, 28.5, 28.6, 28.8, 28.9, 29.0,
N - {4 - [(4 - Hexadecanoylamino)butanoylamino]benzoyl}-
methionine methyl ester (4). Compound 13 (900 mg,
1.88 mmol) was deprotected as described and coupled
with palmitoyl chloride (0.58 mL, 1.88 mmol) using
procedure C. Puri®cation: ¯ash chromatography (ethyl
acetate:hexane, 3:2). Yield: 978 mg (86%); mp 81 ꢀC. IR
1
(KBr) n 3320, 2915, 2850, 1745, 1640 cm 1. H NMR
(400 MHz, DMSO-d6) d 0.76 (t, J=7 Hz, 3H), 1.09±
1.17 (m, 24H), 1.40 (m, 2H), 1.62 (m, 2H), 1.93±1.97 (m,
7H), 2.24 (t, J=7 Hz, 2H), 2.45 (m, 2H), 2.99 (m, 2H),

M. Schlitzer et al. / Bioorg. Med. Chem. 8 (2000) 2399±2406
2405
29.8, 30.1, 31.1, 34.6, 35.7, 38.3, 38.9, 51.5, 51.8, 118.0,
128.2, 142.1, 166.0, 171.1, 172.4. MS (EI) m/z 619 (4,
M+), 545 (13), 457 (19), 339 (22), 338 (100), 120 (65).
Anal. calcd for C34H57N3O5S: C, 65.91; H, 9.20; N,
6.78; found: C, 65.49; H, 9.32; N, 6.82.
MHz, CDCl3) d 1.39 (m, 9H), 2.07±2.13 (m, 5H), 2.44
(m, 2H), 2.54 (m, 2H), 3.38 (m, 2H), 3.76 (s, 3H), 4.44 (d,
J=6 Hz, 2H), 4.88 (m, 1H), 7.30 (m, 2H), 7.73 (m, 2H).
N - {4 - [3 - (Hexadecanoylamino)propanoylaminomethyl]-
benzoyl}methionine methyl ester (9). Compound 18 (500
mg, 1.07 mmol) was deprotected as described and cou-
pled with palmitoyl chloride (0.34 mL, 1.07 mmol) using
procedure C. Puri®cation: ¯ash chromatography (1.
ethyl acetate:hexane, 3:2; 2. ethyl acetate). Yield 356 mg
(55%); mp 147 ꢀC. IR (KBr) n 3300, 2920, 2850, 1745,
N-(4-tert-Butyloxycarbonylaminomethylbenzoyl)methio-
nine methyl ester (16). From 4-tert-butyloxycarbonyl-
aminobenzoic acid8 (2.2 g, 10 mmol) and methionine
methyl ester hydrochloride (2.0 g, 10 mmol) using pro-
cedure A. Yield 2.2 g (55%). 1H NMR (400 MHz,
CDCl3) d 1.45 (s, 9H), 2.07±2.13 (m, 1H), 2.09 (s, 3H),
2.26 (m, 1H), 2.55 (m, 2H), 3.77 (s, 3H), 4.33 (d, J=6
Hz, 2H), 4.91 (m, 1H), 6.92 (d, J=6 Hz, 1H), 7.32±7.34
(m, 2H), 7.74±7.76 (m, 2H).
1
1640 cm 1. H NMR (400 MHz, DMSO-d6) d 0.76 (t,
J=7 Hz, 3H), 1.37 (m, 24H), 1.91±1.98 (m, 5H), 2.23 (m,
2H), 2.41±2.55 (m, 4H), 3.16 (m, 2H), 3.56 (s, 3H), 4.22
(d, J=6 Hz, 2H), 4.51 (m, 1H), 7.24 (m, 2H), 7.69 (m,
1H), 7.74 (m, 2H); 13C NMR (100 MHz, DMSO-d6) d
13.8, 14.5, 22.0, 28.6, 28.7, 28.8, 28.9, 29.0, 29.9, 30.1,
31.2, 35.2, 35.3, 35.4, 41.7, 51.6, 51.8, 126.8, 127.4, 143.1,
166.4, 170.4, 172.0, 172.3. MS (EI) m/z 605 (1, M+), 531
(15), 324 (67). Anal. calcd for C33H55N3O5S: C, 65.45;
H, 9.09; N, 6.94; found: C, 65.52; H, 8.83; N, 7.11.
N-(4-Octadecanoylaminomethylbenzoyl)methionine methyl
ester (7). Compound 16 (594 mg, 1.5) was deprotected
as described and coupled with stearoyl chloride (450
mg, 1.5 mmol) using procedure C. Puri®cation: recrys-
tallization from ethyl acetate. Yield: 440 mg (52%); mp
67 ꢀC. IR (KBr) n 3305, 2920, 2850, 1740, 1700, 1640
1
cm 1. H NMR (400 MHz, DMSO-d6) d 0.85 (m, 3H),
N-{4-{3-[N-(Hexadecanoyl)-N-methylamino]propanoyla-
min}benzoyl}methionine methyl ester (10). Compound
10 was prepared from N-palmitoyl-N-methyl-b-alanine
(512 mg, 1.5 mmol) and 11 (423 mg, 1.5 mmol) using
procedure A. Puri®cation: ¯ash chromatography (ethyl
1.24 (m, 28H), 1.52 (m, 2H), 1.98±2.05 (m, 5H), 2.15 (m,
2H), 2.58 (m, 2H), 3.65 (s, 3H), 4.31 (m, 2H), 4.58 (m,
1H), 7.32 (m, 2H), 7.84 (m, 2H). MS (EI) m/z 562 (10,
M+), 488 (100), 400 (61). Anal. calcd for C32H54N2O4S:
C, 68.28; H, 9.67; N, 5.69; S, 5.69; found: C, 68.06; H,
10.47; N, 5.22; S, 5.38.
1
acetate). Yield 116 mg (13%); oil. H NMR (500 MHz,
CDCl3) d 0.81 (t, J=7 Hz, 3H), 1.17 (m, 24H), 1.56 (m,
2H), 2.04 (m, 5H), 2.28 (m, 2H), 2.52 (m, 2H), 2.67 (m,
2H), 3.00 (s, 3H), 3.69 (m, 3H), 3.72 (s, 3H), 4.85 (m,
1H), 6.85 (d, J=8 Hz, 1H), 7.66 (m, 2H), 7.70 (m, 2H),
9.49 (s, 1H). MS (EI) m/z 605 (5, M+), 525 (15, 215
(16). HRMS calcd for C33H55N3O5S: 605.3862; found:
605.3850.
N-[4-(tert-Butyloxycarbonylaminoacetylaminomethyl)ben-
zoyl]methionine methyl ester (17). Compound 16 (660
mg, 1.67 mmol) was deprotected as described and coupled
with N-Boc-glycine (292 mg, 1.67 mmol) using proce-
dure A. Yield: 600 mg (79%). 1H NMR (400 MHz,
CDCl3) d 1.42 (m, 9H), 2.02±2.10 (m, 5H), 2.57 (m, 2H),
3.68 (s, 3H), 3.82 (d, J=6 Hz, 2H), 4.47 (d, J=6 Hz,
2H), 4.90 (m, 1H), 7.30 (m, 2H), 7.74 (m, 2H).
Farnesyltransferase assay
Yeast farnesyltransferase (FTase) fused to glutathione
S-transferase at the N-terminus of the b-subunit was
expressed in Escherichia coli DH5a grown in LB media
containing ampicillin and additional chloramphenicol
for co-expression of pGEX-DPR1 and pBC-RAM2 for
FTase production.10 The enzyme was puri®ed by stan-
dard protocol using glutathione±agarose beads for
selective binding of the target proteins. FPP was
obtained as ammonium salt solution in methanol:10
mM aq NH4Cl (7:3) from Sigma-Aldrich. Dansyl-
GCVLS was custom synthesized by ZMBH, Heidelberg,
Germany. The assay mixture (100 mL volume) con-
tained 50 mM Tris/HCl, pH 7.4, 5 mM MgCl2, 10 mM
ZnCl2, 5 mM DTT, 7 mM Ds-GCVLS, 20 mM FPP and
approx. 5 nmol yeast GST±FTase and 1% of various
concentrations of the test compounds dissolved in
DMSO. The progress of the enzyme reaction was fol-
lowed by the enhancement of the ¯uorescence emission
at 505 nm (excitation: 340 nm). The reaction was started
by addition of the enzyme and run in a Quartz cuvette
thermostatted at 30 ꢀC. Fluorescence emission was
recorded with a Perkin±Elmer LS50B spectrometer.
IC50s were calculated from initial velocity of three inde-
pendent measurements of each inhibitor concentration
and expressed as mean Æ SD.
N - [4 - (Hexadecanoylaminoacetylaminomethyl)benzoyl]-
methionine methyl ester (8). Compound 17 (600 mg,
1.32 mmol) was deprotected as described and coupled
with palmitoyl chloride (0.4 mL, 1.32 mmol) using pro-
cedure C. Puri®cation: ¯ash chromatography (ethyl
acetate:hexane, 3:2). Yield 548 mg (70%); mp 94 ꢀC. IR
1
(KBr) n 3315, 2920, 2850, 1745, 1640 cm 1. H NMR
(400 MHz, DMSO-d6) d 0.76 (t, J=7 Hz, 3H), 1.13±
1.19 (m, 24H), 1.96±2.02 (m, 5H), 2.04 (m, 2H), 2.45±
2.53 (m, 2H), 3.56 (s, 3H), 3.63 (d, J=6 Hz, 2H), 4.24 (d,
J=6 Hz, 2H), 4.49 (m, 1H), 7.25 (m, 2H), 7.74 (m, 2H),
7.90 (m, 1H); 13C NMR (100 MHz, DMSO-d6) d 13.6,
14.2, 21.7, 24.7, 24.9, 28.3, 28.4, 28.5, 28.6. 28.65, 28.7,
29.6, 29.9, 30.9, 34.9, 41.4, 41.8, 51.3, 51.6, 126.5, 127.1,
131.8, 142.8, 166.1, 169.9, 172.1, 172.3. MS (EI) m/z
591, (10, M+), 517 (50), 296 (30), 118 (37). Anal. calcd
for C32H53N3O5S: C, 64.97; H, 8.96; N, 7.10; found: C,
64.53; H, 9.34; N, 7.52.
N-{4-[3-(tert-Butyloxycarbonylamino)propanoylamino-
methyl]benzoyl}methionine methyl ester (18). Com-
pound 16 (1.19 g, 3 mmol) was deprotected as described
and coupled with N-Boc-b-alanine (567 mg, 3 mmol)
1
using procedure A. Yield 1.05 g (75%). H NMR (400

2406
M. Schlitzer et al. / Bioorg. Med. Chem. 8 (2000) 2399±2406
Drugs 1996, 5, 1589. (f) Lerner, E. C.; Hamilton, A. D.; Sebti, S.
Acknowledgements
M. Anti-Cancer Drug Des. 1997, 12, 229. (g) Qian, Y.; Sebti, S. M.,
Hamilton, A. D. Biopolymers 1997, 43, 25. (h) Sattler, I.; Tama-
noi, F. In Regulation of the RAS Signaling Network; Maruta, H.,
Burgess, A. W., Eds.; Springer: Heidelberg, 1996; pp 95±138.
3. Schlitzer, M.; Sattler, I. Angew. Chem., Int. Ed. 1999, 38,
2032.
4. Strickland, C. L.; Windsor, W. T.; Syto, R.; Wang, L.;
Bond, R.; Wu, R.; Schwartz, J.; Le, H. V.; Beese, L. S.; Weber,
P. C. Biochemistry 1998, 37, 16601.
The pGEX-DPR1 and pBC-RAM2 plasmids were
kindly provided by Prof. F. Tamanoi (UCLA). M.
Schlitzer thanks Ms. K. Burk, Ms. S. Eilers and Ms. K.
Sellinger for technical assistance. I. Sattler wishes to
thank Prof. Dr. S. Grabley and PD Dr. R. Thiericke for
generous support and Ms. S. Egner for excellent tech-
nical assistance.
5. Dunten, P.; Kammlott, U.; Crowther, R.; Weber, D.;
Palermo, R.; Birktoft, J. Biochemistry 1998, 37, 7907.
6. Schlitzer, M.; Bohm, M.; Dahse, H.-M.; Sattler, I. Bioorg.
Med. Chem., in press.
References
1. (a) Casey, P. J.; Seabra, M. C. J. Biol. Chem. 1996, 271,
5289. (b) Clarke, S. Annu. Rev. Biochem. 1992, 61, 355. (c)
Cox, A. D.; Der, C. J. Biochim. Biophys. Acta 1997, 1333, F51.
(d) Gibbs, J. B.; Oli€, A. Annu. Rev. Pharmacol. Toxicol. 1997,
37, 143.
2. (a) Leonard, D. M. J. Med. Chem. 1997, 40, 2971. (b)
Graham, S. L. Exp. Opin. Ther. Pat. 1995, 5, 1269. (c) Ayral-
Kaloustian, S.; Skotnicki, J. S. Annu. Rep. Med. Chem. 1996,
171. (d) Sebti, S. M.; Hamilton, A. D. Drug Discovery Today
1998, 3, 26. (e) Singh, S. B.; Lingham, R. B. Exp. Opin. Invest.
7. Zhao, M. J.; Robert, D.; Jung, L. Eur. J. Med. Chem.
Chim. Ther. 1993, 28, 949.
8. Svahn, C. M.; Merenyi, F.; Karlson, L.; Widlund, L.;
Gralls, M. J. Med. Chem. 1986, 29, 448.
9. Pompliano, D. L.; Gomez, R. P.; Anthony, N. J. J. Am.
Chem. Soc. 1992, 114, 7945.
10. Del Villar, K.; Mitsuzawa, H.; Yang, W.; Sattler, I.;
Tamanoi, F. J. Biol. Chem. 1997, 272, 680.
11. Rarey, M.; Kramer, B.; Lengauer, T.; Klebe, G. J. Mol.
Biol. 1996, 261, 470.