New derivatives of farnesylthiosalicylic acid (Salirasib) for cancer treatment: Farnesylthiosalicylamide inhibits tumor growth in nude mice models

Article abstract of DOI:10.1021/jm801165r

The Ras inhibitor S-trans,trans-farnesylthiosalicylic acid (FTS, Salirasib) interferes with Ras membrane interactions that are crucial for Ras-dependent transformation. It remains unknown whether modifications of the carboxyl group of FTS can affect its a

Full text of DOI:10.1021/jm801165r

J. Med. Chem. 2009, 52, 197–205
197
New Derivatives of Farnesylthiosalicylic Acid (Salirasib) for Cancer Treatment:
Farnesylthiosalicylamide Inhibits Tumor Growth in Nude Mice Models
Liat Goldberg,† Roni Haklai,† Victor Bauer,‡ Aaron Heiss,§ and Yoel Kloog*,†
Department of Neurobiology, The George S. Wise Faculty of Life Sciences, Tel-AViV UniVersity, Tel AViV, Israel, Concordia Pharmaceuticals,
Inc., 1550 Sawgrass Corporate Parkway, Sunrise, Florida, Ricerca Biosciences, 7528 Auburn Road, Concord, Ohio
ReceiVed April 23, 2008
The Ras inhibitor S-trans,trans-farnesylthiosalicylic acid (FTS, Salirasib) interferes with Ras membrane
interactions that are crucial for Ras-dependent transformation. It remains unknown whether modifications
of the carboxyl group of FTS can affect its activity. Here we show that specific modifications of the FTS
carboxyl group by esterification or amidation yield compounds with improved growth inhibitory activity,
compared to FTS, as shown in Panc-1 and U87 cells. The most potent compounds were FTS-methoxymethyl
ester and FTS-amide. However, selectivity toward active Ras-GTP, as known for FTS, was apparent with
the amide derivatives of FTS. FTS-amide exhibited the overall highest efficacy in inhibition of Ras-GTP
and cell growth. This new compound significantly inhibited growth of both Panc-1 tumors and U87 brain
tumors. Thus amide derivatives of the FTS carboxyl group provide potent cell-growth inhibitors without
loss of selectivity toward the active Ras protein and may serve as new candidates in cancer therapy.
Introduction
A strong affirmation to this concept came with the design of
the specific Ras inhibitor, S-trans-trans-farnesylthiosalicylic acid
(Salirasib). FTS was designed to mimic the farnesyl moiety in
the carboxy terminal of Ras23 and was found to induce
accelerated dislodgement of Ras from the cell membrane and
subsequent degradation of the protein.24 FTS was also found
to be highly efficient in inhibiting growth of many types of
cancer cell lines both in vitro and in vivo as well as inhibiting
migration of several glioblastoma cell lines.10,25-29 Biochemical
and gene expression profiling experiments provided strong
support to the notion that the growth inhibitory effects of FTS
are the result of inhibition of Ras-dependent signaling involved
in tumor progression and maintenance.26,30,31
Ras proteins act as key regulators of cell signaling pathways.
TheiractivationisdependentuponbindingtoaGTPa nucleotide1-3
and association to the plasma membrane.4-9 Active Ras proteins
trigger an array of cellular messengers transducing key signals
regulating proliferation,2 migration,10-12 and cell death.13
Termination of these signals involves hydrolysis of the Ras
bound GTP to GDP.14 Several mutated forms of Ras are
defective in their GTP hydrolysis liability and are therefore
constitutively active.1,3 These oncogenic Ras proteins, which
are found in many cancer types, contribute to malignancy and
are therefore considered favored targets for directed therapy.15
Association of Ras to the plasma membrane has been shown
to be crucial for its activity in both the wild type and the mutated
constitutively active forms.2,16,17 At least two structural elements
are required for this association; the first is a farnesylcysteine
carboxy methyl ester at the carboxy terminal of Ras, and the
other element resides at the adjacent upstream sequence and
varies among different Ras isoforms.5,6 Normal Ras activity
requires specifically the farnesyl isoprenoid moiety,16,18 which
acts as a specific recognition unit to allow binding of H-Ras
with galectin-119,20 and K-Ras with galectin-3,21 promoting
strong membrane association and robust signaling. These
galectins possess a hydrophobic binding pocket, the putative
farnesyl binding site.20 Galectin-1 was shown to drive H-Ras-
GTP nanocluster formation and to be an integral component of
the H-Ras-GTP nanocluster, a site at which Raf is activated.22
Together, these findings mark the farnesyl-binding pockets as
outstanding targets for Ras directed therapy, assuming that
compounds that would block such sites will act as Ras inhibitors.
In previous work, efforts were made to find new FTS
derivatives that might improve the anticancer effect of the drug.
These compounds were all based on the FTS backbone and were
modified on different positions such as the benzene ring or
different lengths of the isoprenoid chain. These studies estab-
lished that the minimal length of the isoprenoid group required
for the inhibition of Ras is C15 (farnesyl) because the C10
isoprenoid derivative geranylthiosalicylic acid was inactive.23,32
Although some of the previously described compounds, includ-
ing 5-F-FTS and 5-Cl-FTS, were found to be effective inhibitors
of cell growth in Ras transformed cells, none proved to be
superior to FTS.32 It remains unknown whether modifications
of the carboxyl group of FTS could have an impact on its anti
Ras activity. This prompted us to search for new compounds
that may improve FTS’s potency in various cancer types and
perhaps shed some more light on Ras-membrane interactions.
In the present study, we investigated the efficacy of several new
FTS derivatives with modifications to their carboxyl group.
During the design of these compounds, an important consider-
ation taken into account was the degree of applicability in a
brain tumor model. In a previous work, FTS was shown to
penetrate the BBB in very low concentrations, possibly due to
its free carboxyl end.33 Therefore, the new compounds described
here were designed to allow possible better penetration by means
of various modifications that increase the lipophilicity of the
molecule. The potency of the new derivatives was examined in
* To whom correspondence should be addressed. Phone: 972-3-640-9699.
Fax: 972-3-640-7643. E-mail: kloog@post.tau.ac.il.
†
Department of Neurobiology, The George S. Wise Faculty of Life
Sciences.
‡
Concordia Pharmaceuticals, Inc.
Ricerca Biosciences.
§
a Abbreviations: FTS, farnesyl thiosalicylic acid; FTS-A, FTS-amide;
FTS-MA, FTS-methyl amide; FTS-DMA, FTS-dimethyl amide; FTS-ME,
FTS-methyl ester; FTS-MOME, FTS-methoxymethyl ester; FTS-IE, FTS-
isopropyl ester; FTS-BE, FTS-benzyl ester.
10.1021/jm801165r CCC: $40.75
2009 American Chemical Society
Published on Web 12/15/2008

198 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1
Goldberg et al.
Moreover, some of the modifications, namely addition conver-
sion of FTS to its methyl ester, methoxymethyl ester, amide,
methyl amide, and dimethyl amide derivatives, showed im-
proved efficacy as compared to FTS in inhibiting tumor cell
growth. On the other hand, the benzyl ester or isopropyl ester
showed a decrease in growth inhibition as compared to FTS
itself.
Effect of FTS Derivatives on Ras and Its Downstream
Effectors ERK and Akt. We next wanted to examine whether
the growth inhibitory effects exerted by the FTS derivatives
were associated with down regulation of Ras signaling, as in
the case of FTS.25,26,28,35 We thus used the Ras-GTP pull down
assay (see Experimental procedures) and examined the effects
of the more potent compounds on the levels of all three Ras
isoforms. Typical immunoblots demonstrating the effect of
FTS-A on the level of all active Ras isoforms (total Ras-GTP)
and on the levels of the individual active K-Ras-GTP, H-
RasGTP, and N-Ras-GTP in Panc-1 and U87 cells are shown
in Figure 3. As shown, FTS-A caused a clear decrease in the
levels of total Ras-GTP as well as in K-Ras-GTP in both cell
lines. Reduction in the level of N-Ras-GTP was observed only
in U87 cells, and no effect on H-Ras-GTP was observed in either
cell line (Figure 3). Table 2 summarizes the results obtained
with FTS-ME, FTS-MOME, FTS-A, FTS-MA, and FTS-DMA.
As shown, FTS-A, FTS-MA, and FTS-DMA caused a clear
decrease in the levels of K-Ras-GTP in Panc-1 cells (respectively
56.3 ( 11.2%, 52.3 ( 12.6%, and 47.8 ( 12.4% inhibition)
and in U87 cells (respectively 50.8 ( 9.2%, 32.6 ( 9.4%, and
28 ( 16.5% inhibition). Reduction in the level of N-Ras-GTP
was observed only in U87 cells; the degree of inhibition by
FTS-A, FTS-MA, and FTS-DMA was respectively 37.3 ( 8.9%,
22.5 ( 6.5%, and 41 ( 4.6%. No effect on H-Ras-GTP was
observed in either cell line (Table 2). However, FTS-ME and
FTS-MOME, the more potent growth inhibitors among the ester
derivatives, unexpectedly had a lesser effect on Ras-GTP in
either cell line. Together these results suggested that only the
FTS-amide derivatives mimicked the effect of the parental
compound FTS, while the growth inhibition activity of the FTS
esters may include a mechanism beyond Ras inhibition. That
is to say, the amide derivatives of FTS exhibited both growth
inhibitory activity and affected the target active Ras protein;
the oncogenic K-Ras in Panc-1 cells and the chronically active
N-Ras and K-Ras in U87 cells. Consistent with this notion, we
found that in U87 cells FTS-A caused a significant decrease in
the level of phospho-ERK and phospho-AKT, two classical
downstream effectors of Ras (Figure 3). The impact of FTS-A
on these signaling molecules in Panc-1 cells was less pro-
nounced, presumably because under the conditions used these
pathways were not active.
Figure 1. Structure of new FTS derivatives. Synthesis and purity of
the compounds are detailed under Experimental procedures.
terms of inhibition of in vitro cell growth and Ras activity in
human cancer cell lines. The most potent compound FTS-amide
was then tested in vivo using nude mice tumor models.
Results
Synthesis and Chemical Characterization of New FTS
Derivatives. The following derivatives were prepared: FTS-
methylester (FTS-ME), FTS-methoxymethylester (FTS-MO-
ME), FTS-isopropylester (FTS-IE), FTS-bezylester (FTS-BE),
FTS-amide (FTS-A), FTS-methylamide (FTS-MA), and FTS-
dimethylamide (FTS-DMA). The synthesis descriptions and
purity of the compounds are detailed in Experimental proce-
dures. Structures of the compounds are shown in Figure 1.
Effects of FTS Derivatives on Growth of Human
Cancer Cell Lines. Our first goal was to establish efficacy of
the new compounds in cell growth assays. We performed in
vitro growth inhibition assays using a cancer cell line that
harbors an oncogenic K-Ras, the pancreatic cell line Panc-1,
and a cell line in which Ras is not mutated but is chronically
active due to high activity of growth factor receptors, the
glioblastoma cell line U87. The cells were incubated with and
without each drug for 5 days, and cell number was then
determined by direct counting (see Experimental Section). First
we examined the effects of the FTS-ester derivatives on Panc-1
(Figure 2A) and on U87 cells (Figure 2C). All of the esters
caused a dose-dependent decrease in cell number at the
concentration range of 6-100 µM. These effects were mainly
due to inhibition of cell growth except for the higher dose where
cell death was observed as well. FTS-MOME was the most
potent growth inhibitor with an IC50 of 15 µM in both cell lines.
FTS-ME was somewhat less effective in Panc-1 cells (IC50
)
30 µM) but exhibited the same growth inhibition potency as
FTS-MOME in U87 cells (Figure 2C, Table 1). FTS-IE and
FTS-BE were clearly less potent than the two other esters with
an IC50 > 50 µM in both cell lines (Figure 2A,C, Table 1).
Next, we examined the effects of the FTS-amides on panc-1
and U87 cells (Figure 2C,D). All three FTS amide derivatives
caused a dose-dependent decrease in cell number of both cell
lines at the concentration range of 6-100 µM exhibiting similar
potencies (Figure 2B,D). Cell death was observed at concentra-
tions higher than 50 µM. The IC50s recorded in both cell lines
were at the range of 10-20 µM (Table 1). Interestingly, five of
the seven new compounds were more effective growth inhibitors
of Panc-1 and U87 cells as compared to FTS whose reported
IC50 values 35 µM in Panc-134 and 50 µM in U87 cells.26
Altogether, these results demonstrate that certain modifica-
tions on the carboxyl group of FTS yield active compounds
capable, to various degrees, of inhibiting cancer cell growth.
We have considered the possibility that ester or amide
hydrolysis could have occurred under the culture conditions
used, leading to the formation of FTS itself. Our results do not
support this possibility. First, both FTS-A and FTS-ME
exhibited a significantly higher growth inhibitory activity as
compared with FTS (Table 1); if they were hydrolyzed, then
these compounds would have exhibited potency similar to FTS.
A significantly higher uptake of the FTS derivatives is also
unlikely in light of previous experiments that demonstrated that
cells take up farnesylated peptides with very high efficiency.36
Second, FTS-ME exhibited different pharmacology than FTS
(Table 2); if it was hydrolyzed to FTS, we would have detected
a stronger effect of FTS-ME on Ras-GTP as compared with
FTS, but this was not the case (Table 2). Third, our next set of

FTS DeriVatiVes
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1 199
Figure 2. Inhibition of cell growth by FTS derivatives. Panc-1 and U87 cells were seeded in 24 wells plates at a density of 5000 cells/well. After
24 h, the media were replaced with media containing either the detailed FTS derivatives (6.25, 12.5, 25, 50, and 100 µM) or the control vehicle
(0.1% DMSO). Five days later, the cells were collected and counted. Growth inhibition curves are presented for: (A) Panc-1 cells treated with
FTS-ME, FTS-MOME, FTS-IE, and FTS-BE; (B) Panc-1 cells treated with FTS-A, FTS-MA, and FTS-DMA; (C) U87 cells treated with FTS-ME,
FTS-MOME, FTS-IE, and FTS-BE; (D) U87 cells treated with FTS-A, FTS-MA, and FTS-DMA. The values shown are from 2 independent
experiments (means ( SD, n ) 8).
Table 1. IC50 Values of FTS Derivatives in Panc-1 and U87 Cellsa
IC50 (µM)
FTS derivatives
Panc-1
U87
FTSb
FTS-ME
FTS-MOME
FTS-IE
FTS-BE
FTS-A
FTS-MA
FTS-DMA
35
30
12
70
95
20
10
22
50
10
10
52
30
10
18
20
a The IC50 values of the detailed FTS derivatives were calculated from
the growth inhibitory curves shown in Figure 2. b The IC50 of FTS was
previously calculated for Panc-1 cells34 and for U87 cells.26
pharmacokinetic experiments as described bellow confirmed that
FTS-A is indeed stable.
Altogether, these experiments suggested that FTS-amides act
like FTS as Ras inhibitors. FTS-A seemed to exhibit a relatively
higher activity as compared with FTS (about 3-5 times more
potent) and thus was employed in the subsequent animal
experiments.
Figure 3. Effect of FTS-A on Ras signaling. Panc-1 and U87 cells
were seeded at a density of 1.5 × 106 cells per 10-cm plate. After
24 h, the media were replaced with media containing either FTS-A
(50 µM) or the control vehicle (0.1% DMSO) for 24 h. The cells were
then lysed, and aliquots of the lysates were immunoblotted with anti
pan-Ras, anti K-Ras, anti H-Ras, anti N-Ras, anti phospho-Akt, anti
phospho-ERK, and anti ꢀ-tubulin antibodies. Total Ras-GTP, K-Ras-
GTP, H-Ras-GTP, and N-Ras-GTP in the aliquots were determined
by pull-down assays as described in Experimental Section. Typical
immunoblots are shown for Panc-1 cells (left panel) and U87 cells (right
panel).
Pharmacokinetics of FTS-Amide. First, we evaluated the
pharmacokinetics of FTS-A after oral gavage administration
(100 mg/kg) using rats as a model (see Experimental Section).
Following dosing, blood samples were collected at the indicated
times (10-330 min). Samples were then processed and analyzed
by LC-MS/MS as described in the Experimental Section. A
calibration curve for FTS-A was prepared ranging from 1 to
2500 ng/mL and was linear over this range. FTS-A was readily
detected in rat plasma by LC-MS/MS at a detection limit of 1
ng/mL. Compared with the expected retention time of the parent
ion (FTS-A, 1.01-1.02 min), the retention times of the tested
sample ions was 1.02 min, corroborating the assay method and
confirming that the compound remains stable in the plasma
(Figure 4A). Plasma concentration curves for FTS-A at the
various time points (Figure 4B) were used to calculate the
pharmacokinetic variables by a noncompartmental model as
detailed in the Experimental Section. The results thus obtained,
namely the terminal-phase half-time (elimination half-time, t1/
2), the time at which the maximum plasma concentration (Cmax
)
was reached (Tmax), the Cmax, and the area under the plasma
concentration-time curve (AUC), are summarized in Table 3

200 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1
Table 2. Inhibition of Active Ras by FTS Derivativesa
Goldberg et al.
% inhibition of active Ras levels
Panc-1
U87
FTS
derivatives
total
Ras-GTP
KRas-
GTP
HRas-
GTP
total
Ras-GTP
KRas-
GTP
NRas-
GTP
HRas-
GTP
NRas-GTP
FTS-ME
FTS-MOME
FTS-A
FTS-MA
FTS-DMA
5.3 ( 4.1
11.9 ( 13.8
19.9 ( 41.2
56 ( 11.2c
52 ( 12.6c
48 ( 12.4c
23.8 ( 18.6
4.8 ( 14.8
50.8 ( 9.2c
32.6 ( 9.4b
28 ( 16.5b
18.7 ( 2.8
5.1 ( 23.6
37.3 ( 8.9b
22.5 ( 6.5b
41 ( 4.6c
2.6 ( 3.8
37.4 ( 8.6b
39.7 ( 6.1c
38.6 ( 6.3c
8.6 ( 33.1
18.8 ( 19.3
13.6 ( 7.1
33 ( 12.7b
34.7 ( 7.8b
45.4 ( 8.8c
14 ( 12.7
a Panc-1 and U87 cells were seeded at a density of 1.5 × 106 cells per 10-cm plate. After 24 h, the media were replaced with media containing either 50
µM of the FTS derivatives or the control vehicle (0.1% DMSO) for 24 h. The cells were then lysed and aliquots of the lysates were used to determine total
Ras-GTP, K-Ras-GTP, H-Ras-GTP, and N-Ras-GTP levels by pull-down assays as described in Experimental Section. Protein bands were visualized by
enhanced chemiluminescence and quantified by densitometry using EZQuant-Gel computer software. The percentage of active Ras inhibition was calculated
as the subtraction of the percentage of active Ras levels after drug treatment (compared to control levels) from the percentage of active Ras levels in the
control. Values are presented for Panc-1 cells (left panel) and U87 cells (right panel) as means ( SD. b P < 0.05. c P < 0.01.
FTS-Amide Inhibits Growth of Human Tumors in
Nude Mouse Models. Next we examined the effect of FTS-A
on brain tumor growth using a nude mouse model with human
glioblastoma U87 cells implanted intracranially into the striatum
area. Four days post-implantation, the mice were scanned by
MRI to confirm tumor formation and were divided into two
groups of seven control and seven drug-treated animals.
Treatment started one day after the first MR imaging and
subsequent MR imaging was performed on days 4, 9, and 13
following start of treatment. FTS-A (100 mg/kg) was admin-
istered orally twice daily (see Experimental Section). T1- and
T2-weighted MR images were acquired after Gd-DTPA injection
using the 7-T BioSpec magnet. The T1-weighted images were
used to measure tumor volume. The T2-weighted images were
used to obtain information about inflammation and ventricle
enlargement. Figure 5A shows T1-weighted MR images of
representative control and FTS-A treated mice recorded before
treatment (baseline images) and on day 4 (1st followup) and
day 9 (2nd followup) after the treatment began. The T1-weighted
images of the control mouse demonstrate the rapid increase in
tumor volume over time. The increase in tumor volume recorded
over time in the FTS-A mouse was significantly lower than that
recorded in the control mouse (Figure 5A). The data clearly
imply that more contrast agent molecules accumulated in the
control mice as compared to the FTS-A treated mice. Impor-
Figure 4. Plasma analysis for orally administered FTS-A. Animals
received orally 100 mg/kg FTS-A in corn oil, and plasma samples were
collected and analyzed by LC-MS/MS as detailed in the Experimental
Section. (A) Typical chromatogram for FTS-A plasma analysis of a
50 min sample. (B) Plasma concentration curve for FTS-A was prepared
using a calibration curve. Mean values ((SE, n ) 4) determined at
the indicated times are shown.
tantly, we did not observe any inflammatory response in the
brains of the controls or in the brains of the drug-treated mice
(not shown).
The T1-weighted images were used to evaluate tumor volumes
in the followups of all control and all drug-treated mice
employing the MIA MATLAB program (see Experimental
Section). The data were expressed in terms of “normalized tumor
growth”, namely the subtraction of the initial tumor volume from
the tumor volume on the indicated followup divided by the initial
tumor volume. The results of the analysis for followups 1 and
2 (days 4 and 9 of drug treatment) are presented in Figure 5B
for the controls and FTS-A-treated mice. The mean value of
the “normalized tumor growth” in the FTS-A treated mice (n
) 7) was significantly smaller than that of the controls (n ) 7)
in both followups 1 and 2 (Mann-Whitney, * P < 0.05, ** P
< 0.01) indicating respectively 80% and 50% inhibition (Figure
5B).
Table 3. Pharmacokinetic Parameters for FTS-A and FTSa
parameter
dose (mg/kg)
half-life (min)
Tmax
FTS-A
FTSb
40
141.6
60
100
314
110
Cmax (ng/mL)
AUC0-inf (min·ng/mL)
371
108783
2159.4
372432
a The pharmacokinetic parameters for FTS-A were calculated using the
mean time/concentration data at 11 time points (10-330 min). b The
pharmacokinetic parameters for FTS are from ref 27.
and compared to the previously obtained pharmacokinetic
variables of FTS.27 As shown in Table 3, although the t1/2 value
for FTS-A is more than 2-fold higher than that of FTS, its
availability in the plasma is lower than that of FTS; the Tmax
value for FTS-A is almost 2-fold higher than that of FTS and
the Cmax and AUC values are significantly lower even though
FTS-A was given at a higher dose than FTS.
We also examined the impact of FTS-A on Panc-1 tumor
growth in nude mouse models as described earlier.27,34 Cells
were implanted subcutaneously just above the right femoral joint
of nude mice and oral treatment with FTS-A (100 mg/kg po,
daily) began when palpable tumors were observed. Tumors were
removed 18 days later and weighed. As shown in Figure 5C,

FTS DeriVatiVes
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1 201
Figure 5. FTS-A inhibits growth of U87 and Panc-1 tumors in nude mice. U87 cells (5 × 105) were injected into the striata of nude mice (1
µL/min, 5 min total) as detailed in the Experimental Section. Four days later, the mice were MR imaged to verify tumor development and the mice
were divided into two groups exhibiting similar tumor sizes. Starting from day 5, the mice were treated either with po dosing of 50 µL FTS-A (100
mg/kg) twice a day or with po doses of 50 µL vehicle solutions twice a day. Followup MR images were taken on days 9, 14, and 18. T1-weighted
images from the baseline imaging (day 4), first followup (day 9), and second followup (day 14) are shown in (A) for a typical control mouse (left
panel) and a typical treated mouse (right panel). (B) Tumor volumes were measured from T1-weighted MR images and the normalized ratio of
tumor growth between followups was calculated as the subtraction of the initial tumor volume from the tumor volume in followup(n) divided by
the initial tumor volume (mean ( SD, n ) 7, * P < 0.05, ** P < 0.01). (C) Panc-1 cells (5 × 106) were injected subcutaneously just above the
right femoral joint. Ten days later, tumors were measured and mice were divided into a treated group receiving daily po doses of 100 µL FTS-A
(100 mg/kg) and a control group receiving daily po doses of 100 µL vehicle solution. Eighteen days later, the tumors were extracted and weighed
as shown (mean ( SD, n ) 8, * P < 0.05).
the average tumor weight in the treated group was more than
2-fold smaller than that of the control group (unpaired t test
with Welch correction, * P < 0.05). In conclusion, it appears
that oral FTS-A significantly attenuated both Panc-1 and U87
tumor growth in the nude mouse without causing any toxic side
effects, making it an interesting new Ras inhibitor for potential
treatment of Ras-dependent tumors.
The lowered effect of FTS-ME on Ras is even more surprising
in light of the knowledge that all Ras proteins possess a carboxy-
terminal farnesylcysteine carboxy methyl ester.16,18 Interestingly,
like FTS, N-acetyl farnesylcystine (AFC), which mimics the
carboxy terminal of Ras proteins, is also known to inhibit ERK
activation.39 However, AFC and FTS appear to differ in their
modes of action on Ras proteins and on Icmt, the enzyme that
methylates Ras proteins. AFC acts mostly as an inhibitor of
the Icmt, serving actually as a competitive substrate of the
enzyme.40 FTS is a pure inhibitor of the Icmt enzyme and does
not serve as a substrate for the enzyme.41 At its growth inhibitory
concentrations, (10-50 µM) FTS dislodges the active Ras from
the cell membrane without an effect on methylation.24,41 As
has been discussed previously, AFC inhibits Ras methylation,
thereby blocking Ras trafficking to the cell membrane,39 while
FTS dislodges the mature Ras from the cell membrane.24
Nevertheless, when AFC, but not FTS, enters the cell, it
inevitably becomes methylated, and thus AFC-methyl ester
accumulates at the expense of AFC. When FTS enters the cell,
it retains its free carboxyl group. Because we show here that
the synthetic FTS-methyl ester has a diminished effect on Ras,
we think that, similarly, AFC-methyl ester that accumulates in
cells may not dislodge Ras from the cell membrane because it
does not bind to Ras scaffolds such as galectin-1 and galectin-
3. Accordingly, the more hydrophobic nature of FTS-ME and
of AFC-methyl ester as compared to their free carboxylic acid
derivatives might inhibit the interactions with Ras scaffolds.
This, however, is not the case of the FTS-amides, which may
accommodate better in the putative prenyl binding pockets of
Gal-120 and Gal-3.42
Discussion
We have prepared and characterized new FTS derivatives in
which the carboxyl group has been modified by esterification
or amidation. We found that all of the FTS amides and two of
the FTS esters (FTS-ME and FTS-MOME) strongly inhibited
growth of Panc-1 and of U87 cells in which at least one Ras
isoform is chronically active.37,38 The esters FTS-IE and FTS-
BE were far less active than FTS or the other compounds
examined. This indicates that the bulky group (isopropyl and
benzyl) of these esters may interfere with growth inhibitory
activity of these FTS derivatives. FTS itself acts as a Ras
inhibitor that inhibits Ras membrane anchorage by interfering
with galectin-1- and galectin-3-induced nanoclustering of,
respectively, the oncogenic H-Ras and K-Ras isoforms in the
cell membrane.21,22 This interference blocks the robust signaling
of the Ras nanocluster. This seemed to be also the case of the
new FTS-amides, which reduced the levels of active Ras in
Panc-1 and in U87 cells. However, FTS esters, even those that
exhibited relatively strong growth inhibitory activity, had a much
reduced effect on Ras (Table 2). Therefore we suspect that the
growth inhibitory action of the FTS-ester derivatives (FTS-ME
and FTS-MOME) is a result of interference with Ras-
independent cell growth pathways.

202 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1
Goldberg et al.
The above observations strongly suggest that carboxyl
methylated prenyl analogues, including FTS-ME or FTS-
MOME, inhibit cell growth by interfering with prenyl binding
proteins other than Ras escorts. It is tempting to speculate that
FTS-ME or FTS-MOME would interfere with the action of the
well-described prenyl binding proteins RhoGDIs that bind the
geranylgeranyl isoprenoid moieties of Rac/Rho GTPases, known
to be involved in cell growth and cell transformation.43,44
Consistent with this suggestion are the results of experiments
that determined the binding of isoprenylated acids to RhoGDI.45
These experiments showed that isoprenylated carboxy methyl
esters exhibit strong interactions with RhoGDI while their
corresponding free carboxylic acid derivatives exhibit weak
interactions with RhoGDI.45 For example, AFC-methyl ester
exhibited high affinity to RhoGDI while AFC and FTS barely
interacted with RhoGDI.45 Interestingly, the isoprenoid moiety
of Rac/Rho GTPases accommodates precisely within a hydro-
phobic pocket in RhoGDI43,44 and, indeed, N-acetyl gera-
nylgeranyl cysteine binds strongly to RhoGDI.45 Thus it seems
that strong association of prenylated compounds with RhoGDI
depends on both the length of the isoprenoid group and the
carboxymethyl ester. Important questions that remain to be
answered in future experiments include whether or not such
compounds, as well as FTS-ME, block the hydrophobic pocket
of RhoGDIs in vivo, and if so how does this affect cell growth.
In light of the important functions of RhoGDIs as specific
regulators of Rac/Rho GTPases trafficking to and from the cell
membrane,46 it is tempting to speculate that such blockage will
result in miss localization and functioning of these GTPases.
This would result in inhibition of cell growth.
therefore yields similar and not improved tumor growth inhibi-
tory effects compared to FTS.
Conclusions
Our results along with previous studies that used prenylated
small molecules47,48 show that the structure of such molecules
is a critical functional determinant. While FTS and its amide
derivatives act as Ras inhibitors, FTS esters show lesser activity
as Ras inhibitors and may interfere with the function of other
prenyl-binding proteins such as RhoGDIs and thereby inhibit
cell growth.
Experimental Section
Chemistry. The compounds prepared were FTS-methyl ester
(FTS-ME), FTS-methoxy-methyl ester (FTS-MOME), FTS-iso-
propyl ester (FTS-IE), FTS-benzyl ester (FTS-BE), FTS-amide
(FTS-A), FTS-methyl amide (FTS-MA), and FTS-dimethyl amide
(FTS-DMA). Direct coupling of FTS with the appropriate amine,
alcohol, or methoxymethychloride provided FTS-MOME, FTS-IE,
FTS-BE, FTS-A, FTS-MA, and FTS-DMA. FTS-ME was prepared
from methyl thiosalicylate and farnesol.
S-trans,trans-Farnesylthiosalicylamide (FTS-A). A mixture of
9.0 g of S-trans,trans-farnesylthiosalicylic acid (FTS), 9.6 g of
1-ethyl-3-[dimethylaminopropyl]carbodiimide hydrochloride, 6.75 g
of 1-hydroxybenzotriazole, 2.7 g of ammonium chloride, and 150
mL of dimethylformamide was sonicated for 5 min and then cooled
in an ice bath under argon. Diisopropylethylamine, 12.95 g, was
added dropwise with stirring at 0 °C, and the mixture was stirred
at ice bath temperature for 2.5 h. Ethyl acetate was added, the
mixture was washed with aqueous citric acid, aqueous sodium
bicarbonate, and brine, and the organic layer was dried over
anhydrous sodium sulfate and concentrated to yellow syrup.
Trituration with hexane provided 12.9 g (87%) of off-white solid,
The relatively strong anti-Ras activity of the FTS-amides,
particularly of FTS-amide, suggests that amidation, unlike
methylation, of the carboxyl group is favorable for the interac-
tions with Ras escort proteins. So then, while methylation
abrogated anti-Ras activity, amidation strengthened this activity.
Importantly, FTS-amides retained the selectivity toward active
Ras-GTP of the parental compound FTS. This was apparent in
their inhibitory activities in Panc-1 and in U87 cells; in Panc-1
cells, the most prevalent active Ras isoform is K-Ras-GTP, as
they express the oncogenic K-RasG12V.37 Similarly, we showed
previously that N-Ras-GTP and to a lesser extent K-Ras-GTP
or H-Ras-GTP was inhibited by FTS in U87 cells.26 Of all
1
purity g97%. H NMR (300 MHz, DMSO-d6) δ: 7.75 (bs, 1H),
7.43-7.33 (m, 4H), 7.22-7.12 (m, 1H), 5.25 (t, 1H), 5.06 (bt, 2H),
3.54 (d, 2H), 2.6-1.88 (m, 8H), 1.63 (s, 6H), 1.56 (s, 6H). LC/
MS: (+) ESI m/z ) 358 [M + H]+; 380 [M + Na]+.
N-Methyl-S-trans,trans-Farnesylthiosalicylamide
(FTS-MA). A mixture of 5.0 g of S-trans,trans-farnesylthiosalicylic
acid (FTS), 80 mL of dichloromethane, 16.0 mL of methylamine,
and 6.4 mL of triethylamine was stirred under argon, 6.8 g of
benzotriazol-1-yl-oxy-tris-(dimethylamino)phosphonium hexafluo-
rophosphate was added, and the mixture was stirred for 18 h at
room temperature. The mixture was concentrated under reduced
pressure to an oil, which was purified by flash chromatography,
eluting with hexanes/ethyl acetate (9/1-3/7) to provide 5.18 g
(99%) of yellow oil, purity g98%. 1H NMR (300 MHz, CDCl3) δ:
7.70 (dd, 1H), 7.33 (m, 3H), 6.85 (s, 1H), 5.26 (t, 1H), 5.06 (m,
2H), 3.53 (d, 2H), 3.02 (d, 3H), 2.02 (m, 8H), 1.68 (s, 3H), 1.59
(s, 3H), 1.58 (s, 6H). LC/MS: (+) ESI m/z ) 372 [M + H]+; 394
[M + Na]+.
prenylated derivatives with anti-Ras activity, FTS-A (IC50
10 µM) appears to be the most potent inhibitor of active Ras
)
described thus far.
We therefore examined the impact of FTS-A on tumor growth
in animal models. We used an oral formulation to allow
comparison with results obtained with oral FTS.27 In the Panc-1
nude mouse model, FTS-A exhibited a potency that was
comparable with that of FTS, namely, 100 mg/kg daily FTS-A
caused a ∼60% decrease in tumor weight, while in previous
work, FTS (80 mg/kg, daily) caused approximately the same
effect.27 Similarly, in the intracranial U87 nude mouse model,
treatment with oral FTS-A at daily dose of 100 mg/kg showed
clear tumor growth inhibition (Figure 5). Here, too, the effect
was not stronger than that observed with FTS in the same model
(Goldberg et al. in preparation). Thus, in spite of the in vitro
higher potency of FTS-A as compared with FTS, the in vivo
effects of the two inhibitors were similar. This apparent
discrepancy is clearly associated with differences in pharma-
cokinetics and pharmacodynamics of FTS-A and FTS. Although
FTS-A exhibits a longer elimination half-time than FTS (Table
3), it seems to be less available than FTS in the plasma and
N,N-Dimethyl-S-trans,trans-Farnesylthiosalicylamide
(FTS-DMA). A mixture of 5.0 g of S-trans,trans-farnesylthiosali-
cylic acid (FTS), 80 mL of dichloromethane, 16.0 mL of dimethy-
lamine, and 6.4 mL of triethylamine was stirred under argon, 6.8 g
of benzotriazole-1-yl-oxy-tris-(dimethylamino)phosphonium hexaflu-
orophosphate was added, and the mixture was stirred for 18 h at
room temperature. The mixture was concentrated under reduced
pressure to an oil, which was purified by flash chromatography,
eluting with hexanes/ethyl acetate (9/1-3/7) to provide 5.19 g
1
(96%) of a pale-yellow oil, purity g99%. H NMR (300 MHz,
CDCl3) δ: 7.39 (dd, 1H), 7.27 (m, 3H), 5.28 (t, 1H), 5.08 (bm,
2H), 3.56 (d, 2H), 3.14 (s, 3H), 2.83 (s, 3H), 1.68 (s, 3H), 1.60 (s,
3H), 1.59 (s, 6H). LC/MS: (+) ESI m/z ) 386 [M + H]+; 408 [M
+ Na]+.
Methyl S-trans,trans-Farnesylthiosalicylate (FTS-ME). A
solution of 7.0 g of trans,trans-farnesol, 63 mL of tetrahydrofuran,
and 5.7 mL of triethylamine was cooled to -50 °C. With stirring
under argon, a solution of 2.9 mL of methanesulfonyl chloride in

FTS DeriVatiVes
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1 203
21 mL of tetrahydrofuran was added dropwise over 30 min, and
the mixture was stirred for 2.5 h at -50 °C. Then a solution of 4.3
mL of methyl thiosalicylate, 42 mL of tetrahydrofuran, and 4.4
mL of triethylamine was added dropwise with stirring at -50 °C,
the mixture was stirred at -50 °C for 2.5 h, and at 0 °C for 12 h.
The mixture was filtered and concentrated to a syrup, which was
partitioned between chloroform and 2% hydrochloric acid. The
chloroform layer was washed with water, dried over sodium sulfate,
and concentrated to an oil, which was purified on a filter pad of
flash silica, eluting with hexanes/ethyl acetate (80/1-40/1) to
provide 7.2 g (62%) of colorless oil, purity g96%. 1H NMR (300
MHz, CDCl3) δ: 7.95 (dd, 1H), 7.42 (td, 1H), 7.31 (d, 1H), 7.15
(t, 1H), 5.34 (bt, 1H), 5.09 (bt, 2H), 3.91 (s, 3H), 3.58 (d, 2H),
2.1-2.0 (m, 6H), 1.9-1.8 (m, 2H), 1.73 (s, 3H), 1.68 (s, 3H), 1.60
(s, 6H). LC/MS: (+) ESI m/z ) 373 [M + H]+; m/z ) 395 [M +
Na]+.
Isopropyl S-trans,trans-Farnesylthiosalicylate (FTS-IE). To
a stirred solution of 6.39 g of S-trans,trans-farnesylthiosalicylic
acid (FTS), 2.61 g of dimethylaminopyridine, and 162 mL of
dichloromethane was added 4.78 g of dicyclohexyldiimide. Then
4.09 mL of isopropyl alcohol was added, and the mixture was stirred
at ambient temperature overnight. The mixture was filtered, and
the filtrate was concentrated to an oil, which was chromatographed
on a filterpad of flash silica, eluting with hexanes/ethyl acetate (20/
1) to provided 5.88 g (83%) of colorless oil, purity g98%. 1H NMR
(300 MHz, CDCl3) δ: 7.93 (dd, 1H), 7.40 (td, 1H), 7.31 (d, 1H),
7.14 (td, 1H,), 5.34 (t, 1H), 5.26 (m, 1H), 5.09 (t, 2H), 3.57 (d,
2H), 2.05 (m, 6H), 1.98 (m, 2H), 1.72 (s, 3H), 1.68 (s, 3H), 1.60
(s, 6H), 1.37 (s, 6H). LC/MS (+) ESI: m/z 401 [M + H];1 m/z 423
[M + Na]+.
50, 25, 12.5, and 6.25 µM of each compound, which were then
used in the experiments. For in vivo experiments, FTS-A was
weighed and dissolved in corn oil to a concentration of 100 mg/
kg. The vehicle control was corn oil.
Pharmacokinetic Sampling and Analysis. Assay development
and plasma concentration determinations were conducted by NoAb
BioDiscoveries. Four animals (male Sprague-Dawley rats, 275-285
g) were used to collect samples at each time point (10, 30, 50, 75,
105, 135, 165, 210, 270, and 330 min) taken after a single dose of
100 mg/kg FTS-amide, po. Approximately 0.2 mL of blood was
collected in sodium heparin containing tubes via the femoral artery
while under anesthesia (urethane/chloralose, 33%/1% w/v) at each
of the specified collection times. The blood samples were im-
mediately placed on ice and centrifuged (6800 rpm for 5 min at 4
°C) within 5 min of collection. Plasma was then collected and
immediately stored at -70 °C. Peak plasma concentration, time
peak achieved, and half-life (Cmax, tmax, and t1/2) were calculated.
Area under the plasma concentration-time curves (AUC) was also
determined. These analyses were conducted using Win-Nonlin
(Phasight Corp).
Preparation of Calibration Standards and Extraction of
Plasma for Analysis. Calibration standards of FTS-amide in rat
plasma were prepared at 1, 5, 25, 125, 1250, and 2500 ng/mL. A
sample (60 µL) of each calibration standard or plasma was mixed
with 2 µL of the internal standard (17.5 µL/mL propanolol in
acetonitrile) and then vortex mixed with 600 µL of acetonitrile to
precipitate proteins. The samples were then centrifuged at 12000
rpm and 4 °C for 5 min. the supernatant was then transferred to
glass tubes and evaporated to dryness under nitrogen at 40 °C. The
residue was reconstituted with 60 µL of mobile phase (90%/10%
of acetonitrile/5 mM ammonium acetate with 0.05% formic acid)
and vortexed well. The precipitated samples were then centrifuged
at 12000 rpm and 4 °C for 5 min and transferred into auto sampler
vials.
Liquid Chromatography. Sample analysis was performed using
a PE Sciex API 4000 Q-TRAP LC-MS/MS system equipped with
an Agilent LC system with a binary pump and solvent degasser,
an auto sampler, and a divert valve (VIVI, Valco Instrument Co.
Inc.) installed between the column and the mass spectrometer inlet.
An Onyx Monolithic C-18, 100 mm × 3.0 mm column was used.
Mobile phase was acetonitrile/5 mM ammonium acetate with 0.05%
formic acid (90/10, isocratic elution) at a flow rate of 1 mL/min,
and the injection volume was 10 µL. Propanolol was used as an
internal standard (IS). A sample cleanup method for plasma spiked
with the test compound was developed using protein precipitation.
Method qualification included: the determination of the ion transi-
tion for the compound and the IS (i.e., identification of the parent
and daughter ions), determination of the linear dynamic range using
five calibration standards (1-2500 ng/mL) in duplicate, and the
evaluation of assay performance determined at three quality control
(low, medium, and high) samples in duplicate.
Benzyl S-trans,trans-Farnesylthiosalicylate (FTS-BE). To a
solution of 7.50 g of S-trans,trans-farnesylthiosalicylic acid (FTS),
3.06 g of dimethylaminopyridine, and 189 mL of dichloromethane
was added with stirring 5.61 g of dicyclhexylcarbodiimide. Then
4.76 mL of benzyl alcohol was added, and the mixture was stirred
at ambient temperature overnight. The mixture was filtered, and
the filtrate concentrated to an oil, which was dissolved in ether.
The ether solution was washed with 2% hydrochloric acid and brine,
dried over sodium sulfate, and concentrated to an oil, which was
chromatographed on a filterpad of flash silica, eluting with hexanes/
ethyl acetate (20/1) to afford 6.85 g (73%) of colorless oil, purity
1
g98%. H NMR (300 MHz, CDCl3) δ: 8.00 (dd, 1H), 7.47-7.30
(m, 7H),7.13 (t, 1H), 5.35 (m, 1H), 5.09 (bt, 2H), 3.58 (d, 2H,
2.08-1.96 (m, 8H), 1.72 (s, 3H), 1.68 (s, 3H), 1.60 (s, 6H). LC/
MS (+) ESI: m/z ) 449 [M + H]+; m/z ) 471 [M + Na]+.
Methoxymethyl S-trans,trans-Farnesylthiosalicylate (FTS-
MOME). To a flask containing 0.264 g of sodium hydride under
argon was added with stirring 1 mL of hexamethylphosphoramide.
With cooling, a solution of 3.59 S-trans,trans-farnesylthiosalicylic
acid (FTS) in 10 mL of hexamethylphosphoramide was added
dropwise over 30 min, and the mixture was stirred at room
temperature for 30 min. In one portion, 0.84 mL of chloromethyl
methyl ether was added and stirring was continued for 3 h. The
mixture was poured into saturated aqueous sodium carbonate and
extracted with ether. The ether solution was washed with water
and brine, dried over sodium sulfate, and concentrated to an oil,
which was chromatographed on a filterpad of flash silica, eluting
with hexanes/ethyl acetate (20/1) to provide 3.1 g (78%) of colorless
Cell Cultures. Human glioblastoma U87 cells and Panc-1 cells
(ATCC) were grown in Dulbecco’s modified Eagle medium
(DMEM) containing 10% fetal calf serum (FCS), 2 mM L-
glutamine, 100 U/mL penicillin, and 100 µg/mL streptomycin. The
cells were incubated at 37 °C in a humidified atmosphere of 95%
air and 5% CO2.
Growth Inhibition Experiments. U87 and Panc-1 cells were
plated in 24-well plates at a density of 5000 cells per well and
grown for 24 h. The medium was then replaced by a medium
containing the indicated concentration of each compound or by a
medium containing 0.1% DMSO (vehicle control). The cells were
maintained in culture for 5-7 days and then counted. Data are
expressed in terms of percentage: (number of cells in the drug
treated culture/number of cells in the control × 100). Each
experiment was performed twice in quadruplicates and the means
( SD are presented.
1
oil, purity g99%. H NMR (300 MHz, CDCl3) δ: 8.03 (dd, 1H),
7.45 (t, 1H), 7.33 (d, 1H), 7.17 (td, 1H), 5.49 (s, 2H), 5.34 (bt,
1H), 5.09 (bt, 2H), 3.59 (d, 2H), 3.56 (s, 3H), 2.1-2.0 (m, 6H),
1.99-1.96 (m, 2H), 1.73 (s, 3H), 1.68 (s, 3H), 1.59 (s, 6H). LC/
MS (+) ESI: m/z 425 [M + Na]+.
Preparation of Compounds for the Experiments. The com-
pounds were dissolved in chloroform to yield 0.1 M solutions of
each. Aliquots (10 or 20 µL) of the solutions were kept at -70 °C.
For each experiment, the chloroform was evaporated by a stream
of nitrogen, and the compound was dissolved in 100% DMSO to
yield a 0.1 M solution. This solution was diluted 1:1000 in DMEM/
10%FCS to yield a 100 µM solution, which was then serially diluted
with DMEM/10%FCS/0.1% DMSO to yield solutions containing
Western Immunoblotting. U87 or Panc-1 cells were plated at
a density of 1.5 × 106 cells in 10-cm dishes, and were allowed to
grow overnight in medium containing 10% FCS. The medium was
then replaced by a medium containing either 50 µM of each

204 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1
Goldberg et al.
compound or 0.1% DMSO, and the cells were incubated for 24 h
in the presence of the drugs. We then lysed the cells with
solubilization buffer (50 mmol/L Tris-HCl (pH 7.6), 20 mmol/L
MgCl2, 200 mmol/L NaCl, 0.5% NP40, 1 mmol/L DTT, and
protease inhibitors). Lysates were used to determine the Ras-GTP
content by the glutathione S-transferase (GST)-RBD (Ras-binding
domain of Raf) pull-down assay, followed by Western immunob-
lotting with mouse anti pan-Ras antibody (1:2500) or with Ras-
isoform specific antibodies (mouse anti-H-Ras, anti-K-Ras, and anti
N-Ras (1:100, Calbiochem) as described elsewhere.49 Levels of
total Ras proteins, phosphor-extracellular signal-regulated kinase
(ERK), phospho-Akt, and ꢀ-tubulin in the cell lysates were
determined by immunoblotting using the above anti-Ras antibodies,
rabbit antiphospho-ERK1/2 antibody (1:10000, Sigma-Aldrich),
rabbit antiphospho-Akt antibody (1:200, Cell Signaling), and rabbit
anti ꢀ-tubulin (1:1000, Santa Cruz Biotechnology). Immunoblots
were exposed either to peroxidase-goat antimouse IgG or peroxi-
dase-goat antirabbit IgG (1:2500) accordingly. Protein bands were
visualized by enhanced chemiluminescence and quantified by
densitometry using EZQuant-Gel computer software.
Tumor Implantation. Nude CD1-Nu mice (25-30 g) were
housed in barrier facilities on a 12 h light/dark cycle. Food and
water were supplied ad libitum. On day zero, Panc-1 cells (5 ×
106 cells in 0.1 mL) were implanted subcutaneously just above the
right femoral joint. On day 10, the tumors were measured and the
mice were divided into treated and control groups. The treated
animals received daily po doses of 100 µL FTS-amide (100 mg/
kg), while the control animals received daily po doses of 100 µL
vehicle solution. Approximately 18 days later, the tumors were
extracted and weighed.
For intracranial tumor implantation, mice were anesthetized by
intraperitoneal injection of ketamin (100 mg/kg) and xylazine
hydrochloride (10 mg/kg) solution in 0.1 mL of PBS. Cell
inoculation was performed on day zero by creating a mid-skull
incision, and then a burr hole 1 mm lateral and 1 mm posterior to
the bregma, 3 mm deep, into the striatum area. U-87 cells (0.5 ×
106) were injected using a 1 mL of BASI gastight syringe connected
to a pump, thus allowing slow release of the cells into the brain (1
µL/min, 5 min total). The mice were imaged on day 4 to verify
tumor development as detailed in MRI procedures, and the mice
were divided into two groups exhibiting similar tumor sizes. Drug
treatments started on day 5: the treated group received po dosing
of 50 µL of FTS-amide (100 mg/kg) twice a day, while the control
animals received po doses of 50 µL vehicle solutions twice a day.
Followup MR images were taken on days 9, 14, and 18.
Image Analysis (MIA version 2.4), MATLAB image processing
toolbox. The normalized tumor growth between the follow-ups was
calculated as detailed in the equation below for each animal and
the results were averaged for each treatment group.
normalized tumor growth change )
tumor volume(follow-up(n)) - tumor volume(follow-up 1)
tumor volume(follow-up 1)
Acknowledgment. We thank the Alfredo Federico Strauss
Center for Computational NeuroImaging of Tel Aviv University.
The MRI scanner used in this study was purchased with grants
from The Israel Science Foundation (ISF) and the Raymond
and Beverly Sackler Institute of Biophysics of Tel Aviv
University. Y.K. is the incumbent of the Jack H. Skirball Chair
in Applied Neurobiology. Grant support: supported in part by
The Wolfson Foundation and by the Prajs-Drimmer Institute
for the Development of Anti-Degenerative Disease Drugs
Supporting Information Available: Elemental analyses of new
farnesylthiosalicylic acid derivatives. This material is available free
of charge via the Internet at http://pubs.acs.org.
References
(1) Barbacid, M. Ras genes. Annu. ReV. Biochem. 1987, 56, 779–827.
(2) Boguski, M. S.; McCormick, F. Proteins regulating Ras and its
relatives. Nature 1993, 366, 643–654.
(3) Bos, J. L. p21ras: an oncoprotein functioning in growth factor-induced
signal transduction. Eur. J. Cancer 1995, 31, 1051–1054.
(4) Casey, P. J.; Solski, P. A.; Der, C. J.; Buss, J. E. p21ras is modified
by a farnesyl isoprenoid. Proc. Natl. Acad. Sci. U.S.A 1989, 86, 8323–
8327.
(5) Hancock, J. F.; Cadwallader, K.; Paterson, H.; Marshall, C. J. A CAAX
or a CAAL motif and a second signal are sufficient for plasma
membrane targeting of ras proteins. EMBO J. 1991, 10, 4033–4039.
(6) Hancock, J. F.; Magee, A. I.; Childs, J. E.; Marshall, C. J. All ras
proteins are polyisoprenylated but only some are palmitoylated. Cell
1989, 57, 1167–1177.
(7) Kato, K.; Cox, A. D.; Hisaka, M. M.; Graham, S. M.; Buss, J. E.;
Der, C. J. Isoprenoid addition to Ras protein is the critical modification
for its membrane association and transforming activity. Proc. Natl.
Acad. Sci. U.S.A 1992, 89, 6403–6407.
(8) Leevers, S. J.; Paterson, H. F.; Marshall, C. J. Requirement for Ras
in Raf activation is overcome by targeting Raf to the plasma
membrane. Nature 1994, 369, 411–414.
(9) Stokoe, D.; MacDonald, S. G.; Cadwallader, K.; Symons, M.; Hancock,
J. F. Activation of Raf as a result of recruitment to the plasma
membrane. Science 1994, 264, 1463–1467.
MRI Procedures. In all MRI experiments, the mice were
anesthetized with isofluorane (5% for induction, 1-3% for main-
tenance) in 1 L/min oxygen. Ten minutes before the MR imaging,
the mice were injected with 0.1 mL of gadolinium diethylenetri-
aminepentaacetic acid (Gd-DTPA, Magnetol, Soreq Radiopharma-
ceuticals, Yavne, Israel) solution (0.05M). Once anesthetized, the
animals were installed in a head-holder to ensure reproducible
positioning inside the probe. Rate of respiration was monitored and
was 60-80 breaths min-1 throughout the experimental period.
We used a 7 T BioSpec Magnet 70/30 USR system (Bruker,
Germany) equipped with gradient coil system capable of producing
pulse gradient of up to 40 gauss/cm in each of the three directions.
The MRI protocol included transverse T1- and T2-weighted MR
images. The T1-weighted images were acquired using the spin-echo
imaging sequence with a repetition delay (TR) of 800 ms, an echo
delay (TE) of 12 ms, matrix dimension of 256 × 128 (interpolated
to 256 × 256), and two averages, corresponding to an image
acquisition time of 3 min 24 s. The T2-weighted images were
acquired using the RARE sequence with a TR of 3000 ms, TE
effective of 50 ms, 256 × 128 matrix (interpolated to 256 × 256),
RARE factor of 8, and four averages, corresponding to an image
acquisition time of 3 min 12 s. Then 14 continuous slices with
slice thickness of 1 mm were acquired with a field of view (FOV)
of 2 cm × 2 cm.
(10) Goldberg, L.; Kloog, Y. A Ras Inhibitor Tilts the Balance between
Rac and Rho and Blocks PI3-Kinase-Dependent Glioblastoma Cell
Migration. Cancer Res. 2006, 66, 11709–11717.
(11) Joneson, T.; White, M. A.; Wigler, M. H.; Bar Sagi, D. Stimulation
of membrane ruffling and MAP kinase activation by distinct effectors
of RAS. Science 1996, 271, 810–812.
(12) Walsh, A. B.; Bar-Sagi, D. Differential activation of the Rac pathway
by Ha-Ras and K-Ras. J. Biol. Chem. 2001, 276, 15609–15615.
(13) Downward, J. Ras signalling and apoptosis. Curr. Opin. Genet. DeV.
1998, 8, 49–54.
(14) Scheffzek, K.; Ahmadian, M. R.; Kabsch, W.; Wiesmuller, L.;
Lautwein, A.; Schmitz, F.; Wittinghofer, A. The Ras-RasGAP
complex: structural basis for GTPase activation and its loss in
oncogenic Ras mutants. Science 1997, 277, 333–338.
(15) Bos, J. L. Ras oncogenes in human cancer: a review. Cancer Res.
1989, 49, 4682–4689.
(16) Cox, A. D.; Der, C. J. Protein prenylation: more than just glue. Curr.
Opin. Cell Biol. 1992, 4, 1008–1016.
(17) Marshall, C. J. Ras effectors. Curr. Opin. Cell Biol. 1996, 8, 197–
204.
(18) Cox, A. D.; Hisaka, M. M.; Buss, J. E.; Der, C. J. Specific isoprenoid
modification is required for function of normal, but not oncogenic,
Ras protein. Mol. Cell. Biol. 1992, 12, 2606–2615.
(19) Elad-Sfadia, G.; Haklai, R.; Ballan, E.; Gabius, H. J.; Kloog, Y.
Galectin-1 augments Ras activation and diverts Ras signals to Raf-1
at the expense of phosphoinositide 3-kinase. J. Biol. Chem. 2002, 277,
37169–37175.
Image processing was as follows: The tumor volume was
calculated from the T1-weighted MR images using the Medical
(20) Rotblat, B.; Niv, H.; Andre, S.; Kaltner, H.; Gabius, H. J.; Kloog, Y.
Galectin-1(L11A) predicted from a computed galectin-1 farnesyl-

FTS DeriVatiVes
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 1 205
binding pocket selectively inhibits Ras-GTP. Cancer Res. 2004, 64,
(35) Gana-Weisz, M.; Haklai, R.; Marciano, D.; Egozi, Y.; Ben Baruch,
G.; Kloog, Y. The Ras antagonist S-farnesylthiosalicylic acid induces
inhibition of MAPK activation. Biochem. Biophys. Res. Commun.
1997, 239, 900–904.
(36) Schroeder, H.; Leventis, R.; Rex, S.; Schelhaas, M.; Nagele, E.;
Waldmann, H.; Silvius, J. R. S-Acylation and plasma membrane
targeting of the farnesylated carboxyl-terminal peptide of N-ras in
mammalian fibroblasts. Biochemistry 1997, 36, 13102–13109.
(37) Berrozpe, G.; Schaeffer, J.; Peinado, M. A.; Real, F. X.; Perucho, M.
Comparative analysis of mutations in the p53 and K-ras genes in
pancreatic cancer. Int. J. Cancer 1994, 58, 185–191.
(38) Guha, A.; Feldkamp, M. M.; Lau, N.; Boss, G.; Pawson, A.
Proliferation of human malignant astrocytomas is dependent on Ras
activation. Oncogene 1997, 15, 2755–2765.
(39) Chiu, V. K.; Silletti, J.; Dinsell, V.; Wiener, H.; Loukeris, K.; Ou,
G.; Philips, M. R.; Pillinger, M. H. Carboxyl methylation of Ras
regulates membrane targeting and effector engagement. J. Biol. Chem.
2004, 279, 7346–7352.
3112–3118.
(21) Elad-Sfadia, G.; Haklai, R.; Balan, E.; Kloog, Y. Galectin-3 augments
K-Ras activation and triggers a Ras signal that attenuates ERK but
not phosphoinositide 3-kinase activity. J. Biol. Chem. 2004, 279,
34922–34930.
(22) Belanis, L.; Plowman, S. J.; Rotblat, B.; Hancock, J. F.; Kloog, Y.
Galectin-1 is a novel structural component and a major regulator of
h-ras nanoclusters. Mol. Biol. Cell 2008, 19, 1404–1414.
(23) Marciano, D.; Ben Baruch, G.; Marom, M.; Egozi, Y.; Haklai, R.;
Kloog, Y. Farnesyl derivatives of rigid carboxylic acidssinhibitors
of ras-dependent cell growth. J. Med. Chem. 1995, 38, 1267–1272.
(24) Haklai, R.; Gana-Weisz, G.; Elad, G.; Paz, A.; Marciano, D.; Egozi,
Y.; Ben Baruch, G.; Kloog, Y. Dislodgment and accelerated degrada-
tion of Ras. Biochemistry 1998, 37, 1306–1314.
(25) Barkan, B.; Starinsky, S.; Friedman, E.; Stein, R.; Kloog, Y. The Ras
inhibitor farnesylthiosalicylic acid as a potential therapy for neurofi-
bromatosis type 1. Clin. Cancer Res. 2006, 12, 5533–5542.
(26) Blum, R.; Jacob-Hirsch, J.; Amariglio, N.; Rechavi, G.; Kloog, Y.
Ras inhibition in glioblastoma down-regulates hypoxia-inducible
factor-1alpha, causing glycolysis shutdown and cell death. Cancer Res.
2005, 65, 999–1006.
(27) Haklai, R.; Elad-Sfadia, G.; Egozi, Y.; Kloog, Y. Orally administered
FTS (Salirasib) inhibits human pancreatic tumor growth in nude mice.
Cancer Chemother. Pharmacol. 2007, in press.
(28) Yaari, S.; Jacob-Hirsch, J.; Amariglio, N.; Haklai, R.; Rechavi, G.;
Kloog, Y. Disruption of cooperation between Ras and MycN in human
neuroblastoma cells promotes growth arrest. Clin. Cancer Res. 2005,
11, 4321–4330.
(29) Zundelevich, A.; Elad-Sfadia, G.; Haklai, R.; Kloog, Y. Suppression
of lung cancer tumor growth in a nude mouse model by the Ras
inhibitor Salirasib (farnesylthiosalicylic acid). Mol. Cancer Ther. 2007,
31, 31.
(30) Blum, R.; Elkon, R.; Yaari, S.; Zundelevich, A.; Jacob-Hirsch, J.;
Rechavi, G.; Shamir, R.; Kloog, Y. Gene Expression Signature of
Human Cancer Cell Lines Treated with the Ras Inhibitor Salirasib
(S-Farnesylthiosalicylic Acid). Cancer Res. 2007, 67, 3320–3328.
(31) Blum, R.; Nakdimon, I.; Goldberg, L.; Elkon, R.; Shamir, R.; Rechavi,
G.; Kloog, Y. E2F1 identified by promoter and biochemical analysis
as a central target of glioblastoma cell-cycle arrest in response to Ras
inhibition. Int. J. Cancer 2006, 119, 527–538.
(40) Rando, R. R. Chemical biology of protein isoprenylation/methylation.
Biochim. Biophys. Acta 1996, 1300, 5–16.
(41) Marom, M.; Haklai, R.; Ben Baruch, G.; Marciano, D.; Egozi, Y.;
Kloog, Y. Selective inhibition of Ras-dependent cell growth by
farnesylthiosalisylic acid. J. Biol. Chem. 1995, 270, 22263–22270.
(42) Ashery, U.; Yizhar, O.; Rotblat, B.; Elad-Sfadia, G.; Barkan, B.;
Haklai, R.; Kloog, Y. Spatiotemporal Organization of Ras Signaling:
Rasosomes and the Galectin Switch. Cell. Mol. Neurobiol. 2006, 26,
469–493.
(43) Hoffman, G. R.; Nassar, N.; Cerione, R. A. Structure of the Rho family
GTP-binding protein Cdc42 in complex with the multifunctional
regulator RhoGDI. Cell 2000, 100, 345–356.
(44) Scheffzek, K.; Stephan, I.; Jensen, O. N.; Illenberger, D.; Gierschik,
P. The Rac-RhoGDI complex and the structural basis for the regulation
of Rho proteins by RhoGDI. Nat. Struct. Biol. 2000, 7, 122–126.
(45) Mondal, M. S.; Wang, Z.; Seeds, A. M.; Rando, R. R. The specific
binding of small molecule isoprenoids to rhoGDP dissociation inhibitor
(rhoGDI). Biochemistry 2000, 39, 406–412.
(46) Michaelson, D.; Silletti, J.; Murphy, G.; D’Eustachio, P.; Rush, M.;
Philips, M. R. Differential localization of Rho GTPases in live cells:
regulation by hypervariable regions and RhoGDI binding. J. Cell Biol.
2001, 152, 111–126.
(32) Aharonson, Z.; Gana-Weisz, M.; Varsano, T.; Haklai, R.; Marciano,
D.; Kloog, Y. Stringent structural requirements for anti-Ras activity
of S-prenyl analogues. Biochim. Biophys. Acta 1998, 1406, 40–50.
(33) Shohami, E.; Yatsiv, I.; Alexandrovich, A.; Haklai, R.; Elad-Sfadia,
G.; Grossman, R.; Biegon, A.; Kloog, Y. The Ras inhibitor S-
trans,trans-farnesylthiosalicylic acid exerts long-lasting neuroprotection
in a mouse closed head injury model. J. Cereb. Blood Flow Metab.
2003, 23, 728–738.
(34) Weisz, B.; Giehl, K.; Gana-Weisz, M.; Egozi, Y.; Ben-Baruch, G.;
Marciano, D.; Gierschik, P.; Kloog, Y. A new functional Ras
antagonist inhibits human pancreatic tumor growth in nude mice.
Oncogene 1999, 18, 2579–2588.
(47) Ding, J.; Lu, D. J.; Perez Sala, D.; Ma, Y. T.; Maddox, J. F.; Gilbert,
B. A.; Badwey, J. A.; Rando, R. R. Farnesyl-L-cysteine analogs can
inhibit or initiate superoxide release by human neutrophils. J. Biol.
Chem. 1994, 269, 16837–16844.
(48) Tan, E. W.; Perez Sala, D.; Canada, F. J.; Rando, R. R. Identifying
the recognition unit for G protein methylation. J. Biol. Chem. 1991,
266, 10719–10722.
(49) de Rooij, J.; Bos, J. L. Minimal Ras-binding domain of Raf1 can be
used as an activation-specific probe for Ras. Oncogene 1997, 14,
623–625.
JM801165R