TX-1123: An antitumor 2-hydroxyarylidene-4-cyclopentene-1,3-dione as a protein tyrosine kinase inhibitor having low mitochondrial toxicity

Article abstract of DOI:10.1016/S0968-0896(02)00160-8

A series of 2-hydroxyarylidene-4-cyclopentene-1,3-diones were designed, synthesized, and evaluated with respect to protein tyrosine kinase (PTK) inhibition, mitochondrial toxicity, and antitumor activity. Our results show that the cyclopentenedione-derived TX-1123 is a more potent antitumor tyrphostin and also shows lower mitochondrial toxicity than the malononitrile-derived AG17, a potent antitumor tyrphostin. The O-methylation product of TX-1123 (TX-1925) retained its tyrphostin-like properties, including mitochondrial toxicity and antitumor activities. However, the methylation product of AG17 (TX-1927) retained its tyrphostin-like antitumor activities, but lost its mitochondrial toxicity. Our comprehensive evaluation of these agents with respect to protein tyrosine inase inhibition, mitochondrial inhibition, antitumor activity, and hepatotoxicity demonstrates that PTK inhibitors TX-1123 and TX-1925 are more promising candidates for antitumor agents than tyrphostin AG17. Copyright

Full text of DOI:10.1016/S0968-0896(02)00160-8

Bioorganic & Medicinal Chemistry 10 (2002) 3257–3265
TX-1123:
An Antitumor 2-Hydroxyarylidene-4-cyclopentene-1,3-Dione
as a Protein Tyrosine Kinase Inhibitor Having Low
Mitochondrial Toxicity
Hitoshi Hori,a,* Hideko Nagasawa,a Masaki Ishibashi,a Yoshihiro Uto,a Akihiko Hirata,a
b
Kouichi Saijo,a Kazuto Ohkura,a Kenneth L.Kirk and Yoshimasa Ueharac
aDepartment of Biological Science & Technology, Faculty of Engineering, The University of Tokushima,
Tokushima 770-8506, Japan
bLaboratory of Bioorganic Chemistry, National Institute of Diabetes and Digestive and Kidney Diseases,
National Institute of Health, Bethesda, MD 20892, USA
cDepartment of Bioactive Molecules, National Institute of Infectious Diseases, 1-23-1 Shinjuku-ku, Tokyo 162-8640, Japan
Received 28 December 2001; accepted 11 May 2002
Abstract—A series of 2-hydroxyarylidene-4-cyclopentene-1,3-diones were designed, synthesized, and evaluated with respect to
protein tyrosine kinase (PTK) inhibition, mitochondrial toxicity, and antitumor activity.Our results show that the cyclopentene-
dione-derived TX-1123 is a more potent antitumor tyrphostin and also shows lower mitochondrial toxicity than the malononitrile-
derived AG17, a potent antitumor tyrphostin.The O-methylation product of TX-1123 (TX-1925) retained its tyrphostin-like
properties, including mitochondrial toxicity and antitumor activities.However, the methylation product of AG17 (TX-1927)
retained its tyrphostin-like antitumor activities, but lost its mitochondrial toxicity.Our comprehensive evaluation of these agents
with respect to protein tyrosine kinase inhibition, mitochondrial inhibition, antitumor activity, and hepatotoxicity demonstrates
that PTK inhibitors TX-1123 and TX-1925 are more promising candidates for antitumor agents than tyrphostin AG17. # 2002
Elsevier Science Ltd.All rights reserved.
Introduction
bis-substrate (protein substrate and ATP) quinoline
3
analogues (Fig.1).
The protein tyrosine kinases (PTKs) are members of a
large family of oncoproteins and proto-oncoproteins
that play major roles in signal transduction during nor-
mal cell division, terminal cell differentiation, and
apoptosis.1ꢀ5 Because enhanced PTK activity is asso-
ciated with proliferative disorders such as cancers, PTK
inhibitors have been developed as potential therapeutic
agents for cancer treatment.6ꢀ8 Indeed, PTK inhibitors
such as tyrphostins have been shown to inhibit growth
of tumor cell lines9 as well as tumor growth in vivo.10,11
Tyrphostins were originally described as a group of low-
molecular-weight protein tyrosine kinase inhibitors
designed to compete with substrate rather than
ATP.12,13 Tyrphostins include three structurally diverse
classes of compounds: the dihydroxy-cis-benzylidene-
malononitrile group, lavendustin derived-blockers, and
In general, criteria for therapeutic potential include target
selectivity, cell permeability, bioavailability, appropriate
pharmacokinetic properties, and absence of host toxicity.
In a screening study AG17 (Fig.2) was found to be the
most potent tumor-cell-growth inhibitor among a series
of tyrphostins examined as inhibitors of breast carcinoma
cell growth.14 Prior studies with AG17 had demonstrated
that it could function as an inhibitor of the epidermal
growth factor receptor tyrosine kinase (EGFR-K).15
Studies in intact cells have revealed that AG17 is a potent
inhibitor of platelet-derived growth factor (PDGF)-stimu-
lated growth of rabbit vascular smooth muscle cells as
well as PDGF-mediated tyrosine phosphorylation of
several intracellular substrates.16 Other investigators have
further characterized AG17 as a potent inhibitor of epi-
dermal growth factor-stimulated pancreatic carcinoma cell
growth17 and interleukin-2-stimulated human peripheral
mononuclear cell proliferation.18
*Corresponding author.Te:l. +81-88-656-7514; fax: +81-88-656-
9164; e-mail: hori@bio.tokushima-u.ac.jp
0968-0896/02/$ - see front matter # 2002 Elsevier Science Ltd.All rights reserved.
PII: S0968-0896(02)00160-8

3258
H. Hori et al. / Bioorg. Med. Chem. 10 (2002) 3257–3265
Figure 1. Structures of three classes of tyrphostins.
Figure 2. Structures of KIH-200–202, TX-1123, tyrphostin AG17, and their derivatives and analogues.
Recently, we found that AG17 suppressed insulin action
in rat white adipocytes.19 Sausville and co-authors
found that AG17 was the most potent tumor cell growth
inhibitor among a series of tyrphostins examined as
inhibitors of breast carcinoma cell growth.20 They also
noted the possibility that AG17 may act in part by
altering mitochondrial function and/or structure, and that
impairment of mitochondrial function (mitochondrial
toxicity21) may be exploitable as a potentially useful
mechanism to modulate tumor cell proliferation.20 Two
years later, Levitzki reported that AG17 induces apoptosis
and inhibition of cdk2 activity through a mechanism
that purportedly does not involve reduction of cellular
ATP levels.Shortly thereafter Sausville cautioned against
Levitzki’s use22 of AG17 as a pure protein tyrosine
kinase inhibitor.23,24
In considering potential mechanisms of the anti-
proliferative action of AG17, as Sausville cautioned, we
should note that SF6847 (an alternate name for AG17) was
shown to act as an uncoupler of oxidative phosphorylation
in isolated rat liver and heart mitochondria.25 In this study,
we describe a new tyrphostin, TX-1123, as a tyrphostin
analogue of AG17 that has lower mitochondrial toxicity.
Results and Discussion
Design and synthesis
2-Arylidine-4-cyclopentene-1,3-diones were previously
reported to be antitumor agents.26 This unique soft-
acid-type electrophilic cyclopentene-1,3-dione moiety
was chosen as our lead structure for drug design.We

H. Hori et al. / Bioorg. Med. Chem. 10 (2002) 3257–3265
3259
then developed some 2-hydroxyarylidene-4-cyclopentene-
1,3-diones such as KIH-200, KIH-201, and KIH-202 as
non-nitro radiosensitizers and hypoxic cytotoxins,27 and
nedione moiety in TX-1123 would be more subject to
nucleophilic attack than the less electron-deficient
AG17.We calculated hydrophobicity of the compounds
using parameters such as logP (calculated and observed)
and dGW.In general, higher hydrophobicities coorelate
with more favorable pharmacological activities.dGW, a
measure of solvation free energies, is a new hydro-
phobicity parameter that includes steric factors that are
not included in the calculation of partition coefficients
(logP).dGW values correlate better with obsd log P
than with calculated logP.Thus, hydrophobicities and
biological activities of cyclopentenedione-type inhibitors
such as TX-1123 analogues and methylene-malononitrile-
type inhibitors such as tyrophostin AG17 and TX-1927
correlated with each other within each series.
as potent inhibitors of phosphate transport in mito-
28
chondria (Fig.2).
benzylidene-cyclopentenedione as
We also developed hydroxy-
new phenolic
a
enhancer of chemiluminescence and found KIH-201 to
be the most potent enhancer in the luminol–hydrogen–
horseradish peroxidase peroxide reaction.29 TX-1123
was also tested in this regard, but was found to have no
activity.We here describe the use of TX-1123 and its
analogues such as TX-1925, TX-1926, and TX-1918
(Fig.2) as tyrphostin AG17 analogues.The presence of
cyclopentenedione moiety instead of the malononitrile
moiety found in AG17 was expected to result in reduced
mitochondrial toxicity.
We synthesized TX-1123 and related compounds using
a modified Knoevenagel reaction of the appropriate
o,o0-disubstituted hydroxybenzaldehyde with 4-cyclo-
pentene-1,3-dione under acidic conditions, as shown in
Scheme 1.We found that direct methylation reactions
of TX-1123 and tyrphostin AG17 did not afford the
corresponding O-methyl derivatives, TX-1925 and TX-
1927.These reaction conditions were not optimized.
Their physical data including electron affinity
(ꢀELUMO), hydrophobicity and UV spectra are shown
in Table 1.
As shown in Table 1, TX-1123 and its analogues have
hydrophobicity parameters, such as dGW and calcd or
obsd logP, similar to those of AG17, presumably
reflecting the presence of two tert-butyl groups in each
structure.From the results of studies with of KIH-200,
KIH-201, and KIH-202, which have the same cyclopente-
nedione moieties, relative to SF6847 (AG17), TX-1123 and
analogues could be expected to be more potent inhibitors
of phosphate transport but should have reduced
uncoupling activities in mitochondria.In general, target
kinases appear to be very flexible, and their adenine
pockets are dominated by hydrophobic interactions.
These factors may conspire to produce a permissive
binding site which accepts a range of planar heterocyclic
systems, many of which have been uncovered by high-
throughput screening.30 These TX-1123 analogues having
sterically demanding o,o0-di-tert-butyl groups and
cyclopentenedione moieties are expected to act as
hydrophobic and reactive Michael reaction acceptors.
Biological activities
We examined kinase inhibitory activities, mitochondrial
toxicities, antitumor activities, and hepatotoxicities of
TX-1123 and related compounds.The data are pre-
sented in Tables 2–4, respectively.TX-1918 was the
most potent eukaryotic elongation factor 2 kinase
(eEF2-K) inhibitor (IC50=0.44 mM) among the com-
pounds tested.The inhibitory activity of TX-1123
against Src tyrosine kinase (Src-K), (IC50=2.2 mM) was
stronger than that of AG17 (IC50>350 mM).The dose–
response curves of inhibitory activity of TX-1123 and
AG17 against Src-K are shown in Figure 5.TX-1123
showed weaker uncoupling (Cmax=2 mM with
Vmax=260 natomO/mg/min) and ATP-synthesis inhi-
bitory activities (IC50=5 mM) than did AG17
In order to correlate their electron affinities (EAs) as a
measure of their tyrosine kinase inhibitory activities, we
calculated the orbital energies of the lowest unoccupied
molecular orbital (LUMO) of TX-1123 analogues.The
PM3 MO calculations show that TX-1123 (ELUMO
=
ꢀ0.82 eV) has electron density regions with high coeffi-
cients localized only at the cyclopentenedione moiety,
while AG17 has broad LUMO (ELUMO=ꢀ1.24 eV)
electron density regions at the methylene-malononitrile
moiety (Fig.3).The methylene-cyclopentenedione moiety
in TX-1123 is less sterically hindered than the methylene-
malononitrile moiety in AG17 (Fig.4).Based on these
LUMO coefficients and steric arguments we expected
that the more electrophilic (lower ELUMO) cyclopente-
(Cmax=0.02 mM
with
Vmax=270natomO/mg/min,
IC50=3.5 nM), as shown in Table 3. We propose that
the more potent kinase inhibitory activities of TX-1123
compared to that of AG17 is due to the presence of the
methylene-cyclopentene-1,3-dione moiety as a more
reactive Michael reaction acceptor in comparison with
that of the methylene-malononitrile moiety because of
Table 1. Physical data of TX-1123, tyrphostin AG17, and their derivatives and analogues
Compounds
TX-1123
Mr
Mp (ꢁC)
155–157
EA [ꢀELUMO (eV)]
0.82
dGW (kJ)
logP (calcd)
pKa
lmax (nm) (emax)
312.4
ꢀ60.4
2.33 (4.93)
7.40
473 (6181)
373 (12,230)
339 (9592)
374 (13,700)
508 (15,157)
464 (10,031)
364 (11,101)
336 (9979)
TX-1925
TX-1918
TX-1926
Tyrphostin AG17
326.4
228.3
362.5
282.4
110–112
186–188
154–156
131–134
0.88
0.83
0.84
1.24
ꢀ58.7
ꢀ69.7
ꢀ59.4
ꢀ62.4
2.40 (6.44)
2.12 (3.35)
2.39 (6.30)
2.32 (5.02)
—
7.30
7.85
6.87
TX-1927
296.4
104–105
1.32
ꢀ60.6
2.83 (5.52)
—
278 (24,107)

3260
H. Hori et al. / Bioorg. Med. Chem. 10 (2002) 3257–3265
Figure 3. LUMOs of TX-1123, tyrphostin AG17, TX-1925, TX-1927, TX-1926, and TX-1918.
localized LUMO coefficients of the methylene-cyclo-
pentene-1,3-dione moiety.
The antitumor activity of TX-1123 for HepG2 cells
(EC50=3.66 mM) was stronger than that of AG17
(EC50=34.4 mM for HepG2 cells), while its antitumor
activity against HCT116 cells (EC50=39 mM) was
weaker than that of AG17 (EC50=0.20 mM), as shown
in Table 4.These data show good correlation between
the cytotoxicity and kinase inhibitory activity in HepG2
cells, as well as a good correlation between cytotoxicity
and uncoupling activity in HCT 116 cells.However, the
precise mechanisms of action remain to be delineated.
AG17 (IC50=0.97 mM) also has much stronger hepato-
toxicity than TX-1123 (IC50=57 mM) as shown in Table
4.The difference of hepatotoxicities of AG17 and TX-
1123 are probably related to the potency of their
uncoupling activities.These data suggest that TX-1123
may be a more promising candidate than AG17 as an
antitumor PTK inhibitor agent, especially because of
the absence of mitochondrial- and hepato-toxicity.
With the exception of AG17, the common pharmaco-
phore for most of the known tyrphostins is an unhin-
dered phenolic styrene that can be viewed as a
‘dehydrogenated’ tyrosine-mimic.6 The molecular structure
Figure 4. Methylene-cyclopentenedione moiety in TX-1123 and meth-
ylene-malononitrile moiety in tyrphostin AG17.

H. Hori et al. / Bioorg. Med. Chem. 10 (2002) 3257–3265
3261
Scheme 1. Synthesis of TX-1123, tyrphostin AG17, and their derivatives and analogues.
of AG17 shows that its sterically hindered phenolic
hydroxyl group is expected to act as a protonophore,
but is not able to take part in hydrogen bonding.To
confirm the non-essential nature of the hydroxyl groups
in tyrphostins for their PTK inhibitory activities, we
compared kinase inhibitory activities of the O-methy-
lated TX-1123 (TX-1925) and O-methylated AG17 (TX-
1927, MeO-SF31) with those of TX-1123 and AG17.As
expected TX-1925 showed Src-K inhibitory activity
(IC50=3.1 mM) and antitumor activity (IC50=89 mM
for HCT116 cells; IC50=6.36 mM for HepG2 cells)
comparable to TX-1123.
It is reasonable to expect that these O-methylated phenolic
compounds, TX-1925 and TX-1927, should be weaker
uncouplers because of the absence of a protonophoric
moiety in the molecules.Indeed, TX-1925 showed a
weaker mitochondrial toxicity, with lower uncoupling
activity (Cmax=40 mM; Vmax=290 natomO/mg/min)
and ATP-synthesis inhibitory activity (IC50=26 mM) as
compared to TX-1123.Significantly, TX-1927 was
completely devoid of properties associated mitochondrial
toxicity, as shown by examination of uncoupling activity
(Cmax=200 mM; Vmax=130 natomO/mg/min) and ATP-
synthesis inhibitory activities (IC50=700 mM), and also
showed no hepatotoxicity (IC50=105 mM).Interestingly,
it possessed comparable antitumor activity for HepG2
cells (IC50=34.1 mM), but much lower activity for
HCT116 cells (IC50=22 mM) in comparison with AG17
(IC50=34.4 mM for HepG2 and IC50=0.20 mM for
HCT116 cells).
Table 2. Protein kinase inhibitory activities (IC50, mM) of TX-1123,
tyrphotin AG17, and their derivatives and analogues
Kinases
Drugs
PKA PKC
Src-K
eEF2-K EGFR-K
TX-1123
TX-1925
TX-1918
TX-1926
9.6
31
44
320
31
44
2.2
3.1
4.4
>27
3.2
3.1
0.44
>27
320
31
440
>27
>27 >27
Tyrphostin AG17 >350 >350 >350 (75%) >350 >350 (14%)
TX-1927 >340 >340 >340 >340 >340
Values in parentheses indicate the percent inhibition at the maximum
concentration when the inhibitory activities were measured.
Table 3. Mitochondrial cytotoxicity of TX-1123 & tyrphostin AG17 and their derivatives and analogues
Mitochondrial cytotoxicity
TX-1123
TX-1925
TX-1918
TX-1926
Tyrphostin AG17
TX-1927
Uncoupling activity
Cmax (mM)
Vmax (natomO/mg/min)
ATP synthesis inhibition: IC50 (mM)
2
260
5
40
290
26
70
150
19.5
0.9
310
0.05
0.02
270
0.0035
200
130
700

3262
H. Hori et al. / Bioorg. Med. Chem. 10 (2002) 3257–3265
Table 4. Cytotoxicities of TX-1123 and tyrphostin AG17 and their
derivatives and analogues for rat hepatocytes and tumor cells
HepG2 cells) similar to TX-1123.As expected, TX-1925
has lower mitochondrial toxicity [uncoupling activity
(Cmax=40 mM; Vmax=290natomO/mg/min),
ATP-
Compounds
IC50 (mM)
HCT116
synthesis inhibitory activities (IC50= 26 mM)], and
hepatotoxicity (IC50=230 mM) than does TX-1123.
These data suggest that TX-1123 and TX-1925 are more
useful candidates of PTK-targeted antitumor agent hav-
ing low mitochondrial toxicity than tyrphostin AG17.
We conclude from these results that PTK inhibitors TX-
1123 and TX-1925 are more promising candidates for
antitumor agents than tyrophsotin AG17.
Rat hepatocytes
HepG2
TX-1123
TX-1925
TX-1918
TX-1926
Tyrphostin AG17
TX-1927
57
230
90
120
0.97
105
39
89
230
4.2
0.20
22
3.66
6.36
2.07
8.05
34.4
34.1
Experimental
General procedures
All chemicals and drugs were of high purity grade and
commercially available, purchased from either Wako
Pure Chemical Industries, Ltd.(Osaka, Japan) or
Sigma-Aldrich Japan (Tokyo, Japan).Oxygen con-
sumption and pH were monitored with a YTS multi-ion
monitor (Yamashita-Giken Ltd., Tokushima, Japan),
that included a Clark oxygen electrode and a pH elec-
trode.The n-octanol–water partition coefficient (logP)
was measured according to the method of Fujita et al.31
The calculated hydrophobic parameters such as dGW
(solvation free energy)32 and logP(calcd) of the com-
pounds were estimated using MOPAC (COSMO
method)33 and PrologP 5.1 (logP Prediction Module of
Pallas 3.0, CompuDrug International Inc., CA, USA),
respectively.All melting points were determined by
Yanagimoto micro melting point apparatus without
correction.IR spectra were recorded in KBr pellets on a
Perkin-Elmer 1600 spectrometer.Ultraviolet–visible
(UV–vis) absorption spectra were determined in ethanol
on a Hitachi U-2000 spectrophotometer.The acid dis-
sociation constant (pKa) was measured spectro-
photometrically at 25 ꢁC from the values of absorbance
at a specified wavelength measured at different pH
Figure 5. Dose–response curves of tyrphostin AG17 and TX-1123 on
Src-K activity.
Conclusion
We designed, synthesized, and evaluated the antitumor
activities of 2-hydroxyarylidene-4-cyclopentene-1,3-
diones as PTK inhibitors having low mitochondrial toxi-
city.TX-1918 was the most potent eEF2-K inhibitor
(IC50=0.44 mM) among the compounds tested.TX-
1123 was a more potent Src-K inhibitor (IC50=2.2 mM)
than tyrphostin AG17 (IC50>350 mM).The O-methyl
derivative of TX-1123 (TX-1925) showed a weaker
mitochondrial toxicity, such as uncoupling activity
(Cmax=40 mM; Vmax=290 natomO/mg/min) and ATP-
synthesis inhibitory activity (IC50=26 mM) than did
1
values.Proton nuclear magnetic resonance ( H NMR)
spectra were recorded on
a JEOL JNM-EX400
(400 MHz) spectrometer with tetramethylsilane as the
internal standard, and chemical shifts (ppm) are given in
d
values.Fast atom bombardment mass spectra
(FABMS) and FAB high resolution mass spectra
(FABHRMS) were measured on a JEOL JMS-SX 102A
instrument.Elemental analyses were performed with a
Yanaco CHN Corder (MT-5).Chromatographic
separations were performed on silica gel columns (Kie-
selgel 60, 230–400 mesh, Merck).To obtain LUMO
energy levels (ELUMO) and coefficients, the semi-empiri-
cal molecular orbital (MO) calculations of compounds
TX-1123.Significantly, the
O-methyl derivative of
AG17 (TX-1927) was completely devoid of properties
associated mitochondrial toxicity, such as uncoupling
activity (Cmax=200 mM; Vmax= 130 natomO/mg/min)
and ATP-synthesis inhibitory activities (IC50=700 mM).
The antitumor activity of TX-1123 (IC50=3.66 mM for
HepG2 cells) was stronger than that of AG17
(IC50=34.4 mM for HepG2 cells), while its antitumor
activity (IC50=39 mM for HCT116) was weaker than
that of AG17 (IC50=0.20 mM).AG17 (IC 50=0.97 mM)
also has much higher hepatotoxicity than TX-1123
(IC50=57 mM).TX-1925 has PTK inhibitory activity
(IC50=3.1 mM for Src-K) and antitumor activity
(IC50=89 mM for HCT116 cells; IC50=6.36 mM for
34
were applied by the PM3 method of J.J.P.Stewart
with his program MOPAC 97(Fujitsu WinMOPAC V.3,
Fujitsu, Tokyo, Japan).
Syntheses of drugs
Tyrphostin AG17 (SF6847). A solution of 3,5-di-tert-
butyl-4-hydroxybenzaldehyde (244 mg, 1.0 mmol) and
malononitrile (80 mg, 1.2 mmol) in anhydrous DMF
(3 mL) was refluxed for 10 h.The solvent was distilled

H. Hori et al. / Bioorg. Med. Chem. 10 (2002) 3257–3265
3263
off under reduced pressure and the residue was purified
by column chromatography on silica gel (hexanes/Et2O,
1:1) to give tyrphostin AG17 (SF6847) (233 mg, 83%) as
a yellow solid: mp 131–134 ꢁC. 1H NMR (CDCl3,
400 MHz): d 7.81 (s, 2H), 7.65 (s, 1H), 6.06 (s, 1H), 1.47
(s, 18H).EI–MS: m/z 282 (M+).Anal.calcd for
C18H22N2O(Mr 282.4): C, 76.56; H, 7.85; N, 9.92.
Found: C, 76.34; H, 7.66; N, 9.63. logP: 5.02 (calcd),
2.32 (obsd). dGW: ꢀ62.4 KJ. pKa: 6.87 (ref36, 6.80)
UV–vis max: 464 nm (10,031), 364 nm (11,101).
Synthesis
methylene}-4-cyclopentene-1,3-dione
of
2-{(3,5-di-tert-butyl-4-methoxyphenyl)
(TX-1925).
A
solution of 3,5-di-tert-butyl-4-methoxybenzaldehyde
(1.24 g, 5.00 mmol), 4-cyclopentene-1,3-dione (603 mg,
6.27 mmol) and p-toluenesulphonic acid (311 mg,
1.63 mmol) in anhydrous CHCl3 (30 mL) was refluxed
for 48 h.The reaction mixture was chromatographed on
silica gel with CH2Cl2 to afford TX-1925 (442 mg, 27%)
as a yellow solid: mp 110–112 ꢁC. H NMR (CDCl3,
1
400 MHz): d 1.48 (bs, 18H), 3.74 (s, 3H), 7.23 and 7.27
(AB type, each 1H, J=6 Hz), 7.60 (s, 1H), 8.34 (s, 2H).
IR (KBr) n cmꢀ1: 3432, 2963, 1707, 1683, 1610, 1585,
1413, 1228, 1172, 1117, 1047, 1007, 889, 854.Anal.
calcd for C21H26O3 (Mr 326.4): C, 77.27; H, 8.03.
Found: C, 77.05; H, 7.96. logP: 6.44 (calcd), 2.40 (obsd).
dGW: ꢀ58.7 KJ. UV–vis max: 339 nm (9592).
Synthesis of
methylene}-4-cyclopentene-1,3-dione
2-{(3,5-di-tert-butyl-4-hydroxyphenyl)-
(TX-1123).
A
solution of 3,5-di-tert-butyl-4-hydroxybenzaldehyde
hemihydrate (239.34 mg, 0.98mmol), 4-cyclopentene-1,3-
dione (123.81 mg, 1.29 mmol), p-toluenesulphonic acid
(124.35 mg, 0.72 mmol) and magnesium sulfate (400 mg)
in anhydrous CHCl3 (12 mL) was refluxed for 19.5 h.
The reaction mixture was filtrated and washed with
water.The filtrate was chromatographed on silica gel
with CH2Cl2 to afford TX-1123 (218.4mg, 71%) as a yel-
low solid: mp 155–157 ꢁC. 1H NMR (CDCl3, 400 MHz): d
1.50 (bs, 18H), 5.93 (s, 1H), 7.18 and 7.22 (AB type,
each 1H, J=6.4 Hz), 7.58 (s, 1H), 8.34 (s, 2H). IR (KBr)
n cmꢀ1 : 3629, 3581, 3431, 2959, 1683, 1623, 1587, 1435,
1176, 1135, 1046, 846.Anal.calcd for C 20H24O3 (Mr
312.4): C,76.89; H, 7.74. Found: C, 76.60; H, 7.50. logP:
4.93(calcd), 2.33 (obsd). dGW: ꢀ60.4 KJ. pKa: 7.40.
UV–vis max 473 nm (6181), 373 nm (12,230).
Synthesis of 2-{(3,5-di-tert-butyl-4-hydroxyphenyl)methyl-
ene}-1,3-indanedione (TX-1926). A solution of 3,5-di-tert-
butyl-4-methoxybenzaldehyde (230 mg, 0.945 mmol),
1,3-indandione (160 mg, 1.095 mmol) and p-toluenesul-
phonic acid (110 mg, 0.58 mmol) in CHCl3 (24 mL) was
refluxed for 21 h.After evaporation of solvents, the
reaction mixture was chromatographed on silica gel
with AcOEt/hexane (1:1) to afford TX-1926 (362 mg)
quantitatively: mp 154–156 ꢁC. 1H NMR (CDCl3,
400 MHz): d 1.53 (bs, 18H), 6.06 (s, 1H), 7.26 (s, 1H),
7.76–7.78 (m, 4H), 8.52 (s, 2H). IR (KBr) n cmꢀ1: 3521,
2952, 1723, 1685, 1557, 1433, 1353, 1279, 1199, 1093,
1036, 1004, 948, 883, 812, 780, 738, 651, 604.Anal.
calcd for C25H28O3 (Mr 362.5): C, 79.53; H, 7.23.
Found: C, 79.43; H, 7.40. logP: 6.30 (calcd), 2.39 (obsd).
dGW: ꢀ59.4KJ. pKa: 7.85. UV–vis max: 508nm (15,157).
Synthesis of 2-{(3,5-dimethyl-4-hydroxyphenyl)methylene}-
4-cyclopentene-1,3-dione (TX-1918). A solution of 3,5-
dimethyl-4-hydroxybenzaldehyde (324.5 mg, 2.16mmol),
4-cyclopentene-1,3-dione (210.3 mg, 2.19 mmol) and
p-toluenesulphonic acid (64.82 mg, 0.34 mmol) in anhy-
drous CHCl3 (25 mL) was refluxed for 24 h.The reaction
mixture was washed with water and chromatographed
on silica gel with CH2Cl2/AcOEt/n-hexane (1:1:2) to
afford TX-1918 (166.9 mg, 33.8%) as a yellow solid:
mp 186–188 ꢁC. 1H NMR (CDCl3, 400 MHz): d 2.32
(bs, 6H), 5.32 (s, 1H), 7.19 and 7.23 (AB type, each 1H,
Synthesis of 2-{(3,5-di-tert-butyl-4-methoxyphenyl)me-
thylene}-malononitrile (TX-1927).35 Yellow solid: mp
104–105 ꢁC. H NMR (CDCl3, 400 MHz): d 1.45 (bs,
1
18H), 3.75 (s, 3H), 7.71 (s, 1H), 7.84 (s, 2H). IR (KBr)
n cmꢀ1: 2964, 2226, 1699, 1589, 1557, 1461, 1442, 1408,
1398, 1293, 1266, 1211, 1118, 1003, 927, 891, 632, 619.
EI–MS (m/z): 296 (M+), 281 [M+ꢀ15(CH3)].Anal.
J=5. 6 and 6. 4 Hz), 7. 51 (s, 1H), 8. 10 (s, 2H). IR (KBr) calcd for C19H24N2O(Mr 296.4): C, 76.99; H, 8.16; N,
n cmꢀ1: 3237, 2955, 2840, 1664, 1595, 1489, 1435, 1388,
1347, 1320, 1253, 1209, 1136, 1029, 970, 947, 870, 791, 689.
Anal.calcd for C 14H12O3(Mr 228.3): C,73.67; H, 5.30.
Found: C, 73.44; H, 5.60. logP: 3.35 (calcd), 2.12 (obsd).
dGW: ꢀ69.7 KJ. pKa: 7.30. UV–vis max: 374 nm (13,700).
9.45. Found: C, 76.73; H, 8.14; N, 9.22. logP: 5.52
(calcd), 2.83 (obsd). dGW: ꢀ60.6 KJ. UV–vis max:
336 nm (9979), 278 nm (24,107).
Protein kinase assay
Synthesis of 3,5-di-tert-butyl-4-methoxybenzaldehyde.
To a suspension of 3,5-di-tert-butyl-4-hydroxybenz-
aldehyde hemihydrate (500 mg, 2.06 mmol) and potassium
carbonate (560 mg, 4.05 mmol) in anhydrous acetone
(20 mL), was added methyliodide (2.28 g, 16.1 mmol).
After being refluxed for 15 h, the reaction mixture was
filtered and washed with water.The filtrate was chro-
matographed on silica gel with CH2Cl2 to afford 3,5-di-
tert-butyl-4-methoxybenzaldehyde (164.3 mg, 32.1%) as
a white solid: mp 59 ꢁC. 1H NMR (CDCl3, 400 MHz): d
1.46 (s, 18H), 3.74 (s, 3H), 7.79 (s, 2H), 9.91 (s, 1H). IR
(KBr) n cmꢀ1: 3386, 2967, 2871, 2719, 2557, 1994, 1697,
1594, 1466, 1380, 1286, 1226, 1196, 1113, 1008, 964,
888, 802, 677, 584.Anal.calcd for C 16H24O2(Mr 236.2):
C, 77.38H, 9.74. Found: C, 77.19; H, 9.54.
Materials. [g-32P]ATP was purchased from NEN
Research Products (Wilmington, DE, USA).
Multiple protein kinase assay36
Assay for PKA, PKC, Src-K, eEF2-K. v-src Trans-
formed NIH3T3 cells were incubated for 10 min on ice
in a hypotonic buffer (1 mM Hepes, pH7, 4.5 mM
MgCl2, and 25 mg/mL each of the protease inhibitors
antipain, leupeptin and pepstatin A).The swollen cells
were homogenized by vortexing for 2 min at room tem-
perature.Following an increase of HEPES to 20 mM,
the homogenate was centrifuged at 500 g for 5 min to
sediment nuclei.To this supernatant, addition was made
to give a final concentration of 20 mM HEPES, pH 7.4,

3264
H. Hori et al. / Bioorg. Med. Chem. 10 (2002) 3257–3265
10mM MgCl2, 0. 1 mM NaVO4, 10mM ß-glycerophos-
3
Laboratories (Kumamoto, Japan).Optical density (OD)
was measured on a microplate reader (BioRad model
450, Japan Bio-Rad Laboratories, Tokyo, Japan) using
a 570 nm filter with blanking at 700 nm.The MTT assay
was performed by suspending the cells in Eagle’s MEM
containing 7% NaHCO3, and 4.7% FCS (pH 7.4), then
pouring 300 mL into the wells of a 48-well culture plate.
Drugs in ethanol (2.5 mL) were added and the cells were
incubated at 37 ꢁC for 48 h.Thereafter, drugs were
washed out with culture medium.The culture medium
(270 mL) and 30 mL of the MTT reagent (MTT 5 mg/mL
in phosphate buffered saline without potassium and
magnesium ions) were added and the cells were incu-
bated for 3 h at 37 ꢁC.Formazan was extracted with
300 mL of 0.04 N HCl in isopropanol and the OD was
measured at a wavelength of 570 nm using a microplate
reader (Bio-Rad, Model 550).From the surviving frac-
tion (OD% control) as a function of drug concentra-
tion, the IC50 value was estimated as the index of the
inhibition of drug to tumor cell growth.
phate, 1 mM NaF and 2.5 mg/mL protein, 1 mM phorbol
myristate 13-acetate (PMA) and 20 mM cAMP.Protein
kinase inhibitors were dissolved in DMSO and the final
concentration of DMSO in the reaction mixture was
10% (v/v).The kinase reaction was initiated by addition
of [g-32P]ATP (12.5 mM, 10 mCi), and the solution was
incubated for 15 min at 25 ꢁC.The reaction was termi-
nated by the addition of SDS-PAGE sample buffer.The
phosphorylated proteins were separated by SDS-PAGE
(9% w/v gels).To detect protein tyrosine kinase (PTK)
activity, the dried gels were further treated with 1 N
KOH at 55 ꢁC for 2 h.The results were visualized by
autoradiography or Fuji Film Bio-image Analyzer BAS
2000.From the enzyme activity (% of control) as a
function of drug concentration, the IC50value was esti-
mated as the index of the enzyme inhibition of drug.
Assay for EGFR-K.37 To measure EGF-K activity,
A431 human epithelial carcinoma cells were solubilized
and cell debris was removed.Cell lysate was incubated
with or without EGF at 25ꢁC for 30min, and the reaction
was started by addition of [g-32P]ATP and incubated at
0 ꢁC for 10 min.After the reaction was stopped, the
mixture was washed.To estimate phosphorylation of
the receptor, the 32P radioactivity was measured with a
scintillation counter.
Hepatotoxicity assay40
Preparation of primary cultures of rat hepatocytes.
Hepatocytes for primary culture were prepared from
male rats of the SD strain (body weight: 150 g, 6 weeks
old) by a collagenase perfusion method.Hepatocytes
were isolated and then inoculated, at a density of 1.31 ꢂ
105 cells/cm2 in a 12-well tissue culture polystyrene
plate.
Mitochondrial toxicity
Uncoupling activity. Mitochondria were isolated from
adult male Wistar rats as reported by Myers and Slater.38
The amount of mitochondrial protein was determined
by the Biuret method with bovine serum albumin as the
standard.The respiration of mitochondria was monitored
polarographically with a Clark-type oxygen electrode
(Yellow Spring, YSI 5331).The incubation medium
consisted of 200 mM sucrose and 2 mM MgCl2 in
10 mM potassium phosphate buffer (pH 7.4). Mitochon-
dria were added at 0.7 mg protein/mL in a total volume
of 2.53 mL succinate (10 mM, plus rotenone at 1 mg/mg
protein) was the respiratory substrate.The activity was
determined by measuring changes in the rate of state 4
respiration upon addition of the test compound.
MTT reduction assay in primary cultures of hepatocytes.
After 2 h, the cells were rinsed three times with a Wil-
liam’s E medium, the medium with 10% fetal bovine
serum and drugs in EtOH (2.5 mL) were added and cul-
ꢁ
tured at 37 C under 5% CO2 atmosphere.After 24 h,
the medium was removed, then 490 mL of medium and
10 mL MTT (5 mg/mL) in phosphate buffered saline
(PBS) were added to each well and the plates were
ꢁ
incubated at 37 C for 3 h to produce formazan.The
medium was then removed from the culture wells, and
the cells were gently rinsed three times with Dulbecco’s
PBS.Then 100 mL 0.04 N HCl in isopropanol was added
to each well to extract the formazan in the cells.The
fromazan extracts were transferred to a 96-well plate
and absorbance was measured at 570 nm in a microplate
reader.
ATP synthesis inhibition. ATP synthesis in mitochon-
dria was determined as reported by Nishimura et al.39
by measuring the increase in pH of the medium asso-
ciated with ATP synthesis at 25 ꢁC in medium consisting
of 200 mM sucrose, 20 mM KCl, 3 mM MgCl2, and
3 mM potassium phosphate buffer, pH 7.4, in a total
volume of 1.68 mL. The reaction was started by adding
400 mM ADP to the mitochondrial suspension (0.7 mg
protein/mL) energized with 5 mM succinate (plus rote-
none at 1 mg/mg protein).The pH change was mon-
itored using a pH meter (Horiba, Kyoto, Japan).
Acknowledgements
The authors thank Mrs.M.Nakamura, Mrs.E.
Okayama, Mrs.K.Yamashita of our faculty, Dr.H.
Satake, Center for Cooperative Research, and Dr.H.
Terada, Dr.Y.Shinohara, and Mrs.Y.Yoshioka of the
Faculty of Pharmaceutical Sciences of our University,
for performing the spectral measurements.We also
thank the Screening Committee of New Anticancer
Agents supported by Grant-in-Aid for Scientific
Research on Priority Area ‘Cancer’ from the Ministry
of Education, Science, Sports, Culture and Technology,
Japan for the protein kinase assay.We also thank Dr.
Soko Kasai, Saharuddin B.Mohamad, Mitsutoshi
Tumor cell growthassay using MTT
The HCT116 human colon tumor cells or human hepa-
toma Hep G2 cells were maintained at exponential
growth.MTT [3-(4,5-dimethyithiazol-2-yl)-2,5-diphenyl-
2H-tetrazolium bromide] was purchased from Dojindo

H. Hori et al. / Bioorg. Med. Chem. 10 (2002) 3257–3265
3265
Sakakibara, Yutaka Shimizu, Mie Masui, and Keiko
Katsuura, for their assistance in these investigations.
19.Ohkura, K;.Hori, H. Bioorg. Med. Chem. 2001, 9, 3023.
20. Burger, A. M.; Kaur, G.; Alley, M. C.; Supko, J. G.;
Malspeis, L.; Grever, M. R.; Sausville, E. A. Cancer Res. 1995,
55, 2794.
21. Lafeuillade, A.; Hittinger, G.; Chadapaud, S. Lancet 2001,
357, 280.
References and Notes
22. Palumbo, G. A.; Yarom, N.; Gazit, A.; Sandalon, Z.;
Baniyash, M.; Kleinberger-Doron, N.; Levitzki, A.; Ben-
Yehuda, D. Cancer Res. 1997, 57, 2434.
23.Sausville, E.A. Cancer Res. 1997, 57, 5610.
24.Levitzki, A;.Ben-Yehuda, D. Cancer Res. 1997, 57, 5610.
1.Ullrich, A;.Schlessinger, J. Cell 1990, 61, 203.
2. Richardson, J. M.; Morla, A. O.; Wang, J. Y. Cancer Res.
1987, 47, 4066.
3.Levitzki, A. FABES J. 1992, 6, 3275.
4.Blume-Jensen, P;.Hunter, T. Nature 2001, 411, 355.
25.Muraoka, S;. Terada, H.
275, 271.
Biochim. Biophys. Acta 1972,
5.Hubbard, SR. ;.Till, JH. .
Annu. Rev. Biochem. 2000, 69, 373.
6.Levitzki, A;.Gazit, A. Science 1995, 267, 1782.
7. Umezawa, H.; Imoto, M.; Sawa, T.; Isshiki, K.; Matsuda,
N.; Uchida, T.; Iinuma, H.; Hamada, M.; Takeuchi, T.
J. Antibiot. (Tokyo) 1986, 39, 170.
8. Yaish, P.; Gazit, A.; Gilon, C.; Levitzki, A. Science 1988,
242, 933.
9.Levitzki, A. FASEB J. 1992, 6, 3275.
26. Inayama, S.; Mamoto, K.; Shibata, T.; Hirose, T. J. Med.
Chem. 1976, 19, 433.
27. Hori, H.; Maezawa, H.; Iitaka, Y.; Ohsaka, T.; Shibata,
T.; Mori, T.; Inayama, S. Jpn. J. Cancer Res. (Gann) 1987, 78,
1128.
28. Koike, H.; Hori, H.; Inayama, S. Biochem. Biophys. Res.
Comm. 1988, 155, 1066.
10. Toi, M.; Mukaida, H.; Wada, T.; Hirabayashi, N.; Toge,
T.; Hori, T.; Umezawa, K. Eur. J. Cancer 1990, 26, 722.
11. Yoneda, T.; Lyall, R.; Asline, M. M.; Persons, P. E.;
Spada, A. P.; Leviszki, A.; Zilberstein, A.; Mundy, G. R.
Cancer Res. 1991, 51, 4430.
29. Hori, H.; Fujii, T.; Kubo, A.; Pan, N.; Sako, S.; Tada, C.;
Matsubara, T. Anal. Lett. 1994, 27, 1109.
30.Teague, S.J. Angew. Chem., Int. Ed. 1999, 38, 738.
31. Fujita, T.; Iwasa, J.; Hansch, C. J. Am. Chem. Soc. 1964,
86, 5175.
12. Gazit, A.; Yaish, P.; Gilon, C.; Levitzki, A. J. Med. Chem.
1989, 32, 2344.
13.Levitzki, A. Biochem. Pharmacol. 1990, 40, 913.
14. Supko, J. G.; Malspeis, L.; Sausville, E. A.; Burger, A.;
Alley, M.; Grever, M. R. Proc. Am. Assoc. Cancer Res. 1994,
35, 424.
32.Ohkura, K;.Hori, H. Bioorg. Med. Chem. 1999, 7, 309.
33. Lambardo, F.; Blanke, J. f.; Curatolo, W. J. J. Med.
Chem. 1996, 39, 4750.
34.Stewart, J.J.P.J.
Comput. Chem. 1989, 10, 209.
35. Terada, H.; Fukui, Y.; Shinohara, Y.; Ju-ichi, M. Bio-
chim. Biophys. Acta 1988, 933, 193.
15. Leviztki, A.; Gazit, A.; Osherov, N.; Posner, I.; Gilon, C.
Methods Enzymol. 1991, 201, 347.
36. Fukazawa, H.; Li, P. M.; Mizuno, S.; Uehara, Y. Anal.
Biochem. 1993, 212, 106.
16. Bilder, G. E.; Krawiec, J. A.; McVety, K.; Gazit, A.;
Gilon, C.; Lyall, R.; Zilberstein, A.; Levitzki, A.; Perrone,
M.H;. Schreiber, A.B. Am. J. Physiol. 1991, 260, C721.
17. Gillespie, J.; Dye, J. F.; Schachter, M.; Guillou, P. J. Br. J.
Cancer 1993, 68, 1122.
37. Murakami, Y.; Fukazawa, H.; Mizuno, S.; Uehara, Y.
Biochem. J. 1994, 301, 57.
38.Myers, D.K;.Slater, E.C.
Biochem. J. 1957, 67, 558.
39. Nishimura, M.; Ito, T.; Chance, B. Biochim. Biophys. Acta
1962, 59, 177.
18.Lander, H.M;. Levine, D.M;. Novogrodsky, A.
Immunol. 1992, 145, 146.
Cell.
40. Fujii, T.; Ha, H.; Yokoyama, H.; Hamamoto, H.; Yoon,
S.H;. Hori, H. Biol. Pharm. Bull. 1995, 18, 1446.