Discovery of a new family of bis-8-hydroxyquinoline substituted benzylamines with pro-apoptotic activity in cancer cells: Synthesis, structure-activity relationship, and action mechanism studies

Article abstract of DOI:10.1016/j.ejmech.2008.03.042

Bis-8-hydroxyquinoline substituted benzylamines have been synthesized and screened for their antitumor activity on KB3 cell line model. Synthesis of this series of new analogues was accomplished using a one pot specific methodology which allows the synthesis of both bis- and mono-8-hydroxyquinoline substituted benzylamines. Among the synthesized compounds two compounds (4a and 5a), respectively, named JLK 1472 and JLK 1486, were particularly potent on KB3 cell line. Their CC₅₀ values being, respectively, 2.6 and 1.3 nM. Screened on a panel of cell lines showing various phenotype alterations, both compounds were found inactive on some cell lines such as PC3 (prostate cell line) and SF268 (neuroblastoma cell line) while highly active on other different cell lines. Mechanistic studies reveal that these two analogues did not affect tubulin and microtubules neither they exert a proteasomal inhibition effect. In contrast 4a and 5a activate specifically caspase 3/7 and not caspase 8 and 9, suggesting that their biological target should be located upstream from caspase 3/7. Moreover their cytotoxic effect is potentiated by the pro-apoptotic effects of TRAIL.

Full text of DOI:10.1016/j.ejmech.2008.03.042

Available online at www.sciencedirect.com
European Journal of Medicinal Chemistry 44 (2009) 558e567
http://www.elsevier.com/locate/ejmech
Original article
Discovery of a new family of bis-8-hydroxyquinoline substituted
benzylamines with pro-apoptotic activity in cancer cells: Synthesis,
structureeactivity relationship, and action mechanism studies
a
Vincent Moret , Younes Laras , Thierry Cresteil , Genevieve Aubert , Dou Q. Ping ,
a
b
b
c
`
´
c
Chen Di , Magali Barthelemy-Requin , Christophe Beclin , Vincent Peyrot ,
d
d
e
´ ´
Diane Allegro e, Amandine Rolland a, Francesca De Angelis g, Evelina Gatti g, Philippe Pierre g,
Luca Pasquini f, Eleonora Petrucci f, Ugo Testa f, Jean-Louis Kraus a,
*
a Laboratoire de Chimie Biomole´culaire, IBDML-UMR-6216-CNRS, Faculte´ des Sciences Luminy, Universite´ de la Me´diterrane´e,
Parc Scientifique de Luminy, Case 901, 13288 Marseille Cedex 9, France
b Institut de Chimie des Substances Naturelles, CNRS UPR 2301, Avenue de la Terrasse, 91198 Gif sur Yvette, France
c The Prevention Program, Barbara Ann Karmanos Cancer Institute, Department of Pathology, School of Medicine, Wayne State University,
540.1 HWCRC, 4100 John R. Road, Detroit, MI 48201-2013, USA
d IMVT-IBDML, Campus de Luminy, Case 907, 13288 Marseille Cedex 09, France
e FRE CNRS 2737, Faculte´ de Pharmacie, 27 Boulevard Jean Moulin, 13385 Marseille Cedex 5, France
f Department of Hematology, Oncology and Molecular Medicine, Instituto Superiore di Sanita`, 00161 Rome, Italy
g Centre d’immunologie Marseille-Luminy, CNRS-UMR-6102-INSERM 631, Case 906, 13288-Marseille Cedex 9, France
Received 7 January 2008; received in revised form 21 March 2008; accepted 27 March 2008
Available online 10 April 2008
Abstract
Bis-8-hydroxyquinoline substituted benzylamines have been synthesized and screened for their antitumor activity on KB3 cell line model.
Synthesis of this series of new analogues was accomplished using a one pot specific methodology which allows the synthesis of both bis- and
mono-8-hydroxyquinoline substituted benzylamines. Among the synthesized compounds two compounds (4a and 5a), respectively, named JLK
1472 and JLK 1486, were particularly potent on KB3 cell line. Their CC50 values being, respectively, 2.6 and 1.3 nM. Screened on a panel of cell
lines showing various phenotype alterations, both compounds were found inactive on some cell lines such as PC3 (prostate cell line) and SF268
(neuroblastoma cell line) while highly active on other different cell lines. Mechanistic studies reveal that these two analogues did not affect
tubulin and microtubules neither they exert a proteasomal inhibition effect. In contrast 4a and 5a activate specifically caspase 3/7 and not caspase
8 and 9, suggesting that their biological target should be located upstream from caspase 3/7. Moreover their cytotoxic effect is potentiated by the
pro-apoptotic effects of TRAIL.
Ó 2008 Elsevier Masson SAS. All rights reserved.
Keywords: Bis-hydroxyquinoline; Benzylamine; Cytotoxic; Synthesis; KB3 cell line; Apoptosis
1. Introduction
chemical biology approach, or target-based approach, consists
first in the identification of a compound that activates or inhibits
a specific target and then in the search for a phenotype induced
by the compound in vivo. The forward chemical biology ap-
proach, or phenotype-based approach, starts with an outward
physical characteristic or phenotype and ends with the identifi-
cation of the responsible target; in this case, molecules are
screened for ‘‘hits’’ that induce a particular phenotype before
There are generally two different approaches used for anti-
proliferative activity compound hit identification. The reverse
* Corresponding author. Tel.: þ33 491 82 91 41; fax: þ33 491 82 94 16.
E-mail address: kraus@luminy.univ-mrs.fr (J.-L. Kraus).
0223-5234/$ - see front matter Ó 2008 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2008.03.042

V. Moret et al. / European Journal of Medicinal Chemistry 44 (2009) 558e567
559
the cellular target is identified. Both reverse and forward
approaches have yielded clinically useful agents for cancer
management.
which involved the direct condensation of 5-chloromethyl 8-
hydroxyquinoline (2) and various substituted primary benzyl-
amines. This reaction was performed in the presence of
potassium carbonate in acetonitrile and gives a mixture of
mono- and bis-hydroxyquinoline derivatives. The bis-
quinoline compounds (3ae9a) are first isolated from the
mixture using their insolubility in diethyl ether. The mono
substituted compounds are isolated from the remaining mix-
ture using the Boc protection procedure. The Boc protected
analogues (5c, 7c and 8c) is isolated by column chromatog-
raphy and then deprotected to give the corresponding mono-
hydroxyquinoline analogues (3be5b, 7be8b). The whole
compounds including the Boc analogues were fully charac-
terized and submitted to KB3-1 cell line model screening.
For mechanistic and structureeactivity relationship
purposes, we have also synthesized a representative quinoline
analogue (14), which structure differs from the most potent an-
alogue by the absence of the hydroxyl group at the 8-position
of the heterocyclic ring. Its synthesis is depicted in Scheme 2.
Starting from the quinoline-5-carboxylic acid (10) which
was first transformed into its corresponding ethyl ester (11)
and reduced with lithium aluminium hydride, the correspond-
ing alcohol 12 was obtained according to a known procedure
[6]. Chloration of 12 in the presence of thionyl chloride lead
to the corresponding chloromethylquinoline 13 (67% yield).
Condensation of 13 with 4-trifluoromethylbenzylamine in
the presence of potassium carbonate lead to the expected
bis-quinoline 14.
The approach we have used for the discovery of new active
anticancer drugs is the forward chemical biology approach.
Our house chemical library of bioactive molecules was first
screened on KB3-1 (human mouth epidermal carcinoma)
cell lines as primary screening in order to identify analogues
which present cytotoxic activity on this cell line model [1].
From this, we have identified a new series of small molecules,
incorporating in their structure a hydroxyquinoline moiety,
which presented potent cytotoxic effect on KB3-1 cells. In-
deed, this cell line appears to be a representative and suitable
model, since it is known to be very sensitive to drugs and also
because a P-gp induced cell line, is also available [2].
Starting from these preliminary results, in this paper we re-
port the synthesis of a series of compounds based on this hy-
droxyquinoline scaffold, and their in vitro nanomolar range
cytotoxic activity on KB3 and other cell lines from different
histogenetic origin having various phenotype alterations. We
also report the ability of this new family of compounds to in-
duce caspase activation since it is known that caspases play an
important role in apoptosis induced by genotoxic stress, and
that some anticancer drugs trigger mitochondrial outer mem-
brane permeabilization resulting in cytochrome c release and
caspase activation [3,4]. We have also investigated the possi-
ble activity for these new analogues to act as proteasome in-
hibitors as well as promotor of a, b tubulin assembly.
3. Results and discussion
2. Chemistry
3.1. Primary screening: cytotoxic activity
on KB3-1 cell line
Among the tested compounds from our in house library, an-
alogues whose general structure included an 8-hydroxyquino-
line moiety linked to a substituted benzylamine (Fig. 1) were
found cytotoxic for KB3-1 cells at concentrations ranging
around 1 mM.
In order to optimize the cytotoxic effect found for some
members of this family of compounds, we synthesized several
new analogues based on this scaffold, in which specific sub-
stituents R1 and R2 were introduced.
The synthetic scheme which has been developed in order to
elaborate this library is outlined in Scheme 1.
Starting from 8-hydroxyquinoline (1), the required starting
material 5-chloromethyl 8-hydroxyquinoline (2) was obtained
in good yield, using a known procedure [5]. The news ana-
logues have been synthesized, using standard methodologies,
All compounds were evaluated for their cytotoxic effect on
KB3 cell line according to a procedure described in Section 5.
The obtained results are presented in Tables 1 and 2.
From the CC50 values presented in Tables 1 and 2, it ap-
pears that derivatives belonging to the bis-hydroxyquinoline
series (3ae9a) are more active than the mono-hydroxyquino-
line analogues. The CC50 values of the former ranged from 1.3
to 123 nM whereas the cytotoxic activity values of the
corresponding mono-hydroxyquinoline derivatives (3be5b,
7be9b) were found far less active, since their cytotoxic activ-
ity values lie between 1 and 10 mM.
The importance of the substituents (R1) at various positions
of the benzyl moiety was first examined (Table 1). Introduc-
tion of a trifluoromethyl substituent at the 4-position of the
aromatic ring led to the most potent analogue (5a,
CC50 ¼ 1.3 nM). Varying the inductive electronic effect (elec-
tron-withdrawing substituent NO2 (6a) and electron-donating
substituent CH3 (4a)) did not change drastically the potency
of the resulting analogues. Their CC50 values varied only by
one order magnitude, from 1.3 to 42 nM, indicating that the
compounds are still potent cytotoxic drugs. It could also be
observed that the para position, on which CF3 group was in-
troduced on the benzyl group, appears to be the most
R1
R2
OH
N
N
N
OH
Fig. 1. General structure of the new cytotoxic drugs.

560
V. Moret et al. / European Journal of Medicinal Chemistry 44 (2009) 558e567
3a: R1 = phenyl,
R2 = H
R2 = H
R2 = H
R2 = H
R2 = H
R1
R2
OH
N
Cl
4a: R1 = 4-CH3-phenyl,
5a: R1 = 4-CF3-phenyl,
6a: R1 = 4-NO2-phenyl,
7a: R1 = 2-CF3-phenyl,
8a: R1 = 4-CH3-phenyl,
9a: R1 = 1-naphthyl,
N
i
ii
N
R1-CH(R2)-NH2
N
OH
R2 = (S) CH3
R2 = H
N
OH
OH
1
2
+
3b: R1 = phenyl,
R2 = H
R2 = H
R2 = H
R2 = H
R2 = (S) CH3
R2 = H
R1
R2
R1
R2
R1
R2
NH
4b: R1 = 4-CH3-phenyl,
5b: R1 = 4-CF3-phenyl,
7b: R1 = 2-CF3-phenyl,
8b: R1 = 4-CH3-phenyl,
9b: R1 = 1-naphthyl,
NH
N
Boc
iv
iii
N
N
N
OH
OH
OH
5c: R1 = 4-CF3-phenyl, R2 = H
7c: R1 = 2-CF3-phenyl, R2 = H
8c: R1 = 4-CH3-phenyl, R2 = (S) CH3
Scheme 1. Reagents and conditions: (i) HCHO, 37% HCl in H2O, HCl (gas), rt, overnight, 80%; (ii) CHCl3, rt, 24 h; or K2CO3, CH3CN, rt, 24 h; (iii) (tBoc)2O,
EtOAc; (iv) HCl, Et2O, rt, 4 h.
favourable position for optimum cytotoxic activity. Indeed
compound 5a in which CF3 is located at the para position be-
ing two logs more active than its corresponding analogue in
which CF3 group is introduced at the ortho position (7a).
Moreover, introduction of a methyl group on the methylene
of the benzyl group (8a, CC50 ¼ 41 nM) only slightly dimin-
ished the resulting cytotoxic activity compared to its corre-
sponding analogue (4a, CC50 ¼ 2.6 nM). Encouraged by
these results, we next prepared the corresponding analogue
of the most potent compound 5a in which both hydroxy sub-
stituents were removed on the quinoline heterocycles leading
to compound 14. We found that this latter analogue was two
logs less active than its corresponding compound 5a, indicat-
ing that the presence of the hydroxyquinoline moiety was re-
quired for cytotoxic activity.
3.2. Activity of the new drugs on various cancer cell lines
The cytotoxic activity of compounds JLK 1472 and JLK
1486 was further investigated on a panel of human cancer
cell lines and on two non-cancerous cell lines (Table 3).
From these results several observations can be made.
- The prostate cell line PC-3 required higher CC50 for both
drugs than other cell lines.
- SF268 (neuroblastoma cell line) is also poorly affected by
both drugs.
- OV3 are ovarian tumor cell lines.
- A549 are human alveolar epithelial cell lines.
- HepG2 are human hepatocellular liver carcinoma cell
lines.
The whole obtained results from this primary screening led
us to select compounds 4a and 5a, which are, respectively,
named JLK 1472 and JLK 1486, as representative hits which
could be submitted to further mechanistic investigations.
- MCF7R and HCT15 are derived from MCF7 (epithelial
breast cancer cell line) and HCT116 cell lines (colonic
epithelial cell line), respectively, after chemical exposure.
It has been shown that these two cell lines expressed the
Cl
HO
O
OH
O
O
N
N
i
ii
iv
iii
N
N
N
N
N
10
11
12
13
CF3
14
Scheme 2. Reagents and conditions: (i) HCl gas, Et2O, reflux, 20 h; (ii) LiAlH4, Et2O, 0 ꢀC, 1.5 h; (iii) SOCl2, reflux, 2 h, 67%; (iv) 4-trifluoromethylbenzylamine,
K2CO3, DMF, 70 ꢀC, 4 h, 65%.

V. Moret et al. / European Journal of Medicinal Chemistry 44 (2009) 558e567
561
Table 1
Cytotoxic effect on KB3-1 cell lines of bis-hydroxyquinoline compounds
Table 3
In vitro cell growth inhibition by analogues JLK 1472 and JLK 1486 in a panel
of cell lines of different histogenic origin and having various phenotype
alterations
R2
R1
OH
N
Cell lines
Compounds CC50 (nM)
N
JLK 1486
JLK 1472
KB3
1.5 ꢁ 0.1
1.0 ꢁ 0.1
0.8 ꢁ 1.0
1.2 ꢁ 0.3
4.7 ꢁ 1.3
3.0 ꢁ 0.1
206 ꢁ 113
0.9 ꢁ 0.2
91 ꢁ 49
8.5 ꢁ 2.9
3.2 ꢁ 1.2.
6.0 ꢁ 7.3
5.4 ꢁ 0.9
37.3 ꢁ 28
8.0 ꢁ 0.4
343 ꢁ 125
4.3 ꢁ 0.1
178 ꢁ 83
32.7 ꢁ 6.6
40.8 ꢁ 22
52.1 ꢁ 14.5
HCT116
HCT15
MCF7
MCF7R
OV3
N
OH
Compound
R1
R2
CC50 (nM)
PC3
A549
3a
4a (JLK 1472)
Phenyl
H
H
H
H
H
15
2.6
1.3
42
4-CH3ephenyl
4-CF3ephenyl
4-NO2ephenyl
2-CF3ephenyl
4-CH3ephenyl
1-Naphthyl
SF268
HL60
K562
5a (JLK 1486)
10.9 ꢁ 4.9
5.2 ꢁ 8.2
22.8 ꢁ 39.2
6a
7a
8a
9a
123
41
HepG2
(S )CH3
H
5
we investigate if compound JLK 1486 could act as TaxolÒ-like
drug. Indeed, suppression of the cellular microtubule dynam-
ics by TaxolÒ is the main cause of cell division inhibition
and of tumor cell death [7].
Docetaxel
0.25
P-gp protein. The level of expression of P-gp, much higher
in MCF7R than in HCT15 cells, could be associated with
the shift of CC50 values displayed compared to the native
MCF7 and HCT116 cells. However, the variation in the
CC50 is relatively moderate and would suggest that ana-
logues JLK 1472 and JLK 1486 could be possible P-gp sub-
strates but with a low affinity.
In this perspective, we evaluated the ability of JLK 1486 to
modify the in vitro assembly of tubulin by a standard turbidi-
metric method. The tubulin self-assembly into microtubules
in the presence of JLK 1486 is shown (Fig. 2). In our conditions,
when 20 mM of tubulin with 20 mM of JLK 1486 compound
(curve B, dashed line) and without (curve A, solid line) are
incubated at 37 ꢀC the turbidity time courses are similar, dem-
onstrating that JLK 1486 did not inhibit microtubule formation.
Moreover, when the sample in the presence of JLK 1486
compound is cooled down to 4 ꢀC, there is a complete disas-
sembly, indicating a lack of tubulin dimer aggregation or
microtubule stabilization by the ligand. TaxolÒ is known to
promote tubulin dimer assembly by binding to b-tubulin in
microtubules. We checked if JLK 1486 induced microtubule
formation at a tubulin concentration which did not permit mi-
crotubule formation. Curve B showed that JLK 1486 (dashed
line) was unable to induce microtubule formation.
- Leukaemia cells (HL60 and K562) are also sensitive to
both drugs.
3.3. Action mechanism studies
3.3.1. Tubulin and microtubule studies
As far as the observed effects of paclitaxel or TaxolÒ on our
model of KB3 cells were found very potent (CC50 ¼ 0.1 nM),
Table 2
Cytotoxic effect on KB3-1 cell lines of mono-hydroxyquinoline compounds
R1
R2
N
R3
N
OH
Compound
R1
R2
R3
CC50 (mM)
3b
4b
5b
7b
8b
9b
5c
7c
8c
Phenyl
H
H
H
H
H
>1
>10
10
4-CH3ephenyl
4-CF3ephenyl
3-CF3ephenyl
4-CH3ephenyl
1-Naphtyl
H
H
H
0.2
>1
>1
10
(S )CH3
H
H
Fig. 2. Effects of JLK 1486 on the turbidity time course of in vitro microtubule
assembly. The tubulin assembly curve is given as a solid line, whereas tubulin
assembly with JLK 1486 (ratio tubulin: JLK 1486 1:1) is given as a dashed
line. Tubulin concentration was 20 mM for A and 15 mM for B. Assembly is
started by heating the solution from 4 to 37 ꢀC (first arrow) and the disassem-
bly is induced by switching the temperature to 4 ꢀC (second arrow).
H
4-CF3ephenyl
3-CF3ephenyl
4-CH3ephenyl
H
tBoc
tBoc
tBoc
H
1
>1
(S )CH3
Docetaxel
0.25

562
V. Moret et al. / European Journal of Medicinal Chemistry 44 (2009) 558e567
No direct inhibitory concentration effect of the drug on the
For this purpose we explored their capacity to activate cas-
pase by a direct measurement of caspase activities in HL60
cells after a 48 h exposure to these compounds [4]. To assess
the role of JLK 1472 and JLK 1486 on the caspase cascade we
used specific substrates: DEVD for caspases 3/7, IEDT for
caspase 8 and LEHD for caspase 9. Colchicine was used as
a positive reference inducer, and compounds JLK 1472 and
JLK 1486 were tested at three concentrations 1, 0.1 and
0.01 mM. JLK 1472 and JLK 1486 had no effect upon caspase
3/7 activation at the concentration of 0.01 mM, but elicited
a marked effect at 0.1 and 1 mM (Fig. 6). Assayed in parallel
in the same samples, caspase 8 and caspase 9 were not acti-
vated by both compounds even at 1 mM. Thus JLK 1472 and
JLK 1486 activated caspases 3/7 in a dose-dependent
manner but did not induce caspase 8 nor 9 activation.
tubulin assembly was observed, neither a stabilizing effect on
microtubule formation. These results suggest that the new dis-
covered drugs are not inhibitors of microtubule formation
Taxol-like drugs [8].
3.3.2. Proteasomal activity
We next examined whether the most active analogues could
affect the proteasome activity. It has been reported that
hydroxyquinolines are compounds capable of forming stable
complexes with cooper(II) ions [9]. Similarly, clioquinol,
a known copper chelatant, which structure included an 8-
hydroxyquinoline moiety (Fig. 3) used as an antibiotic for
treating diarrhea and skin infection [10], has been identified
as an androgen receptor inhibitor and an inducer of cellular ap-
optosis in relation with its observed antiprostate tumor effect
[11]. Clioquinol, after binding to copper, can inhibit the pro-
teasomal chymotrypsin-like activity and can induce cell death
in human cancer LNCaP and C4-2B cell lines [11].
So it can be hypothesized that JLK 1486 (5a) and JLK 1472
(4a) could exert their antiproliferative activities through an in-
hibition of the cellular proteasome in a copper-dependent
manner. To test this possible mode of action we assessed the
antiproliferative activity of compounds JLK 1486 and JLK
1472 on the HL60 cell line and compare their effect with
that of clioquinol used as a model.
As shown in Fig. 4A, clioquinol alone had no antiprolifer-
ative activity in HL60 cells when incubated for 24 h in the ab-
sence of CuCl2, but elicited a dramatic reduction of cell
proliferation and viability (as shown by the trypan blue exclu-
sion assay) in the presence of CuCl2 (Fig. 4A). This clearly
suggests that JLK 1472 and JLK 1486, unlike clioquinol, acted
through an independent mechanism. Concomitantly, it was ca-
pable to promote the cleavage of PARP in the presence of
CuCl2 whereas JLK 1486 had no or very limited effect on
the processing of PARP (Fig. 4B). Since PARP is a caspase
3 substrate, it did not represent a direct proteasome marker,
we have performed a direct in vitro proteasome inhibitory
test using both chymotrypsin and trypsin sensitive peptides
in the presence of various concentrations of JLK 1486 from
1 to 1 mM. No activity proteasomal inhibitory activity was
found (Fig. 5), establishing definitively the lack of effect of
JLK 1486 on proteasomal activity.
3.3.4. Drug cytotoxic effect potentiation by TRAIL (tumor
necrosis factor related apoptosis inducing ligand) [12]
Since it was shown above that JLK 1486 and JLK 1472 in-
duce caspase 3/7 activation, we evaluated whether their cyto-
toxic effect is potentiated by the death ligand TRAIL. In this
purpose, we have used four cancer cell lines, either scarcely
(M4, A2780/WT and A2780/ADR) or moderately (HL60 and
MCF7) sensitive to the pro-apoptotic effects of TRAIL.
As shown in Fig. 7 it can be seen that the simultaneous addi-
tion of JLK 1486 and TRAIL, even using JLK 1486 at low con-
centrations (10 nM for HL60 cells or 100 nM for A2780, M4
and MCF7cells) that were only moderately cytotoxic, resulted
in a strong apoptotic effect with a high rate of cell death (Fig. 7).
4. Conclusion
From invitro cellular screening on KB3 cell line, we have dis-
covered a new family of potent cytotoxic drugs, which belong to
the class of bis-8-hydroxyquinoline substituted benzylamine,
which have never been described before. From the elaboration
of a small focussed library, two analogues of this family were
found particularly potent on KB3 cell lines (analogue JLK
1472 and JLK 1486); their cytotoxic activity (CC50 ranging
from 1 to 5 nM) was only one order magnitude less potent than
that of the standard clinically used anticancer drug, docetaxel.
Experiments on HL60 cell line using different substrates
have shown that these compounds can activate caspase 3/7 be-
tween 1 and 0.1 mM but not at 0.01 mM. These range of con-
centrations correspond to the concentrations that are able to
induce cytotoxicity in the same cell line, suggesting that cell
death induced by these compounds occurs through a caspase
3/7 apoptosis pathway. This assumption is further supported
by the observation that cytotoxicity in HL60 is further en-
hanced in the presence of TRAIL which is also able to induce
caspase 3/7. Since different pathways may induce apoptosis in
cell, we have investigated the possible involvement of two
pathways known to be targeted by different clinical anticancer
drugs, microtubule assembly and proteasome. However none
of these pathways appeared to be affected by JLK 1472 and
JLK 1486, showing that these two compounds act through
different pro-apoptotic pathways.
3.3.3. Caspase activation
We next investigated whether the cytotoxic activity of JLK
1472 and JLK 1486 on HL60 cell line depends on caspase
induction.
Cl
I
N
OH
Fig. 3. Clioquinol.

V. Moret et al. / European Journal of Medicinal Chemistry 44 (2009) 558e567
563
Fig. 4. Comparative cytotoxicity of clioquinol, JLK 1472 and JLK 1486 in the absence or presence of copper (A) and Western blotting (B). The reduction of cell
viability by JLK 1472 and JLK 1486 on human leukaemia HL60 cells was compared with that of clioquinol (CQ). HL60 cells were treated for 24 h with JLK 1472,
JLK 1486 or CQ (A), or treated with a mixture of JLK 1472, JLK 1486 or CQ with CuCl2 in the indicated concentrations (A), followed by trypan blue assay (A) or
Western blotting (B) at 10 mM drug concentration.
Further investigations are needed in order to identify the bi-
ological target on which these new drugs are acting. Moreover
pharmacokinetic studies should be performed in order to eval-
uate a possible therapeutical interest for these new drugs.
5. Materials and methods
5.1. Chemistry
5.1.1. General procedures
Unless otherwise noted, starting materials and reagents
were obtained from commercial suppliers and were used with-
out purification. Methylene dichloride (CH2Cl2) was distilled
over P2O5 just prior to use. Acetonitrile was of anhydrous
quality from commercial suppliers (Aldrich, Carlo Erba Re-
agents). As far as the products are obtained as a mixture of
multiple compounds (2 or 3), the given yields are the calcu-
lated yields for each compound and not the overall yield of
1
the reaction. H NMR spectra were recorded at 250 MHz on
a Bruker AC-250 spectrometer. Chemical shifts are expressed
as d units (parts per million) downfield from TMS (tetrame-
thylsilane). Electro-spray mass spectra were obtained on a Wa-
ters Micromass ZMD spectrometer by direct injection of the
sample solubilized in acetonitrile. Elemental analyses were
within ꢁ0.4% of theoretical values for all compounds. All re-
actions were monitored by thin-layer chromatography. Analyt-
ical thin-layer chromatographies (TLC) were performed using
silica gel plates of 0.2 mm thick (60F254 Merck). Preparative
flash column chromatographies were carried out on silica gel
Fig. 5. Proteasomal activity of analogues JLK 1472 and JLK 1486.
(230e400 mesh, G60 Merck).

564
V. Moret et al. / European Journal of Medicinal Chemistry 44 (2009) 558e567
ethyl acetate (10 ml) is stirred at 50 ꢀC overnight. It is cooled
to 0 ꢀC and filtered; the filter cake is washed with cold ethyl ac-
etate (5 ml). The filtrate is concentrated in vacuo, diluted in di-
ethyl ether (5 ml) and centrifuged. The etheral phase is removed
and the obtained solid is washed two more times by centrifuga-
tion with diethyl ether at 0 ꢀC to give 168 mg of 3a (30% yield)
1
as a green powder; H NMR (CDCl3) d 3.66 (s, 2H), 3.99 (s,
4H), 7.20e7.35 (m, 6H), 7.42e7.45 (m, 3H), 7.55e7.58 (m,
2H), 8.03e8.07 (m, 2H), 8.89 (s, 2H); 13C NMR (CDCl3)
d 56.4, 58.5, 108.8, 120.8, 125.2, 127.2, 127.6, 128.1, 129.7,
129.8, 134.3, 138.6, 139.0, 147.5, 151.8; ESI-MS m/z ¼ 422
([M þ H]þ, 100%). Combined ethereal phases are concentrated
in vacuo. The obtained product is dissolved in dichloromethane
(4 ml) and Boc2O (171 mg, 0.786 mmol) is added. The solution
is stirred overnight at room temperature then washed with water
(4 ml) and brine (2 ml). It is dried over MgSO4 then concen-
trated in vacuo and purified by chromatography eluting with
cyclohexaneedichloromethane (1:1). The obtained product is
dissolved in diethyl ether (2 ml) and hydrochloric acid in
diethyl ether (1 M, 2 ml) is added. The solution is centrifuged
to give 70 mg 3b as a white solid (21% yield); 1H NMR
(MeOH-d4) d 4.2 (s, 2H), 4.6 (s, 2H), 7.2e8.0 (m, 8H), 8.9 (s,
2H); 13C NMR (MeOH-d4) d 41.2, 43.6, 49.1, 113.9, 116.7,
120.8, 127.4, 128.1, 130.6, 132.7, 140.9, 141.1, 146; ESI-MS
m/z ¼ 265 (free base, [M þ H]þ, 100%).
5.1.3. 5,50-(4-Methylbenzylazanediyl)bis(methylene)
diquinolin-8-ol (JLK 1472, 4a) and 5-
((4-methylbenzylamino)methyl)quinolin-8-ol
dihydrochloride (4b)
Thecompoundsweresynthesizedfrom4-methylbenzylamine
following the procedure described for 3a and 3b. Compound 4a:
yield 22%; 1H NMR (CDCl3) d 2.33 (s, 3H), 3.44 (s, 2H), 3.81 (s,
4H), 6.95e7.17 (m, 8H), 7.36e7.45 (m, 2H), 7.88e7.98(m, 2H),
8.68e8.76 (m, 2H); 13C NMR (CDCl3) d 20.9, 56.1, 58.0, 108.4,
120.6, 125.1, 127.3, 128.5, 129.4, 129.4, 134.1, 135.6, 136.5,
138.3, 147.2, 151.5; ESI-MS m/z ¼ 436 ([M þ H]þ, 100%).
1
Compound 4b: yield 16%; H NMR (MeOD-d4) d 2.2 (s, 3H),
4.3 (s, 2H), 4.7 (s, 2H), 7.2e9.6 (m, 9H); 13C NMR (MeOD-
d4) d 21.8, 48.5, 48.8, 49.1, 49.8, 50.2, 50.5, 117.1, 124.7,
129.7, 131.4, 132.0, 137.0, 141.4, 145.5, 151.5; ESI-MS m/z ¼
279 (free base, [M þ H]þ, 100%).
5.1.4. 5,50-(4-(Trifluoromethyl)benzylazanediyl)
bis(methylene)diquinolin-8-ol (JLK 1486, 5a), 5-
((4-(trifluoromethyl)benzylamino)methyl)quinolin-
8-ol (5b) and tert-butyl (8-hydroxyquinolin-5-yl)
methyl(4-(trifluoromethyl)benzyl)carbamate (5c)
A solution of 5-chloromethylquinolin-8-ol dihydrochloride
(300 mg, 1.3 mmol) and 4-(trifluoromethyl)benzylamine
(683 mg, 3.9 mmol) in ethyl acetate (10 ml) is stirred at 50 ꢀC
overnight. It is cooled to 0 ꢀC and filtered; the filter cake is
washed with cold ethyl acetate (5 ml). The filtrate is concen-
trated in vacuo, diluted in diethyl ether (5 ml) and centrifuged.
The ethereal phase is removed and the obtained solid is washed
two more times by centrifugation with diethyl ether at 0 ꢀC to
Fig. 6. Caspase activation by analogues JLK 1472 and JLK 1486.
5.1.2. 5,50-(Benzylazanediyl)bis(methylene)diquinolin-
8-ol (3a) and 5-((benzylamino)methyl)quinolin-
8-ol dihydrochloride (3b)
A solution of 5-chloromethylquinolin-8-ol dihydrochloride
(300 mg, 1.3 mmol) and benzylamine (419 mg, 3.91 mmol) in
1
give 5a (32% yield); H NMR (CDCl3) d 3.55 (s, 2H), 3.84

V. Moret et al. / European Journal of Medicinal Chemistry 44 (2009) 558e567
565
Fig. 7. JLK 1486 markedly potentiates the cytotoxic effect of TRAIL on various cancer cell lines. (A) HL60 cell has been incubated for 24 h in the presence of
increasing concentrations of JLK 1486 added alone or in combination with recombinant human TRAIL (50 ng/ml), the percentage of apoptotic cells was deter-
mined by the Annexin V binding assay. Mean ꢁ SEM are observed in three separate experiments. (B) Melanoma M4 and breast carcinoma MCF7 cells have been
grown for 24 h either in the absence (C) or in the presence of JLK 1486 (100 nM or 1 mM) added alone or in combination with TRAIL (50 ng/ml) or TRAIL at the
above concentration and the proportion of apoptotic cells was evaluated as above. Mean values ꢁ SEM are observed in three separate experiments. (C) Ovarian
cancer A2780WT and its chemoresistant variant (A2780ADR) cell lines have been grown for various times (0e3 days) as reported in (B). The proportion of ap-
optotic cells was evaluated as above.
(s, 4H), 7.07e7.12 (m, 2H), 7.17 (m, 2H), 7.41 (m, 2H), 7.51
(m, 4H), 7.86 (m, 2H), 8.73 (m, 2H); 13C NMR (CDCl3)
d 30.9, 56.6, 58.1, 76.5, 77.0, 77.5, 108.7, 121.0, 124.7,
127.4, 129.8, 129.9, 133.9, 138.6, 147.6, 152.0; ESI-MS m/z ¼
490 ([M þ H]þ, 100%). Combined etheral phases are concen-
trated in vacuo. The obtained product is dissolved in dichloro-
methane (4 ml) and Boc2O (171 mg, 0.786 mmol) is added.
The solution is stirred overnight at room temperature then
washed with water (4 ml) and brine (2 ml). It is dried over
MgSO4 then concentrated in vacuo and purified by chromatog-
raphy eluting with cyclohexaneedichloromethane (1:1) to give
5b (15% yield) as a white solid; 1H NMR (DMSO-d6) d 4.40 (s,
2H), 4.67 (s, 2H), 7.44 (m, 1H), 7.82 (m, 3H), 7.95 (m, 2H), 9.07
(s, 1H), 9.24 (m, 1H), 9.93 (m, 1H); ESI-MS m/z ¼ 333
([M þ H]þ, 100%). Compound 5c: yield 19%; 1H NMR
(DMSO-d6) d 1.49 (s, 9H), 4.28 (s, 2H), 4.71 (s, 2H), 7.08 (m,
1H), 7.19 (m, 3H), 7.52 (m, 3H), 8.60 (m, 1H), 8.79 (m, 1H);
ESI-MS m/z ¼ 433 ([M þ H]þ, 100%).
5.1.5. 5,50-(4-Nitrobenzylazanediyl)
bis(methylene)diquinolin-8-ol (6a)
A solution of 5-chloromethylquinolin-8-ol dihydrochloride
(300 mg, 1.3 mmol), 4-nitrobenzylamine (123 mg, 0.65 mmol),
and potassium carbonate (361 mg, 2.6 mmol) in dimethylfor-
mamide (20 ml) is stirred at 50 ꢀC for 4 h. It is cooled to 0 ꢀC
and filtered; the filtered cake is washed with ethyl acetate
(40 ml). The filtrate is washed with water (two times with
40 ml then six times with 20 ml), then the organic phase is
dried over magnesium sulfate. It is then concentrated in vacuo,
dissolved in diethyl ether (10 ml) and centrifuged. The

566
V. Moret et al. / European Journal of Medicinal Chemistry 44 (2009) 558e567
ethereal phase is removed and the obtained solid is washed
two more times by centrifugation with diethyl ether at 0 ꢀC
to give 160 mg of 6a (56% yield) as a green powder; 1H
NMR (CDCl3) d 3.5 (s, 2H), 3.9 (s, 4H), 7.1e8.8 (m, 14H);
13C NMR (CDCl3) d 56.8, 57.9, 108.9, 121.1, 123.2, 124.4,
127.4, 130.0, 130.1, 133.6, 138.6, 147.0, 147.1, 147.6,
152.1; ESI-MS m/z ¼ 467 ([M þ H]þ, 100%).
4.8 (s, 2H), 7.4e8.2 (m, 12H); ESI-MS m/z ¼ 315 (free
base, [M þ H]þ, 100%).
5.1.10. Ethyl quinoline-5-carboxylate (11)
It was synthesized from quinoline-5-carboxylic acid and
the procedure has been described in Ref. [6]. Compound 11:
1
yield 70%; H NMR (CDCl3) d 1.44 (t, 3H), 4.45 (q, 2H),
7.49 (m, 1H), 7.72 (m, 1H), 8.27 (m, 2H), 8.93 (m, 1H),
9.32 (m, 1H); 13C NMR (MeOH-d4) d 14.3, 61.2, 122.4,
127.0, 127.1, 128.0, 130.6, 134.4, 134.7, 148.1, 150.5,
166.5; ESI-MS m/z ¼ 202 ([M þ H]þ, 100).
5.1.6. 5,50-(2-(Trifluoromethyl)benzylazanediyl)
bis(methylene)diquinolin-8-ol (7a)
It was synthesized from 2-(trifluoromethyl)benzylamine
following the procedure described for 3a (33% yield); 1H
NMR (CDCl3) d 3.68 (s, 2H), 3.86 (s, 4H), 6.98 (m, 2H),
7.21e7.31 (m, 3H), 7.45 (m, 3H), 7.62 (m, 1H), 7.96 (m,
2H), 8.81 (m, 2H); ESI-MS m/z ¼ 490 ([M þ H]þ, 100%).
5.1.11. Quinolin-5-ylmethanol (12)
It was synthesized from ethyl quinoline-5-carboxylate and
the procedure has been described in Ref. [6]. Compound 12:
1
yield 62%; H NMR (CDCl3) d 5.13 (d, 2H), 7.43 (m, 1H),
7.55 (m, 1H), 7.65 (m, 1H), 8.03 (m, 1H), 8.50e8.53 (m,
1H), 8.87 (m, 1H); 13C NMR (MeOH-d4) d 63.0, 121.1,
125.8, 126.5, 129.0, 129.7, 132.6, 136.8, 148.4, 150.1; ESI-
MS m/z ¼ 160 ([M þ H]þ, 100).
5.1.7. 5-((3-(Trifluoromethyl)benzylamino)methyl)quinolin-
8-ol (7b) and 5-tert-butyl(8-hydroxyquinolin-5-yl)methyl
(3-(trifluoromethyl)benzyl)carbamate (7c)
The compounds were synthesized from 3-(trifluoromethyl)-
benzylamine following the procedure described for 5b and 5c.
1
Compound 7b: yield 15%; H NMR (DMSO-d6) d 4.49 (s,
2H), 4.71 (s, 2H), 7.14 (m, 1H), 7.61e7.84 (m, 5H), 7.91
(m, 1H), 8.56 (m, 1H), 8.87 (m, 1H); ESI-MS m/z ¼ 333
([M þ H]þ, 100%). Compound 7c: yield 19%; 1H NMR
(CDCl3) d 1.49 (s, 9H), 4.28 (s, 2H), 4.82 (s, 2H), 7.07 (m,
1H), 7.21e7.26 (m, 2H), 7.30e7.37 (m, 3H), 7.42e7.49 (m,
3H), 8.78 (m, 1H); ESI-MS m/z ¼ 433 ([M þ H]þ, 100%).
5.1.12. 5-(Chloromethyl)quinoline (13)
A solution of quinolin-5-ylmethanol (160 mg, 1.0 mmol) in
thionyl chloride (5 ml) is stirred 10 min at room temperature,
then left 2 h at reflux. It is then concentrated in vacuo, dissolved
in ethyl acetate and washed with brine. The organic phase is
concentrated in vacuo to give 120 mg of crude 13 (67% yield)
as a brown solid; 1H NMR (CDCl3) d 5.00 (s, 2H), 7.46e7.67
(m, 3H), 8.10 (m, 1H), 8.48 (m, 1H), 8.95 (m, 1H); 13C NMR
(MeOH-d4) d 43.3, 121.4, 126.3, 128.0, 128.9, 130.9, 132.2,
133.4, 148.4, 150.3; ESI-MS m/z ¼ 178 ([M þ H]þ, 100).
5.1.8. (S )-5,50-(1-Phenylethylazanediyl)bis(methylene)
diquinolin-8-ol (8a), (S )-5-((1-phenylethylamino)methyl)
quinolin-8-ol (8b), and (S )-tert-butyl(8-hydroxyquinolin-
5-yl)methyl(1-phenylethyl)carbamate (8c)
The compounds were synthesized from (S )-1-phenylethan-
amine following the procedure described for 5a, 5b and 5c.
1
Compound 8a: yield 32%; H NMR (CDCl3) d 1.34 (d, 3H),
5.1.13. 5,50-(4-(Trifluoromethyl)benzylazanediyl)
bis(methylene)diquinolin-8-ol (14)
It was synthesized from crude 5-(chloromethyl)quinoline
and 4-(trifluoromethyl)benzylamine following the procedure
described for 6a. Compound 14: yield 65%; 1H NMR
(CDCl3) d 5.23 (s, 2H), 5.56 (s, 4H), 7.28e7.39 (m, 3H),
7.51e7.85 (m, 5H), 8.41 (m, 2H), 8.72 (m, 2H), 8.95 (m,
4H); ESI-MS m/z ¼ 458 ([M þ H]þ, 100).
3.68e3.78 (m, 6H), 6.92 (m, 3H), 7.1e7.3 (m, 8H), 8.16
(m, 2H), 8.58 (m, 2H); ESI-MS m/z ¼ 436 ([M þ H]þ,
100%). Compound 8b: yield 14%; 1H NMR (MeOH-d4)
d 1.79 (d, 3H), 4.47 (s, 2H), 4.73 (m, 2H), 7.45e7.65 (m,
6H), 7.91 (m, 1H), 8.14 (m, 1H), 9.0e9.15 (m, 2H); 13C
NMR (MeOH-d4) d 20.5, 47.3, 61.0, 116.7, 120.2, 124.5,
124.6, 129.5, 130.3, 131.0, 136.6, 137.4, 145.4, 145.6,
151.0; ESI-MS m/z ¼ 279 (free base, [M þ H]þ, 100%). Com-
5.2. Biology
1
pound 8c: yield 15%; H NMR (CDCl3) d 1.27 (s, 9H), 1.45
5.2.1. In vitro KB3-1 cell line assay
(d, 3H), 4.55 (s, 2H), 4.82 (m, 2H), 7.1e7.4 (m, 6H), 8.15e
8.45 (m, 2H), 8.65e8.75 (m, 2H); ESI-MS m/z ¼ 379
([M þ H]þ, 100%).
Human mouth epidermal carcinoma (KB) cell line was incu-
bated for 72 h with different concentrations of drugs. The final
volume in each experiment was made up with the media
containing 1% DMSO final volume. Docetaxel was used as
positive control. The experiments were performed in triplicate.
Cell growth inhibition was determined by the MTS assay
according to the recommendations of the manufacturer [Prom-
ega]. The optical density was measured at 490 nm. The number
of viable cells was proportional to the extent of formazan pro-
duction. The percent cytotoxicity index [(OD490 treated/OD490
control) ꢂ 100] was calculated from three experiments. For
CC50 determinations, the cytotoxicity index was plotted against
5.1.9. 5,50-(Naphthalen-1-ylmethylazanediyl)bis(methylene)
diquinolin-8-ol (9a) and 5-((naphthalen-1-ylmethylamino)
methyl)quinolin-8-ol dihydrochloride (9b)
The compounds were synthesized from 1-naphthylmethyl-
amine following the procedure described for 3a and 3b. Com-
1
pound 9a: yield 4%; H NMR (CDCl3) d 4.0e4.4 (m, 6H),
7.0e9.8 (m, 17H); ESI-MS m/z ¼ 472 ([M þ H]þ, 100%).
1
Compound 9b: yield 2%; H NMR (MeOH) d 4.5 (s, 2H),

V. Moret et al. / European Journal of Medicinal Chemistry 44 (2009) 558e567
567
the drug concentration ranged over 10e0.5 nM and the value
resulting in 50% cytotoxicity was determined.
Acknowledgments
The authors thank E. Coste and M. Toure for technical
assistance and Prof. K. Dudley for critical reading of the man-
5.2.2. Tubulin and microtubule studies
´
uscript. This work was supported by: Conseil Regional Pro-
Preparation of lamb brain tubulin. Tubulin was extracted
from lamb brain by ammonium sulfate fractionation and ion-
exchange chromatography and stored in liquid nitrogen. The pro-
tein was prepared as described [13e15]. Protein concentrations
were determined spectrophotometrically with a PerkineElmer
spectrophotometer Lambda 800 and an extinction coefficient
at 275 nm of 1.09 L/g/cm in 6 M guanidine hydrochloride.
ˆ
vence Alpes Cote d’Azur, INSERM (Institut National de la
sante et de la Recherche Medicale) (VM fellowship), CNRS
´
´
(Centre National de la Recherche Scientifique), and IBDML
´
(Institut de Biologie du Developpement de Marseille-Luminy).
References
5.2.3. Tubulin assembly
[1] S. Akiyama, A. Fojo, J.A. Hanover, I. Pastan, M.M. Gottesman, Isolation
and genetic characterization of human KB cell lines resistant to multiple
drugs, Somat. Cell Mol. Genet. 11 (1985) 117e126.
Microtubule assembly was performed in 20 mM sodium
phosphate buffer, 1 mM EGTA, 8 mM MgCl2, 3.4 M glycerol,
and 1 mM GTP pH 6.5. The reaction was started by warming
the samples at 37 ꢀC in thermostatted cuvettes (1 ꢂ 0.2 cm),
and the mass of polymer formed was monitored by turbidim-
etry at 350 nm with a Beckman DU7400 spectrophotometer
[15]. Samples containing the compound and their controls
had less than 1% residual Me2SO.
[2] A.Y. Chen, L.F. Liu, Design of topoisomerase inhibitors to
overcome MDR1-mediated drug resistance, Adv. Pharmacol. 29B
(1994) 245e256.
[3] A.K. Samraj, D. Sohn, K. Schulze-Osthoff, I. Schmitz, Loss of caspase-9
reveals its essential role for caspase-2 activation and mitochondrial
membrane depolarization, Mol. Biol. Cell 18 (2007) 84e93.
[4] P.S. Schwartz, D.J. Waxman, Cyclophosphamide induces caspase 9-
dependent apoptosis in 9L tumor cells, Mol. Pharmacol. 60 (2001)
1268e1279.
5.2.4. Proteasomal chymotrypsin-like assay
[5] J.H. Burckhalter, R.I. Leib, Amino- and chloromethylation of 8-
quinolinol. Mechanism of preponderant ortho substitution in phenols
under Mannich conditions, J. Org. Chem. 26 (1961) 4078e4083.
[6] C. Kaiser, P.A. Dandridge, E. Garvey, K.E. Flaim, R.L. Zeid, J.P. Hieble,
Dopamine receptor agonist activity of some 5-(2-aminoethyl)carbostyril
derivatives, J. Med. Chem. 28 (1985) 1803e1810.
Whole cell extracts (10 mg) of cells treated with indicated
compounds or mixtures were incubated for 2 h at 37 ꢀC in
100 ml of assay buffer (50 mmol/L TriseHCl, pH 7.5) with
10 mmol/L of fluorogenic substrate Suc-LLVY-AMC as de-
scribed previously [16]. After incubation, production of free
hydrolyzed 7-amino-4-methylcoumarin (AMC) groups liber-
ated by substrate hydrolysis was fluorometrically measured
using a Victor 3 Multilabel Counter with an excitation filter
of 380 nm and an emission filter of 460 nm (PerkineElmer,
Boston, MA, USA).
[7] M.A. Jordan, L. Wilson, Microtubules and actin filaments: dynamic targets
for cancer chemotherapy, Curr. Opin. Cell Biol. 10 (1998) 123e130.
[8] J.A. Evangelio, M. Abal, I. Barasoain, A.A. Souto, M.P. Lillo, A.U. Acuna,
F. Amat-Guerri, J.M. Andreu, Fluorescent taxoids as probes of the micro-
tubule cytoskeleton, Cell Motil. Cytoskeleton 39 (1998) 73e90.
[9] M. Di Vaira, C. Bazzicalupi, P. Orioli, L. Messori, B. Bruni, P. Zatta,
Clioquinol, a drug for Alzheimer’s disease specifically interfering with
brain metal metabolism: structural characterization of its zinc(II) and
copper(II) complexes, Inorg. Chem. 43 (2004) 3795e3797.
[10] D.A. Richards, Prophylactic value of clioquinol against travellers’
diarrhoea, Lancet 1 (1971) 44e45.
5.2.5. Caspase activity assay
Caspase activities were assayed in HL60 cell line after treat-
ment with chemicals for 48 h. HL60 cells (20,000 cells per well
in 180 ml RPMI medium) were plated in black 96-well culture
microplates and treated with chemicals dissolved in DMSO
(1% final volume). Plates were kept under 5% CO2 for 48 h.
Twenty microliters of 10ꢂ lysis buffer consisting in 250 mM
Hepes (pH 7.5), 5 mM EDTA, 0.5% NP40, 0.1% SDS and
50 mM dithiothreitol was added before caspase substrates dis-
solved in water (DEVD-AMC and LEHD-AMC) or DMSO
(IETD-AMC) at a final concentration of 50 mM. Plates were in-
cubated at 37 ꢀC and fluorescence was recorded (lexc ¼ 360 nm,
lem ¼ 435 nm) at time 0, 30, 60, 120 and 180 min. Reaction
rates were calculated from the slope of the linear time-depen-
dent reaction and expressed as the fold-activation over the con-
trol (HL60 without chemical treatment). Colchicine 10ꢃ8 M
was used as a positive control and processed as JLK 1486 and
JLK 1472 added at concentrations 1, 0.1 and 0.01 mM.
[11] D. Chen, Q.C. Cui, H. Yang, R.A. Barrea, F.H. Sarkar, S. Sheng, B. Yan,
G.P. Reddy, Q.P. Dou, Clioquinol, a therapeutic agent for Alzheimer’s
disease, has proteasome-inhibitory, androgen receptor-suppressing,
apoptosis-inducing, and antitumor activities in human prostate cancer
cells and xenografts, Cancer Res. 67 (2007) 1636e1644.
[12] E. Petrucci, L. Pasquini, A. Petronelli, E. Saulle, G. Mariani, R. Riccioni,
M. Biffoni, G. Ferretti, P. Benedetti-Panici, F. Cognetti, G. Scambia,
R. Humphreys, C. Peschle, U. Testa, A small molecule Smac mimic
potentiates TRAIL-mediated cell death of ovarian cancer cells, Gynecol.
Oncol. 105 (2007) 481e492.
[13] R.C. Weisenberg, G.G. Borisy, E.W. Taylor, The colchicine-binding
protein of mammalian brain and its relation to microtubules, Biochemis-
try 7 (1968) 4466e4479.
[14] J.C. Lee, R.P. Frigon, S.N. Timasheff, The chemical characterization of
calf brain microtubule protein subunits, J. Biol. Chem. 248 (1973)
7253e7262.
[15] J.M. Andreu, M.J. Gorbunoff, J.C. Lee, S.N. Timasheff, Interaction of
tubulin with bifunctional colchicine analogues: an equilibrium study,
Biochemistry 23 (1984) 1742e1752.
[16] C. Bailly, C. Bal, P. Barbier, S. Combes, J.P. Finet, M.P. Hildebrand,
V. Peyrot, N. Wattez, Synthesis and biological evaluation of 4-
arylcoumarin analogues of combretastatins, J. Med. Chem. 46 (2003)
5437e5444.
5.2.6. Cytotoxic potentiation effect by TRAIL
The experimental procedure for these assays is described
by Petrucci et al. [12].