Arylthioindole inhibitors of tubulin polymerization. 3. Biological evaluation, structure-activity relationships and molecular modeling studies

Article abstract of DOI:10.1021/jm061479u

The new arylthioindole (ATI) derivatives 10, 14-18, and 21-24, which bear a halogen atom or a small size ether group at position 5 of the indole moiety, were compared with the reference compounds colchicine and combretastatin A-4 for biological activity. Derivatives 10, 11, 16, and 21-24 inhibited MCF-7 cell growth with IC₅₀ values <50 nM. A halogen atom (14-17) at position 5 caused a significant reduction in the free energy of binding of compound to tubulin, with a concomitant reduction in cytotoxicity. In contrast, methyl (21) and methoxy (22) substituents at position 5 caused an increase in cytotoxicity. Compound 16, the most potent antitubulin agent, led to a large increase (56%) in HeLa cells in the G₂/M phase at 24 h, and at 48 h, 26% of the cells were hyperploid. Molecular modeling studies showed that, despite the absence of the ester moiety present in the previously examined analogues, most of the compounds bind in the colchicine site in the same orientation as the previously studied ATIs. Binding to β-tubulin involved formation of a hydrogen bond between the indole and Thr179 and positioning of the trimethoxy phenyl group in a hydrophobic pocket near Cys241.

Full text of DOI:10.1021/jm061479u

J. Med. Chem. 2007, 50, 2865-2874
2865
Arylthioindole Inhibitors of Tubulin Polymerization. 3. Biological Evaluation,
Structure-Activity Relationships and Molecular Modeling Studies
Giuseppe La Regina,†,‡ Michael C. Edler,§ Andrea Brancale,| Sahar Kandil,| Antonio Coluccia,† Francesco Piscitelli,†
Ernest Hamel,§ Gabriella De Martino,† Ruth Matesanz, Jose´ Fernando D´ıaz, Anna Ivana Scovassi,# Ennio Prosperi,#
Antonio Lavecchia,X Ettore Novellino,X Marino Artico,† and Romano Silvestri*,†
Istituto Pasteur - Fondazione Cenci Bolognetti, Dipartimento di Studi Farmaceutici, Sapienza UniVersita` di Roma, Piazzale Aldo Moro 5,
I-00185 Roma, Italy, Welsh School of Pharmacy, Cardiff UniVersity, King Edward VII AVenue, Cardiff, CF10 3XF, United Kingdom,
Toxicology and Pharmacology Branch, DeVelopmental Therapeutics Program, DiVision of Cancer Treatment and Diagnosis, National Cancer
Institute at Frederick, National Institutes of Health, Frederick, Maryland 21702, Dipartimento di Chimica Farmaceutica e Tossicologica,
UniVersita` di Napoli “Federico II”, Via Domenico Montesano 49, I-80131, Napoli, Italy, Centro de InVestigaciones Biolo´gicas, Consejo
Superior de InVestigaciones Cientificas, C/Ramiro de Maeztu 9, E-28040 Madrid, Spain, and Istituto di Genetica Molecolare - Consiglio
Nazionale delle Ricerche, Via Abbiategrasso 207, I-27100 PaVia, Italy
ReceiVed December 27, 2006
The new arylthioindole (ATI) derivatives 10, 14-18, and 21-24, which bear a halogen atom or a small
size ether group at position 5 of the indole moiety, were compared with the reference compounds colchicine
and combretastatin A-4 for biological activity. Derivatives 10, 11, 16, and 21-24 inhibited MCF-7 cell
growth with IC50 values <50 nM. A halogen atom (14-17) at position 5 caused a significant reduction in
the free energy of binding of compound to tubulin, with a concomitant reduction in cytotoxicity. In contrast,
methyl (21) and methoxy (22) substituents at position 5 caused an increase in cytotoxicity. Compound 16,
the most potent antitubulin agent, led to a large increase (56%) in HeLa cells in the G2/M phase at 24 h, and
at 48 h, 26% of the cells were hyperploid. Molecular modeling studies showed that, despite the absence of
the ester moiety present in the previously examined analogues, most of the compounds bind in the colchicine
site in the same orientation as the previously studied ATIs. Binding to â-tubulin involved formation of a
hydrogen bond between the indole and Thr179 and positioning of the trimethoxy phenyl group in a
hydrophobic pocket near Cys241.
Chart 1. Reference and Arylthioindole Derivatives
Introduction
Microtubules have essential roles in vital cellular functions,
such as motility, division, shape maintenance, and intracellular
transport. Drugs that interact with tubulin, the protein subunit
of microtubules, cause mitotic arrest, interfering with the
dynamic equilibrium of these organelles by either inhibiting
tubulin polymerization or blocking microtubule disassembly.
Inhibitors of tubulin assembly include colchicine (1), combre-
tastatin A-4 (CAS4, 2a), and the Catharanthus alkaloids
vincristine and vinblastine (Chart 1). At high concentrations, 1
and 2a interact with R,â-tubulin dimers at the interface between
alpha and beta1 and cause microtubule destabilization and
apoptosis. Taxoids and epothilones bind as well to R,â-tubulin
to a lumenal site on the â-subunit2,3 and probably to a recently
described microtubule stabilizing agent binding site4 in the pore
on the microtubule surface formed by two different R and â
tubulin subunits. Paclitaxel stimulates microtubule polymeri-
zation and stabilization at high concentrations, whereas at lower
concentrations, the drug inhibits microtubule dynamics with little
effect on the proportion of tubulin in polymer.5,6 Independent
of precise mechanism of action, clinical use of antitubulin drugs
is associated with problems of drug resistance, toxicity, and
bioavailability.7
In the past few years, several antitubulin agents that target
the colchicine site have been intensively investigated as vascular-
disrupting antitumor drugs.8 For example, combretastatin A-4
phosphate (2b) and ZD6126 (3) stop blood flow through tumor
capillaries, probably caused by rapid disruption of endothelial
cell morphology, and consequently, the tumor is starved of
nutrients and rapid tumor cell death occurs.9,10 These vascular-
disrupting agents are currently in ongoing clinical trials for either
single-drug or multi-drug combination antitumor therapy.10,11
Arylthioindoles (ATIs, general structure 4) are a new class
of potent tubulin assembly inhibitors that bind to the colchicine
site on â-tubulin close to its interface with R-tubulin.12
* To whom correspondence should be addressed. Phone: +39 06 4991
3800. Fax: +39 06 491 491. E-mail: romano.silvestri@uniroma1.it.
† Universita` di Roma.
‡ Research performed at Cardiff University.
§ National Cancer Institute at Frederick.
| Cardiff University.
Consejo Superior de Investigaciones Cientificas, Madrid.
# Consiglio Nazionale delle Ricerche, Pavia.
X Universita` di Napoli.
10.1021/jm061479u CCC: $37.00 © 2007 American Chemical Society
Published on Web 05/12/2007

2866 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 12
Chart 2. Design of Arylthioindoles 9-28
La Regina et al.
Table 1. Structures of ATIs 9-31
cmpd
R1
R2
R3
R4
R5
R6
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
H
H
CH3
H
H
H
CH3
H
H
H
H
H
H
H
CH3
H
H
H
H
H
OCH3
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Cl
Cl
OCH3 OCH3 OCH3
OCH3 OCH3 OCH3
H
H
H
H
CH3
CH3
OCH3 OCH3 OCH3 Cl
OCH3 OCH3 OCH3 Cl
OCH3 OCH3 OCH3 Br
OCH3 OCH3 OCH3
OCH3 OCH3 OCH3
OCH3 OCH3 OCH3 NO2
OCH3 OCH3 OCH3 NH2
OCH3 OCH3 OCH3 CH3
OCH3 OCH3 OCH3 OCH3
OCH3 OCH3 OCH3 OCH3
OCH3 OCH3 OCH3 OCH2CH3
OCH3 OCH3 OCH3 OCH(CH3)2
OCH3 OCH3 OCH3 OH
OCH3 OCH3 OCH3 OCH2CH2OH
OCH3 OCH3 OCH3 OCH2CH2OCH2Ph
OCH3 OCH3 OCH3
I
F
Structure-activity relationship (SAR) analysis clarified struc-
tural requirements for good activity in this class of inhibitors.
Essential structural features for an active agent have included
(A) a small-size ester function at position 2 of the indole, (B)
the 3-arylthio group, (C) the sulfur atom bridge, and (D) a
substituent at position 5 of the indole (Chart 1).13 We carried
out molecular modeling studies and dynamics simulations that
helped explain the experimental data. We therefore used our
molecular model in designing and synthesizing new ATI
derivatives.13
H
H
H
H
H
H
H
H
H
COOCH3
COOCH3
COOCH3
H
OCH3 OCH3 OCH3 Cl
OCH3 OCH3 OCH3 OCH3
Scheme 1. Synthesis of Arylthioindole Derivatives 9-28a
Recent studies have focused on the synthesis of aminoderiva-
tives related to CSA4.14 The potent antitubulin activity displayed
by these analogues (for example 5,15 6,16 7,16 and 817, Chart 2)
attracted our attention. Compounds 5-8 share, as a common
structural feature, an amino group located ortho to the bridging
group (either carbonyl or cis-ethenyl group). We hypothesized
that this ortho-substituted aniline might resemble the indole
nucleus of ATI derivatives (see Chart 2), with the indole ring
acting as a bioisostere of the ortho-substituted aniline. These
observations prompted us to design new ATI derivatives 9-28.
Predictive docking simulations using our model13 showed that,
despite the absence of the ester moiety at position 2 of the indole
ring, most of the compounds should bind in the colchicine site
of tubulin in the same orientation as the previously studied ATIs.
The new ATI derivatives, like those described previously, were
potent inhibitors of tubulin polymerization and of the growth
of cancer cells, with activities comparable with those of
colchicine and combretastatin A-4. Finally, we should note the
recent paper of Hsieh and co-workers,18 which included a group
of 3-aroylthioindoles. These compounds are significantly dif-
ferent from the ATI’s we have prepared, because there are major
differences in our SAR findings and those of the Hsieh group.
Chemistry
a Reagents and reaction conditions: (a) 3 N NaOH, D-(+)-glucose, EtOH,
65 °C, 2 h; (b) appropriate indole, I2-KI, EtOH-H2O, 25 °C, 1 h; (c) (2-
bromoethoxy)-tert-butyldimethylsilane, K2CO3, acetone, 24 h, reflux; (d)
PTSA, methanol, 25 °C, 30 min; (e) alkyl halide, K2CO3, acetone, 24 h,
reflux; (f) 2-benzyloxyethanol, DEAD, PPh3, THF, overnight, reflux.
The structures of ATI derivatives 9-31 are shown in Table
1. Compounds 10, 11, 14-26, and 28 were synthesized by a
two-step procedure (Scheme 1). O-Ethyl-S-(3,4,5-trimethoxy-
phenyl)carbonodithioate19 was transformed into 3,4,5-trimethoxy-
thiophenol by heating at 65 °C in aqueous ethanol in the
presence of sodium hydroxide. This mixture was made acidic
with 6 N HCl and treated at 25 °C with the appropriate indole
while adding dropwise an aqueous iodine-potassium iodide

Arylthioindole Inhibitors of Tubulin Polymerization
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 12 2867
Chart 3. Reference Structures for Binding Studies
Table 2. Inhibition of Tubulin Polymerization, Growth of MCF-7
Human Breast Carcinoma Cells, and Colchicine Binding by Compounds
9-31
tubulina
IC50 ( SD
(µM)
MCF-7b
IC50 ( SD
(nM)
inhibition
colchicine bindingc
(% ( SD)
cmpd
9d
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29d
30d
31d
1
15 ( 0.7
2.6 ( 0.2
6.8 ( 0.6
11 ( 2
9.4 ( 0.3
2.6 ( 0.2
2.7 ( 0.5
1.6 ( 0.3
2.7 ( 0.3
3.3 ( 0.3
16 ( 0.4
13 ( 2
2.7 ( 0.2
4.1 ( 0.6
3.3 ( 0.2
2.1 ( 0.1
19 ( 0.6
6.3 ( 0.8
6.8 ( 0.8
>40
2.9 ( 0.1
2.3 ( 0.3
2.0 ( 0.2
3.2 ( 0.4
2.2 ( 0.2
>2500
nde
68 ( 0.8
61 ( 4
nd
33 ( 3
51 ( 4
59 ( 5
65 ( 3
61 ( 2
39 ( 6
nd
34 ( 9
46 ( 4
>2500
1200 ( 100
77 ( 7
82 ( 10
43 ( 7
68 ( 7
Introduction of a chlorine atom at position 5 of the indole
(14, IC50 ) 2.6 µM) did not change the activity of 10 as an
inhibitor of assembly, but inhibition of cell growth was reduced
greater than 2-fold. When a chlorine atom was introduced into
11 (compound 15), inhibition of tubulin assembly increased
while inhibition of MCF-7 cell growth decreased. (Alternatively,
15 can be viewed as an introduction of a C-2 methyl group
into compound 14, resulting in no change in activity, as
compared with the loss of activity when the same methyl group
was introduced into 10.) With the other three halogen atoms
(16-18), a bromine atom (16) resulted in a compound more
inhibitory than 10 in the tubulin assembly assay and essentially
identical to 10 as an inhibitor of MCF-7 cell growth, an iodine
atom (17) was almost indistinguishable in its effects from the
chlorine atom, and a fluorine atom (18) resulted in the least
active compound in the halogen series, although with the latter
compound there was a greater loss in antiproliferative than in
antitubulin activity.
Introduction into 10 of a methyl (21, IC50 ) 2.7 µM), a
methoxy (22, IC50 ) 4.1 µM; 23, with a C-2 methyl group also,
IC50 ) 3.3 µM), or an ethoxy (24, IC50 ) 2.1 µM) group at
position 5 of the indole resulted in little change in inhibitory
effect on tubulin assembly and small increases in antiprolifera-
tive activity relative to 10. These compounds, too, were
essentially equipotent with 1 and 2a, both as inhibitors of tubulin
assembly and MCF-7 cell growth. Bulkier ether groups at
position 5 of the indole (25, 27, 28), or even a hydroxyl group
at this position (26), yielded compounds with reduced activity
in both assays.
160 ( 50
560 ( 70
260 ( 30
16 ( 6
nd
56 ( 3
61 ( 4
69 ( 0.2
76 ( 5
nd
26 ( 0.5
31 ( 2
nd
22 ( 2
18 ( 4
16 ( 5
1500 ( 700
190 ( 40
95 ( 8
260 ( 20
25 ( 1
74 ( 2
57 ( 2
90 ( 1
42 ( 1
13 ( 3
13 ( 3
2a
17 ( 10
99 ( 1
a Inhibition of tubulin polymerization. b Inhibition of growth of MCF-7
human breast carcinoma cells.20 c Inhibition of [3H]colchicine binding.
Tubulin was at 1 µM, both [3H]colchicine and inhibitor were at 5 µM.20
d Data from lit.12,13 e nd ) not determined.
solution. Compounds 12 and 13 were prepared similarly, starting
from the corresponding commercially available carbonodithioate.
Compound 27 was prepared by treating 26 with (2-bromo-
ethoxy)-tert-butyldimethylsilane in the presence of potassium
carbonate in boiling acetone; the intermediate silyloxy derivative
was stirred with para-toluensulfonic acid in methanol at room
temperature. The 5-ethoxy- (32) and 5-isopropyloxy- (33)
indoles were obtained by alkylation of 5-hydroxyindole with
iodoethane or 2-iodopropane, respectively, in the presence of
potassium carbonate. 5-(2-Benzyloxy)ethoxyindole (34) was
prepared by reaction of 5-hydroxyindole with 2-benzyloxyetha-
nol in the presence of diethyl azodicarboxylate and triphen-
ylphosphine in boiling THF.
In the assay measuring inhibition of [3H]colchicine binding,
all compounds were examined at 5 µM, with 1 µM tubulin and
5 µM 1.20 None of the new compounds approached the standard
2a as an inhibitor in this assay. With the inhibitors at 1 µM, 2a
inhibited colchicine binding 88%, and the most active of the
ATI’s, compound 24, inhibited 42%.
Structure-Affinity-Cytotoxicity Relationships. We also
determined binding constants at 20 °C for the colchicine site
on tubulin for the new ATIs that were highly active as inhibitors
of tubulin polymerization. Inhibition of tubulin polymerization
does not directly correlate with binding affinity. On the other
hand, binding affinity can be an important parameter involved
in cytotoxicity.21-25
Results and Discussion
Biological data for the inhibition of tubulin polymerization,
binding of [3H]colchicine to tubulin (more active compounds
only), and growth of MCF-7 human breast carcinoma cells by
arylthioindoles 9-28 in comparison with the reference com-
pounds 1 and 2a and ATIs 29-3112,13 are summarized in Table
2.
Replacing the 3-phenylthio of 9 with the 3-(3,4,5-trimethoxy-
phenyl)thio group (10) resulted in a 5.8-fold increase in
inhibition of tubulin polymerization. This value (IC50 ) 2.6 µM)
Compounds 10, 11, 14-17, 21-23, 29, and 30 were
compared with 1 and 2-methoxy-5-(2,3,4-trimethoxyphenyl)-
2,4,6-cycloheptatrien-1-one (MTC, 35, Chart 3), an analogue
of colchicine lacking the B ring that rapidly reaches an
equilibrium in its binding reaction with tubulin (Table 3), while
losing part of the free energy of binding due to the entropy
contribution needed for the immobilization of the A and C rings
in the site.26 As expected from their cytotoxicities and their
activities as inhibitors of [3H]colchicine binding, the compounds
bound tightly to the colchicine site of tubulin, as they displaced
compound 35 already bound to the colchicine site (Figure 1).
We explored structure-affinity relationships of the substit-
uents at positions 2 and 5 of the indole nucleus. Invariably, the
was very close to those of 1 (IC50 ) 3.2 µM) and 2a (IC50
)
2.2 µM). Most importantly, this chemical modification resulted
in great improvement in antiproliferative activity against MCF-7
cells. The IC50 obtained with 10 was 34 nM, a value only 2.6-
and 2-fold higher than those obtained with reference compounds
1 (IC50 ) 13 nM) and 2a (IC50 ) 17 nM), respectively. The
2-methyl group of 11 caused a 2.6-fold reduction in activity as
an inhibitor of tubulin assembly (IC50 ) 6.8 µM) relative to
10, while inhibition of MCF-7 cell growth (IC50 ) 46 nM) was
only marginally affected.

2868 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 12
La Regina et al.
Table 3. Binding Constants at 20 °C of Compounds 10, 11, 14-17,
21-23, 29, and 30 and the Reference Compounds 1 and 35 for the
Colchicine Site on Tubulin
Table 4. Incremental Thermodynamic Parameters of Binding of
Compounds 11, 14-17, 21-23, 29 and 30 to the Colchicine Site
MCF-7
binding constant
∆Gapp 20 °C
(kJ mol-1
reference
cmpd
single group
modificationa
∆∆G 20 °C
(kJ mol-1
∆IC50
cmpd
(×105 M-1
)
)
cmpd
)
(nM)
10a
11a
14a
15a
16a
17a
21a
22a
23a
29b
30b
35c
1c
52.1 ( 2.8
55.0 ( 3.0
8.1 ( 0.3
20.7 ( 2.4
27.6 ( 1.3
23.1 ( 2.1
19.6 ( 0.5
30.5 ( 1.9
63.8 ( 1.6
5.7 ( 0.4
13 ( 2
-37.7 ( 0.1
-37.8 ( 0.1
-33.1 ( 0.1
-35.4 ( 0.3
-36.1 ( 0.1
-35.7 ( 0.2
-35.3 ( 0.1
-36.4 ( 0.1
-38.2 ( 0.1
-32.3 ( 0.2
-34.3 ( 0.3
-31.8 ( 0.2
-46.0
11
14
15
16
17
21
22
23
29
30
10
10
10
11
10
10
10
10
22
10
14
35
2-CH3
5-Cl
2-CH3-5-Cl
5-Br
5-I
-0.1 ( 0.2
+4.5 ( 0.2
+2.4 ( 0.4
+1.5 ( 0.2
+1.9 ( 0.3
+2.3 ( 0.2
+1.2 ( 0.2
-1.8 ( 0.2
+5.3 ( 0.2
-1.2 ( 0.4
-5.8 ( 0.3
+12
+43
+36
+9
+34
-18
-12
-4
-9
-35
5-CH3
5-OCH3
2-CH3-5-CH3O
2-COOCH3
2-COOCH3-5-Cl
4.7 ( 0.3
1600
a Single group modification (H f substituent) highlighted in bold on
indole nucleus with respect to the indicated reference compound.
a Data obtained measuring compound 35 displacement from the binding
site.27 b Data measured using quenching of tubulin fluorescence.28,29 c Data
from lit.28
22 were more cytotoxic. The methyl group at position 2
increased binding affinity (negative value of ∆∆G 20 °C, Table
4), but this positive effect was not associated with an increase
in cytotoxicity (compare 10 with 11 and 22 with 23), probably
due to a negative effect in solubility of the compound. The
2-methoxycarbonyl group of 29 and 30 caused opposite effects
on the binding affinity of 10 and 14, respectively. However,
both 29 and 30 were more cytotoxic.
These results indicate that cytotoxicity is generally correlated
with binding affinity for the colchicine site. However, some
results were not fully explained in terms of binding affinity,
suggesting that others factors, such as specific effects of the
compound on the tubulin molecule,22 may play a role in the
cytotoxic activities of these derivatives.
We also analyzed the data from Table 2, combined with data
from our previous studies,12,13 comparing effects of ATIs on
MCF-7 cell proliferation with inhibitory effects on tubulin
assembly (Figure 2A) and with inhibitory effects on the binding
of [3H]colchicine to tubulin (Figure 2B). As noted above,
assembly inhibition correlated poorly with inhibition of the
growth of this cell line.
There was much better correlation with inhibition of colchi-
cine binding, which indirectly measures relative affinity of the
compounds for the colchicine site. Note, too, that 2a and two
compounds prepared by Flynn et al.20 (previously cited by us12)
also fit into the overall pattern generated by the ATIs.
Cell Cycle Analysis. The most potent antitubulin agent 16
was selected for cell cycle studies in HeLa cells. After treatment
for 24 h with 0.1, 1.0, and 10 µM 16, the cells showed a dose-
dependent reduction in cell growth. At the highest concentration
used, cell growth was reduced about 50% (not shown). Cell
cycle analysis (Figure 3 and Table 5) following treatment with
10 µM 16 showed that 56% of the cells were arrested in the
G2/M phase after 24 h. Following replacement of the original
medium with medium not containing the drug and incubation
for a further 24 h, 53% of the cells remained in G2/M and an
additional 26% of the cells had DNA content >4C. These
findings indicate a continuing impairment of cell division, as
would be expected following treatment with a tubulin inhibitor.
The morphological features of 16-treated cells were analyzed
by Hoechst staining of cellular DNA. The nuclei of untreated
HeLa cells showed the typical diffuse pattern of chromatin
distribution, whereas cells treated with 16 became multinucleated
as an effect of perturbation of tubulin function (Figure 4A).
Although there was a small fraction of cells in the sub-G1 region
(the A0 compartment in Table 5), other markers of apoptosis
were not observed. There was neither a DNA ladder (Figure
Figure 1. (A) Displacement of 35 from the colchicine site. Fluores-
cence emission spectra of 10 µM 35 and 10 µM tubulin in 10 mM
phosphate-0.1 mM GTP buffer pH 7.0, in the presence of compound
14: (a) 0 µM, (b) 2 µM, (c) 5 µM, (d) 10 µM, (e) 20 µM, (f) 50 µM.
(B) Displacement isotherm at 20 °C of 35 by 14. The data points were
fit to the best value of the binding equilibrium constant of 14, assuming
0.8 sites per tubulin dimer.28
presence of a halogen atom (14-17) or a methyl (21) or
methoxy (22) group at position 5 resulted in reduction of binding
affinity (positive value of ∆∆G 20 °C, Table 4). This reduction
was accompanied by a concomitant reduction of cytotoxicity
in 14-17 relative to 10, while, in contrast, compounds 21 and

Arylthioindole Inhibitors of Tubulin Polymerization
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 12 2869
Figure 2. Correlation of MCF-7 cytotoxicity data with inhibition of tubulin assembly (A) and inhibition of colchicine binding (B). Data represented
by open inverted triangles are taken from Table 2 of this paper. The open upright triangles represent ATI compounds described in ref 13 and the
open circles represent ATI compounds described in ref 12. The solid squares represent data obtained with 2a. The solid circles represent data
obtained with two compounds synthesized by Flynn et al.20
4B) nor evidence for PARP-1 proteolysis (Figure 4C), both of
which were instead observed following a 3 h treatment with
etoposide.
tions: compounds 9 and 10 showed a slightly different
conformation, with the indole buried deeper in the binding
pocket, although both compounds still form the interactions
described above. In the case of compounds 19 and 28, on the
other hand, the docking simulation did not yield a reasonable
pose within the active site. While with compound 28 it is
possible to rationalize this observation as being caused by steric
hindrance attributable to the substituent at position 5 of the
indole, with compound 19, the best explanation is that electro-
static effects cause a different positioning of the inhibitor in
the binding site. If the conformation of compound 19 in the
binding site was similar to the poses of the other analogues, its
nitro group would be just 3 Å away from the phosphate groups
of the nonexchangeable GTP molecule bound to the R-tubulin
in the dimer. This GTP site on R-tubulin is near the colchicine
site on â-tubulin, and the close proximity of the two ligands
would result in an unacceptable electrostatic repulsion between
negatively charged groups.
These data suggested that tubulin polymerization inhibition
induced in HeLa cells by 16 did not cause apoptosis but rather
impaired cell viability through a “mitotic catastrophe”.
Molecular Modeling. To investigate the possible binding
mode for this new series of compounds, we performed docking
simulations, using the FlexX module included in Sybyl 7.2.30
We previously reported the putative binding for ATIs bearing
an ester moiety at position 2 of the indole, describing the main
interaction between the inhibitors and the colchicine site of
tubulin, which included a hydrogen bond between the carbonyl
group of the ester function and a lysine in the binding site.12,13
With the series reported here, despite the absence of the ester
moiety, most of the compounds bind in the same orientation as
the previously studied ATIs, forming a hydrogen bond between
the indole and Thr179 (residue numbers, as in ref 1, describing
the crystal structure we used) and with the trimethoxyphenyl
group positioned in a hydrophobic pocket close to Cys241
(Figure 5).
Conclusions
We synthesized new ATI derivatives, many of which strongly
inhibited tubulin assembly, with activity in the low micromolar
range, comparable to the effects of 1 and 2a. Derivatives 10,
It should be noted that, while most of the compounds in this
series dock in a very similar fashion, there are a few excep-

2870 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 12
La Regina et al.
Figure 3. Cell cycle analysis of HeLa cells treated with 16. A typical experiment is shown. Cells were harvested after treatment with 16 (10 µM)
for 24 h and after further recovery in drug-free medium for 24 h (24 h + 24 h). The percentage of cells in each cell cycle phase was quantified
(Table 5).
Table 5. Cell Cycle Distribution of 16-Treated HeLa Cellsa
24 h
24 h + 24 h
cell cycle
phase
16b
control
DMSOc
16d
control
DMSO
e
A0
G1
S
6.5
1.8
8.9
56.1
4.2
0.2
69.6
17.4
12.0
0
0.1
73.9
15.2
10.1
0
4.5
3.5
10.5
53.1
26.0
0.4
81.4
6.3
11.5
0
0.2
76.7
8.1
14.4
0
G2/M
>4C
a Data are expressed as % of cells in each cell cycle phase. A typical
experiment is shown. b Cells were treated with 16 at 10 µM for 24 h.
c Parallel samples incubated with 0.1% DMSO (the same final concentration
used with 16 at 10 µM) did not significantly alter cell cycle distribution.
d Cells were further incubated in drug-free medium for 24 h. e Indicates
cells with a sub-G1 DNA content, probably representing a small population
of apoptotic cells.
14-18, and 21-24, bearing a halogen atom or a small alkyl or
ether group at position 5 of the indole, were also potent
inhibitors of MCF-7 cell growth. The most active derivatives
(10, 11, 16, and 21-24) inhibited cell growth with IC50 values
<50 nM. SAR studies indicated reasonable, albeit imperfect,
correlation between cytotoxicity and binding affinity for the
colchicine site. In particular, a halogen atom at position 5
decreased the free energy of binding of ATIs to tubulin, with
concomitant reduction in cytotoxicity. In contrast, methyl (21)
and methoxy (22) substituents at position 5 resulted in more
cytotoxic compounds. Compound 16, the most potent inhibitor
of tubulin assembly, induced accumulation of HeLa cells in the
G2/M phase of the cell cycle at 24 h and polyploidization at 48
h. At 24 h, inhibition of tubulin polymerization by 16 had not
caused extensive apoptosis, suggesting that impaired cell
viability might occur through a “mitotic catastrophe”. Molecular
modeling studies showed that, despite the absence of the ester
moiety, most of the compounds appear to bind in the same
orientation as the previously studied ATIs,12,13 forming a
hydrogen bond between the indole and Thr179 and with the
trimethoxyphenyl group positioned in a hydrophobic pocket near
Cys241. These findings induce us to continue our investigations
of the SAR among ATI derivatives in the expectation of
developing more potent and selective analogues.
Figure 4. Morphological and biochemical evaluation of apoptotic
parameters of HeLa cells treated with 16 and etoposide. (A) Hoechst
staining for DNA. (B) Agarose gel electrophoresis; Mr, molecular
weight markers; 1, control cells; 2, 16-treated cells (10 µM, 24 h + 24
h of recovery); 3, etoposide-treated cells (100 µM, 3 h + 24 h of
recovery). (C) Western blot for PARP-1; 1, control cells; 2, 16-treated
cells (10 µM, 24 h); 3, etoposide-treated cells (100 µM, 3 h + 24 h of
recovery); 116 kDa, intact PARP-1; 89 kDa, caspase-cleaved PARP-
1.
Experimental Section
Chemistry. Melting points (mp) were determined on a Bu¨chi
510 apparatus and are uncorrected. Infrared spectra (IR) were run
on a SpectrumOne FT spectrophotometer. Band position and
absorption ranges are given in cm-1. Proton nuclear magnetic

Arylthioindole Inhibitors of Tubulin Polymerization
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 12 2871
5-Chloro-3-[(2-methoxyphenyl)thio]-1H-indole (12). Com-
pound 12 was synthesized as 10 using 2-methoxythiophenol, yield
1
54%, mp 145-148 °C (from ethanol). H NMR (CDCl3): δ 3.97
(s, 3H), 6.59 (dd, J ) 7.78 and 1.62 Hz, 1H), 6.68-6.72 (m, 1H),
6.86 (dd, J ) 8.14 and 1.07 Hz, 1H), 7.05-7.09 (m, 1H), 7.22
(dd, J ) 8.64 and 2.04 Hz, 1H), 7.37 (d, J ) 8.63 Hz, 1H), 7.50
(d, J ) 2.63 Hz, 1H), 7.60 (d, J ) 2.02 Hz, 1H), 8.49 ppm (broad
s, disappeared on treatment with D2O, 1H). IR: ν 3433 cm-1. MS:
ES+ ) 312 (MNa+). Anal. (C15H12ClNOS (289.79)) C, H, Cl, N,
S.
5-Chloro-3-[(3,5-dimethylphenyl)thio]-1H-indole (13). Com-
pound 13 was synthesized as 10 using 3,5-dimethythiophenol and
5-chloro-1H-indole, yield 49%, mp 135-138 °C (from ethanol).
1H NMR (DMSO-d6): δ 2.12 (s, 6H), 6.66 (s, 2H), 6.70 (s, 1H),
7.18 (dd, J ) 8.61 and 2.09 Hz, 1H), 7.35 (d, J ) 2.06 Hz, 1H),
7.51 (d, J ) 8.14 Hz, 1H), 7.83 (s, 1H), 11.89 ppm (broad s,
disappeared on treatment with D2O, 1H). IR: ν 3357 cm-1. MS:
ES+ ) 310 (MNa+). Anal. (C16H14ClNS (287.81)) C, H, Cl, N, S.
5-Chloro-3-[(3,4,5-trimethoxyphenyl)thio]-1H-indole (14). Com-
pound 14 was synthesized as 10 using 5-chloro-1H-indole, yield
1
59%, mp 135-139 °C (from ethanol). H NMR (CDCl3): δ 3.68
(s, 6H), 3.79 (s, 3H), 6.36 (s, 2H), 7.20 (d, J ) 8.73 Hz, 1H), 7.35
(d, J ) 8.43 Hz, 1H), 7.51 (s, 1H), 7.62 (s, 1H) 8.69 ppm (broad
s, disappeared on treatment with D2O, 1H). IR: ν 3247 cm-1. MS:
ES+ ) 350 (MH+). Anal. (C17H16ClNO3S (349.84)) C, H, Cl, N,
S.
5-Chloro-2-methyl-3-[(3,4,5-trimethoxyphenyl)thio]-1H-in-
dole (15). Compound 15 was synthesized as 10 using 5-chloro-2-
methyl-1H-indole, yield 54%, mp 178-182 °C (from ethanol). 1H
NMR (DMSO-d6): δ 2.48 (s, 3H), 3.59 (s, 9H), 6.28 (s, 2H), 7.12
(dd, J ) 8.53 and 2.00 Hz, 1H), 7.32 (d, J ) 4.68 Hz, 1H), 7.40
(d, J ) 8.54 Hz, 1H), 11.84 ppm (broad s, disappeared on treatment
with D2O, 1H) IR: ν 3324 cm-1. MS: ES+ ) 364 (MH+). Anal.
(C18H18ClNO3S (363.86)) C, H, Cl, N, S.
Figure 5. Putative binding mode of different ATIs: compound 31 in
green, compound 24 in magenta, compound 10 in cyan.
resonance (1H NMR) spectra were recorded on Bruker 200 and
400 MHz FT spectrometers in the indicated solvent. Chemical shifts
are expressed in δ units (ppm) from tetramethylsilane. Mass spetra
were recorded on Bruker MicroTOF LC. Column chromatography
was performed on columns packed with alumina from Merck (70-
230 mesh) or silica gel from Merck (70-230 mesh). Aluminum
oxide TLC cards from Fluka (aluminum oxide precoated aluminum
cards with fluorescent indicator at 254 nm) and silica gel TLC cards
from Fluka (silica gel precoated aluminum cards with fluorescent
indicator at 254 nm) were used for thin layer chromatography
(TLC). Developed plates were visualized by a Spectroline ENF
260C/F UV apparatus. Organic solutions were dried over anhydrous
sodium sulfate. Concentration and evaporation of the solvent after
reaction or extraction was carried out on a Bu¨chi Rotavapor rotary
evaporator operating at reduced pressure. Elemental analyses were
found within (0.4% of the theoretical values. Compound 9 was
synthesized as we previously reported.12
5-Bromo-3-[(3,4,5-trimethoxyphenyl)thio]-1H-indole (16). Com-
pound 16 was synthesized as 10 using 5-bromo-1H-indole, yield
1
59%, mp 154-156 °C (from ethanol). H NMR (CDCl3): δ 3.70
(s, 6H), 3.82 (s, 3H), 6.39 (s, 2H), 7.31 (d, J ) 8.60 Hz, 1H), 7.35
(dd, J ) 8.64 and 1.82 Hz, 1H), 7.50 (d, J ) 2.64 Hz, 1H), 7.80
(s, 1H) 8.82 ppm (broad s, disappeared on treatment with D2O,
1H). IR: ν 3353 cm-1. MS: ES+ ) 416, 418 (MNa+). Anal.
(C17H16BrNO3S (394.28)) C, H, Br, N, S.
Method A. General Procedure for the Synthesis of Com-
pounds 10-26 and 28. Example: 3-[(3,4,5-Trimethoxyphenyl)-
thio]-1H-indole (10). D-(+)-glucose (1.07 g, 0.006 mol) and 3 N
NaOH (3.55 mL) were added to a solution of O-ethyl-S-(3,4,5-
trimethoxyphenyl)carbonodithioate19 (1.54 g, 0.005 mol) in ethanol
(20 mL). The reaction mixture was heated at 65 °C for 2 h while
stirring. After cooling, water (18 mL) and 6 N HCl (1.58 mL) were
poured into the reaction mixture, and indole (0.5 g, 0.0043 mol)
was added while stirring. A solution of iodine (1.08 g, 0.0043 mol)
and potassium iodide (3.06 g, 0.02 mol) in water (11.5 mL) was
dropped into the reaction, and it was stirred at 25 °C for 1 h. Water
(20 mL) and a saturated solution of sodium hydrogen carbonate
(15 mL) were added, and the mixture was extracted with chloro-
form. The organic layer was washed with brine, dried, and filtered.
Evaporation of the solvent gave 10, yield 47%, mp 125-128 °C
(from ethanol). 1H NMR (DMSO-d6): δ 3.56 (s, 6H), 3.57 (s, 3H),
6.38 (s, 2H), 7.08 (t, J ) 7.44 Hz, 1H), 7.18 (t, J ) 7.56 Hz, 1H),
7.45 (d, J ) 7.90 Hz, 1H), 7.48 (d, J ) 8.07 Hz, 1H), 7.77 (s, 1H),
11.67 ppm (broad s, disappeared on treatment with D2O, 1H). IR:
ν 3356 cm-1. MS: ES+ ) 338 (MNa+). Anal. (C17H17NO3S
(315.39)) C, H, N, S.
5-Iodo-3-[(3,4,5-trimethoxyphenyl)thio]-1H-indole (17). Com-
pound 17 was synthesized as 10 using 5-iodo-1H-indole, yield 60%,
mp 178-180 °C (from ethanol). 1H NMR (CDCl3): δ 3.71 (s, 6H),
3.81 (s, 3H), 6.39 (s, 2H), 7.23 (d, J ) 8.52 Hz, 1H), 7.47 (d, J )
2.65 Hz, 1H), 7.53 (dd, J ) 8.54 and 1.66 Hz, 1H), 8.02 (s, 1H)
8.66 ppm (broad s, disappeared on treatment with D2O, 1H). IR:
ν 3354 cm-1. MS: ES+ ) 464 (MNa+). Anal. (C17H16INO3S
(441.29)) C, H, I, N, S.
5-Fluoro-3-[(3,4,5-trimethoxyphenyl)thio]-1H-indole (18). Com-
pound 18 was synthesized as 10 using 5-fluoro-1H-indole, yield
1
57%, mp 160-163 °C (from ethanol). H NMR (CDCl3): δ 3.68
(s, 6H), 3.78 (s, 3H), 6.37 (s, 2H), 6.98-7.04 (m, 1H), 7.29 (dd, J
) 9.16 and 2.51 Hz, 1H), 7.35 (dd, J ) 8.83 and 4.21 Hz, 1H),
7.54 (d, J ) 2.68 Hz, 1H), 8.52 ppm (broad s, disappeared on
treatment with D2O, 1H). IR: ν 3344 cm-1. MS: ES+ ) 356
(MNa+). Anal. (C17H16FNO3S (333.38)) C, H, F, N, S.
5-Nitro-3-[(3,4,5-trimethoxyphenyl)thio]-1H-indole (19). Com-
pound 19 was synthesized as 10 using 5-nitro-1H-indole, yield 6%,
1
yellow oil. H NMR (DMSO-d6): δ 3.58 (s, 3H), 3.60 (s, 6H),
6.44 (s, 2H), 7.68 (d, J ) 8.54 Hz, 1H), 8.06-8.09 (m, 2H), 8.32
(d, J ) 2.33 Hz, 1H), 12.48 ppm (broad s, disappeared on treatment
with D2O, 1H). IR: ν 3284 cm-1. MS: ES+ ) 383 (MNa+). Anal.
(C17H16N2O5S (360.39)) C, H, N, S.
2-Methyl-3-[(3,4,5-trimethoxyphenyl)thio]-1H-indole (11). Com-
pound 11 was synthesized as 10 using 2-methyl-1H-indole, yield
1
53%, mp 135-137 °C (from ethanol/water). H NMR (DMSO-
5-Amino-3-[(3,4,5-trimethoxyphenyl)thio]-1H-indole (20). Com-
pound 20 was synthesized as 10 using 5-amino-1H-indole, yield
d6): δ 2.44 (s, 3H), 3.57 (s, 9H), 6.29 (s, 2H), 7.03 (t, J ) 7.03
Hz, 1H), 7.11 (t, J ) 7.27 Hz, 1H), 7.37-7.39 (m, 2H), 11.61
ppm (broad s, disappeared on treatment with D2O, 1H). IR: ν 3312
cm-1. MS: ES+ ) 352 (MNa+). Anal. (C18H19NO3S (329.42)) C,
H, N, S.
1
17%, mp 128-131 °C (from ethanol). H NMR (CDCl3): δ 3.60
(broad s, disappeared on treatment with D2O, 2H), 3.69 (s, 6H),
3.79 (s, 3H), 6.38 (s, 2H), 6.72 (dd, J ) 8.55 and 2.20 Hz, 1H),
6.91 (d, J ) 2.17 Hz, 1H), 7.24 (d, J ) 8.54 Hz, 1H), 7.42 (d, J )

2872 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 12
La Regina et al.
2.66 Hz, 1H) 8.30 ppm (broad s, disappeared on treatment with
D2O, 1H). IR: ν 3393 cm-1. MS: ES+ ) 331 (MH+). Anal.
(C17H18N2O3S (330.41)) C, H, N, S.
the organic layer was washed with brine, dried, and filtered.
Evaporation of the solvent gave 5-[2-(tert-butyldimethylsilyloxy)-
ethoxy]-3-[(3,4,5-trimethoxy phenyl)thio]-1H-indole (yield 41% as
a brown oil), which was used without further purification. To a
solution of the latter compound (0.11 g, 0.225 mol) in methanol
(1.13 mL) was added para-toluenesulfonic acid monohydrate (0.01
g, 0.05 mmol). The reaction mixture was stirred at 25 °C for 30
min, neutralized with a saturated solution of sodium hydrogen
carbonate, and extracted with ethyl acetate; the organic layer was
washed with brine, dried, and filtered. Evaporation of the solvent
gave a residue that was purified by silica gel column chromatog-
raphy (ethyl acetate-n-hexane 7:1 as eluent) to furnish 27, yield
22%, as a yellow oil. 1H NMR (CDCl3): δ 2.19 (broad s,
disappeared on treatment with D2O, 1H), 3.69 (s, 6H), 3.80 (s, 3H),
3.94-3.99 (m, 2H), 4.10 (t, J ) 4.53 Hz, 2H), 6.39 (s, 2H), 6.94
(dd, J ) 8.79 and 2.43 Hz, 1H), 7.11 (d, J ) 2.26 Hz, 1H), 7.33
(d, J ) 8.79 Hz, 1H), 7.48 (d, J ) 2.68 Hz, 1H), 8.65 ppm (broad
s, disappeared on treatment with D2O, 1H). IR: ν 3336 cm-1. MS:
ES+ ) 398 (MNa+). Anal. (C19H21NO5S (375.45)) C, H, N, S.
5-Methyl-3-[(3,4,5-trimethoxyphenyl)thio]-1H-indole (21). Com-
pound 21 was synthesized as 10 using 5-methyl-1H-indole, yield
45%, oil which solidified on standing, mp 81-84 °C (aqueous
1
ethanol). H NMR (CDCl3): δ 2.45 (s, 3H), 3.69 (s, 6H), 3.83 (s,
3H), 6.40 (s, 2H), 7.10 (d, J ) 7.69 Hz, 1H), 7.34 (d, J ) 8.24 Hz,
1H), 7.45 (s, 1H), 7.48 (d, J ) 2.58 Hz, 1H), 8.38 ppm (broad s,
disappeared on treatment with D2O, 1H). IR: ν 3346 cm-1. MS:
ES+ ) 330 (MH+). Anal. (C18H19NO3S (329.42)) C, H, N, S.
5-Methoxy-3-[(3,4,5-trimethoxyphenyl)thio]-1H-indole (22).
Compound 22 was synthesized as 10 using 5-methoxy-1H-indole,
1
yield 30%, 99-101 °C (from ethanol). H NMR (DMSO-d6): δ
3.57 (s, 3H), 3.58 (s, 6H), 3.71 (s, 3H), 6.39 (s, 2H), 7.82 (dd, J )
8.76 and 2.43 Hz, 1H), 6.90 (d, J ) 2.38 Hz, 1H), 7.38 (d, J )
8.76 Hz, 1H), 7.71 (d, J ) 2.70 Hz, 1H), 11.54 ppm (broad s,
disappeared on treatment with D2O, 1H). IR: ν 3356 cm-1. MS:
ES+ ) 368 (MNa+). Anal. (C18H19NO4S (345.42)) C, H, N, S.
5-Methoxy-2-methyl-3-[(3,4,5-trimethoxyphenyl)thio]-1H-in-
dole (23). Compound 23 was synthesized as 10 using 5-methoxy-
2-methyl-1H-indole, yield 29%, mp 138-142 °C (from ethanol).
1H NMR (DMSO-d6): δ 2.44 (s, 3H), 3.58 (s, 9H), 3.71 (s, 3H),
6.30 (s, 2H), 6.74 (dd, J ) 8.66 and 2.43 Hz, 1H), 6.84 (d, J )
2.20 Hz, 1H), 7.27 (d, J ) 8.68 Hz, 1H), 11.48 ppm (broad s,
disappeared on treatment with D2O, 1H). IR: ν 3339 cm-1. MS:
ES+ ) 382 (MNa+). Anal. (C19H21NO4S (359.45)) C, H, N, S.
5-Ethoxy-3-[(3,4,5-trimethoxyphenyl)thio]-1H-indole (24). Com-
pound 24 was synthesized as 10 using 5-ethoxy-1H-indole (32),
5-Ethoxy-1H-indole (32). Iodoethane (1.23 g, 0.63 mL, 0.0079
mol) and potassium carbonate (1.46 g, 0.01 mol) were added to a
solution of 5-hydroxy-1H-indole (0.7 g, 0.0053 mol) in acetone
(49 mL). The reaction mixture was refluxed overnight. Iodoethane
(1.23 g, 0.63 mL, 0.0079 mol) and potassium carbonate (1.46 g,
0.01 mol) were added, and the reaction mixture was stirred at the
same temperature for an additional 12 h. After cooling, the reaction
mixture was filtered, and the resulting solution was diluted with
ethyl acetate (30 mL) and washed with 3 N NaOH. The organic
layer was washed with brine and dried. Evaporation of the solvent
gave a residue that was purified by silica gel column chromatog-
raphy (chloroform as eluent) to furnish 32, yield 64%, yellow oil.
1H NMR (CDCl3): δ 1.49 (t, J ) 6.98 Hz, 3H), 4.12 (q, J ) 6.98
Hz, 2H), 6.50-6.52 (m, 1H), 6.91 (dd, J ) 8.78 and 2.12 Hz, 1H),
7.16-7.18 (m, 2H), 7.28 (d, J ) 8.78 Hz, 1H), 8.08 ppm (broad s,
1
yield 31%, brown oil. H NMR (CDCl3): δ 1.41 (t, J ) 6.98 Hz,
3H), 3.68 (s, 6H), 3.79 (s, 3H), 4.04 (q, J ) 6.99 Hz, 2H), 6.38 (s,
2H), 6.92 (dd, J ) 8.79 and 2.43 Hz, 1H), 7.08 (d, J ) 2.37 Hz,
1H), 7.33 (d, J ) 8.78 Hz, 1H), 7.47 (d, J ) 2.68 Hz, 1H), 8.48
ppm (broad s, disappeared on treatment with D2O, 1H). IR: ν 3337
cm-1. MS: ES+ ) 382 (MNa+). Anal. (C19H21NO4S (359.45)) C,
H, N, S.
disappeared on treatment with D2O, 1H). IR: ν 3409 cm-1
.
5-Isopropoxy-1H-indole (33). Compound 33 was synthesized
as 32 using 2-iodopropane, yield 25%, yellow oil. 1H NMR
(CDCl3): δ 1.38 (d, J ) 6.08 Hz, 6H), 4.55 (m, 6.00-6.07 Hz,
1H), 6.48-6.50 (m, 1H), 6.88 (dd, J ) 8.75 and 2.36 Hz, 1H),
7.17-7.19 (m, 2H), 7.28 (d, J ) 8.14 Hz, 1H), 8.09 ppm (broad s,
5-Isopropoxy-3-[(3,4,5-trimethoxyphenyl)thio]-1H-indole (25).
Compound 25 was synthesized as 10 using 5-isopropoxy-1H-indole
1
(33), yield 31%, brown oil. H NMR (DMSO-d6): δ 1.20 (d, J )
6.00 Hz, 6H), 3.57 (s, 9H), 4.44-4.48 (m, 1H), 6.39 (s, 2H), 6.79
(dd, J ) 8.75 and 2.36 Hz, 1H), 6.86 (d, J ) 1.87 Hz, 1H), 7.36
(d, J ) 8.72 Hz, 1H), 7.69 (d, J ) 2.66 Hz, 1H), 11.51 ppm (broad
s, disappeared on treatment with D2O, 1H). IR: ν 3392 cm-1. MS:
ES+ ) 396 (MNa+). Anal. (C20H23NO4S (373.43)) C, H, N, S.
5-Hydroxy-3-[(3,4,5-trimethoxyphenyl)thio]-1H-indole (26).
Compound 26 was synthesized as 10 using 5-hydroxy-1H-indole,
yield 28%, 184-186 °C (from ethanol). 1H NMR (CDCl3): δ 3.34
(s, 9H), 6.35 (s, 2H), 6.67 (dd, J ) 8.62 and 2.30 Hz, 1H), 6.76 (d,
J ) 3.21 Hz, 1H), 7.27 (d, J ) 8.64 Hz, 1H), 7.63 (d, J ) 2.70
Hz, 1H), 8.83 (broad s, disappeared on treatment with D2O, 1H),
11.37 ppm (broad s, disappeared on treatment with D2O, 1H). IR:
ν 3340, 3279 cm-1. MS: ES+ ) 354 (MNa+). Anal. (C17H17NO4S
(331.39)) C, H, N, S.
disappeared on treatment with D2O, 1H). IR: ν 3412 cm-1
.
5-[2-(Benzyloxy)ethoxy]-1H-indole (34). A solution of diethyl
azodicarboxylate (40% in toluene, 0.64 g, 1.60 mL, 0.0037 mol)
was added dropwise to a mixture of 2-benzyloxyethanol (0.56 g,
0.53 mL, 0.0037 mol), 5-hydroxy-1H-indole (0.5 g, 0.0037 mol),
and anhydrous triphenylphosphine (0.97 g, 0.0037 mol) in anhy-
drous tetrahydrofuran (23 mL). The reaction mixture was refluxed
overnight. After evaporation of the solvent, water (15 mL) and ethyl
acetate (15 mL) were added; the organic layer was washed with
brine and dried. Removal of the solvent gave a residue that was
purified by silica gel column chromatography (chloroform as eluent)
to furnish 34, yield 82%, brown oil. 1H NMR (CDCl3): δ 3.89 (t,
J ) 4.92 Hz, 2H), 4.23 (t, J ) 4.93 Hz, 2H), 4.69 (s, 2H), 6.48-
6.50 (m, 1H), 6.93 (dd, J ) 8.79 and 2.44 Hz, 1H), 7.15 (d, J )
2.37 Hz, 1H), 7.18-7.19 (m, 1H), 7.30-7.43 (m, 6H), 8.10 ppm
5-[2-(Benzyloxy)ethoxy]-3-[(3,4,5-trimethoxyphenyl)thio]-1H-
indole (28). Compound 28 was synthesized as 10 using 5-(2-
(benzyloxy)ethoxy)-1H-indole (34), yield 35%, brown oil. 1H NMR
(CDCl3): δ 3.68 (s, 6H), 3.79 (s, 3H), 3.85 (t, J ) 4.88 Hz, 2H),
4.18 (t, J ) 4.83 Hz, 2H), 4.65 (s, 2H), 6.38 (s, 2H), 6.98 (dd, J )
8.80 and 2.43 Hz, 1H), 7.11 (d, J ) 2.32 Hz, 1H), 7.30-7.40 (m,
6H), 7.48 (d, J ) 2.67 Hz, 1H), 8.42 ppm (broad s, disappeared on
treatment with D2O, 1H). IR: ν 3334 cm-1. MS: ES+ ) 488
(MNa+). Anal. (C26H27NO5S (465.57)) C, H, N, S.
2-[3-[(3,4,5-Trimethoxyphenyl)thio]-1H-indol-5-yloxy]etha-
nol (27). (2-Bromoethoxy)-tert-butyldimethylsilane (0.17 g, 0.16
mL, 0.724 mmol) and potassium carbonate (0.1 g, 0.72 mmol) were
added to a solution of 26 (0.2 g, 0.603 mmol) in acetonitrile (30
mL). The reaction was refluxed overnight. (2-Bromoethoxy)-tert-
butyldimethylsilane (0.17 g, 0.16 mL, 0.724 mmol) and potassium
carbonate (0.1 g, 0.72 mmol) were added, and the reaction was
refluxed for an additional 12 h. After cooling, water (10 mL) was
added, and the reaction mixture was extracted with ethyl acetate;
(broad s, disappeared on treatment with D2O, 1H). IR: ν 3409 cm-1
.
Biology. Tubulin Assembly. The reaction mixtures contained
0.8 M monosodium glutamate (pH 6.6 with HCl in 2 M stock
solution), 10 µM tubulin, and varying concentrations of drug.
Following a 15 min preincubation at 30 °C, samples were chilled
on ice, GTP to 0.4 mM was added, and turbidity development was
followed at 350 nm in a temperature controlled recording spectro-
photometer for 20 min at 30 °C. The extent of the reaction was
measured. Full experimental details were previously reported.31
[3H]Colchicine Binding Assay. The reaction mixtures contained
1.0 µM tubulin, 5.0 µM [3H]colchicine, and 5.0 µM inhibitor and
were incubated 10 min at 37 °C. Complete details were described
previously.32
MCF-7 Cell Growth. The above paper32 can also be referenced
for methodology of MCF-7 cell growth.

Arylthioindole Inhibitors of Tubulin Polymerization
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 12 2873
Binding Constants of the Ligands. The binding constants of
the ligands were measured at 20 °C either by displacement of
compound 3527 in a Shimadzu RF540 fluorimeter, with 5 nm
excitation and emission slits and with excitation at 350 nm and
emission at 422 nm. To check if any fluorescence or inner filter
effect could interfere with the assay results, the spectra of all
compounds dissolved in ethanol were determined in a Hitachi
U-2000 spectrophotometer. Only compound 30 showed absorbance
at the excitation wavelength (350 nm) of compound 35, as well as
emission at 422 nm. Therefore, its binding constant was measured
by quenching of the intrinsic fluorescence of tubulin.28,29 The data
were analyzed using the software package Equigra v5.25
Cell Culture. HeLa cells were grown at 37 °C in a humidified
atmosphere containing 5% CO2 in DMEM (GIBCO BRL, U.K.)
supplemented with 10% fetal calf serum (Hyclone, NL), 100 U/mL
of penicillin and streptomycin, and 2 mM glutamine (all reagents
were from Celbio, Italy). Cells were trypsinized when subconfluent,
seeded in T75 flasks at a concentration of 2.5 × 105 cells/mL in
complete medium and treated with compound 16 at 0.1-10 µM
for 24 h. Cells were harvested either immediately at the end of the
treatment or after a further incubation in drug-free medium for 24
h. In some experiments, HeLa cells were treated with 100 µM
etoposide for 3 h, followed by a 24 h recovery period.
scoring.svl script35 was used to identify interaction types between
ligand and protein.
Acknowledgment. G. La R. thanks Istituto Pasteur -
Fondazione Cenci Bolognetti for his Borsa di Studio per
Ricerche all’Estero. G. De M. thanks Italian Miur for her
Progetto Mobilita` Studiosi Italiani all’Estero. This research was
funded by Istituto Pasteur - Fondazione Cenci Bolognetti and
Universita` di Roma “La Sapienza”, Ricerche di Ateneo. Authors
also thank FIRC (Federazione Italiana per la Ricerca sul Cancro)
for its contribution. This work was supported by Grants
BFU2004-00358 from the Direccio´n General de Investigacio´n
Cient´ıfica y Tecnolo´gica (DGICYT) and CAM200520M061
from Comunidad Autonoma de Madrid. This research was also
partially supported by a grant of the Fondazione Banca del
Monte di Lombardia (Pavia, Italy) to A.I.S. For S.K. we would
like to acknowledge the Embassy of the Arab Republic of Egypt
for the award of a PhD scholarship.
Supporting Information Available: Elemental analyses of new
derivatives 10-28. This material is available free of charge via
Internet at http//pubs.acs.org.
Viability and Morphology Assays. Viability was assessed by
staining cells with trypan blue. Permeable cells were counted in a
hemocytometer and considered as nonviable. To evaluate cell
morphology, cells grown on glass coverslips were fixed for 10 min
in ice-cold 70% ethanol, washed several times with ice-cold PBS,
and stained for 10 min at room temperature with 0.1 µg/mL Hoechst
33258 (Sigma). Samples were washed with PBS, mounted on a
glass slide in a drop of Mowiol (Calbiochem, Inalco, Italy), and
observed by fluorescence microscopy.
Cell Cycle Analysis. Cells were detached by careful trypsiniza-
tion (to obtain single-cell suspensions to be processed for flow
cytometry), then resuspended in cold 0.9% NaCl and fixed with
cold 70% (final concentration) ethanol. Cells were stained in a
solution of PBS containing 30 µg/mL propidium iodide and 2 mg/
mL RNase A (Sigma) and analyzed with an Epics XL flow
cytometer (Beckman-Coulter Corp., U.S.A.). At least 10 000 cells/
sample were measured.
Evaluation of apoptosis. To investigate DNA degradation, cells
were rinsed twice in cold PBS containing 5 mM EDTA. Genomic
DNA was extracted from 2.5 × 106 cells and analyzed by agarose
gel electrophoresis.33 PARP-1 proteolysis was used as a marker of
caspase activation. Total extracts for western blots were prepared
from 2.5 × 106 cells. The western blot analysis was performed as
previously reported33 with the mAb C-2-10 against PARP-1
(Alexis, Vinci-Biochem, Italy). An HRP-conjugated antimouse IgG
antibody (Sigma) was used as the secondary antibody. Visualization
was performed by the ECL Detection System (Sigma).
Molecular Modeling. All molecular modeling studies were
performed on a RM Innovator with Pentium IV 3 GHz processor,
running Linux Fedora Core 4 using Molecular Operating Environ-
ment (MOE) 2006.0834 and the FlexX module in Sybyl 7.2.30 The
structure of two tubulin dimers, cocrystalized with a stathmin-like
domain and N-deacetyl-N-(2-mercaptoacetyl)-colchicine (DAMA-
colchicine), was downloaded from the PDB data bank (http://
www.rcsb.org/pdb/index.html; PDB code: 1SA0).1 Ligand struc-
tures were built with MOE and minimized using the MMFF94x
forcefield until a RMSD gradient of 0.05 kcal mol-1 Å-1 was
reached. The partial charges were automatically calculated and the
structure was saved as a mol2 file. Docking experiments were
carried out using the FlexX docking program of Sybyl 7.2. The
DAMA-colchicine present between chain C and chain D of the
structure was used to define the binding site, which included all
residues within 6.5 Å of the crystallized ligand. The GTP molecule
situated at the edge of the binding site on R-tubulin was included
as a heteroatom file. DAMA-colchicine was automatically removed
from the active site by the software before the docking simulation.
The output of FlexX docking was visualized in MOE, and the
References
(1) Ravelli, R. B.; Gigant, B.; Curmi, P. A.; Jourdain, I.; Lachkar, S.;
Sobel, A.; Knossow, M. Insight into tubulin regulation from a
complex with colchicine and a stathmin-like domain. Nature 2004,
428, 198-202.
(2) Nogales, E.; Whittaker, M.; Milligan, R. A.; Downing, K. H. High-
resolution model of the microtubule. Cell 1999, 96, 79-88.
(3) Nettles, J. H.; Li, H.; Cornett, B.; Krahn, J. M.; Snyder, J. P.;
Downing, K. H. The binding mode of epothilone A on alpha, beta-
tubulin by electron crystallography. Science 2004, 305, 866-869.
(4) Buey, R. M.; Calvo, E.; Barasoain, I.; Pineda, O.; Edler, M. C.;
Matesanz, R.; Cerezo, G.; Vanderwal, C. D.; Day, B. W.; Sorensen,
E. J.; Lopez, J. A.; Andreu, J. M.; Hamel, E.; Diaz, J. F. Cyclostreptin
binds covalently to microtubule pores and lumenal taxoid binding
sites. Nat. Chem. Biol. 2007, 3, 117-125.
(5) Lin, M. C.; Ho, H. H.; Pettit, G. R.; Hamel, E. Antimitotic natural
products combretastatin A-4 and combretastatin A-2: Studies on the
mechanism of their inhibition of the binding to colchicine to tubulin.
Biochemistry 1989, 28, 6984-6991.
(6) Beckers, T.; Mahboobi, S. Natural, semisynthetic and synthetic
microtubule inhibitors for cancer therapy. Drugs Future 2003, 28,
767-785.
(7) Sridhare, M.; Macapinlac, M. J.; Goel, S.; Verdier-Pinard, D.; Fojo,
T.; Rothenberg, M.; Colevas, D. The clinical development of new
mitotic inhibitors that stabilize the microtubule. Anticancer Drugs
2004, 15, 553-555.
(8) Jordan, M. A.; Wilson, L.; Microtubules as a target for anticancer
drugs. Nat. ReV. Cancer 2004, 4, 253-265.
(9) Mealy, N. E.; Balcells, L. M. Drugs under development for the
treatment of breast cancer. Drugs Future 2006, 31, 541-564.
(10) Davis, P. D.; Dougherty, G. J.; Blakey, D. C.; Galbraith, S. M.; Tozer,
G. M.; Holder, A. L.; Naylor, M. A.; Nolan, J.; Stratford, M. R. L.;
Chaplin, D. J.; Hill, S. A. ZD6126: A novel vascular-targeting agent
that causes selective destruction of tumor vasculature. Cancer Res.
2002, 62, 7247-7253.
(11) Chaplin, D. J.; Horsman, M. R.; Siemann, D. W. Current development
status of small-molecule vascular disrupting agents. Curr. Opin.
InVest. Drugs 2006, 7, 522-528.
(12) De Martino, G.; La Regina, G.; Coluccia, A.; Edler, M. C.; Barbera,
M. C.; Brancale, A.; Wilcox, E.; Hamel, E.; Artico, M.; Silvestri, R.
Arylthioindoles, potent inhibitors of tubulin polymerization. J. Med.
Chem. 2004, 47, 6120-6123.
(13) De Martino, G.; Edler, M. C.; La Regina, G.; Coluccia, A.; Barbera,
M. C.; Barrow, D.; Nicholson, R. I.; Chiosis, G.; Brancale, A.; Hamel,
E.; Artico, M.; Silvestri, R. Arylthioindoles, potent inhibitors of
tubulin polymerization. 2. Structure-activity relationships and mo-
lecular modeling studies. J. Med. Chem. 2006, 49, 947-954.
(14) Tron, G. C.; Pirali, T.; Sorba, G.; Pagliai, F.; Busacca, S.; Genazzani,
A. A. Medicinal chemistry of combretastatin A4: Present and future
directions. J. Med. Chem. 2006, 49, 3033-3044.
(15) Liou, J.-P.; Chang, C.-W.; Song, J.-S.; Yang, Y.-N.; Yeh, C.-F.;
Tseng, H.-Y.; Lo, Y.-K.; Chang, Y.-L.; Chang, C.-M.; Hsieh, H.-P.
Synthesis and structure-activity relationship of 2-aminobenzophen-
one derivatives as antimitotic agents. J. Med. Chem. 2002, 45, 2556-
2562.

2874 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 12
La Regina et al.
(16) Chang, J.-Y.; Yang, M.-F.; Chang, C.-Y.; Chen, C.-M.; Kuo, C.-C.;
Liou, J.-P. 2-Amino and 2′-aminocombretastatin derivatives as potent
antimitotic agents. J. Med. Chem. 2006, 49, 6412-6415.
(17) (a) Romagnoli, R.; Baraldi, P. G.; Pavani, M. G.; Tabrizi, M. A.;
Preti, D.; Fruttarolo, F.; Piccagli, L.; Jung, M. K.; Hamel, E.; Borgatti,
M.; Gambari, R. Synthesis and biological evaluation of 2-amino-3-
(3′,4′,5′-trimethoxybenzoyl)-5-aryl thiophenes as a new class of potent
antitubulin agents. J. Med. Chem. 2006, 49, 3906-3915. (b)
Romagnoli, R.; Baraldi, P. G.; Remusat, V.; Carrion, M. D.; Cara,
C. L.; Preti, D.; Fruttarolo, F.; Pavani, M. G.; Tabrizi, M. A.;
Tolomeo, M.; Grimaudo, S.; Balzarini, J.; Jordan, M. A.; Hamel, E.
Synthesis and biological evaluation of 2-(3′,4′,5′-trimethoxybenzoyl)-
3-amino 5-aryl thiophenes as a new class of tubulin inhibitors. J.
Med. Chem. 2006, 49, 6425-6428.
(18) Liou, J.-P.; Chang, Y.-L.; Kuo, F.-M.; Chang, C.-W.; Tseng, H.-Y.;
Wang, C.-C.; Yang, Y.-N.; Chang, J.-Y.; Lee, S.-J.; Hsieh, H.-P.
Concise synthesis and structure-activity relationships of combret-
astatin A-4 analogues, 1-aroylindoles and 3-aroylindoles, as novel
classes of potent antitubulin agents. J. Med. Chem. 2004, 47, 4247-
4257.
(19) Offer, J.; Boddy, C. N. C.; Dawson, P. E. Extending synthetic access
to proteins with a removable acyl transfer auxiliary. J. Am. Chem.
Soc. 2002, 124, 4642-4646.
(20) Flynn, B. L.; Flynn, G. P.; Hamel, E.; Jung, M. K. The synthesis
and tubulin binding activity of thiophene-based analogues of com-
bretastatin A-4. Bioorg. Med. Chem. Lett. 2001, 11, 2341-2343.
(21) Perez-Ramirez, B.; Shearwin, K. E.; Timasheff, S. N. The colchicine-
induced GTPase activity of tubulin: State of the product. Activation
by microtubule-promoting cosolvents. Biochemistry 1994, 33, 6253-
6261.
(22) Perez-Ramirez, B.; Andreu, J. M.; Gorbunoff, M. J.; Timasheff, S.
N. Stoichiometric and substoichiometric inhibition of tubulin self-
assembly by colchicine analogues. Biochemistry 1996, 35, 3277-
3285.
(23) Barbier, P.; Peyrot, V.; Leynadier, D.; Andreu, J. M. The active GTP-
and ground GDP-liganded states of tubulin are distinguished by the
binding of chiral isomers of ethyl 5-amino-2-methyl-1,2-dihydro-3-
phenylpyrido[3,4-b]pyrazin-7-yl carbamate. Biochemistry 1998, 37,
758-768.
(24) Buey, R. M.; Diaz, J. F.; Andreu, J. M.; O’Brate, A.; Giannakakou,
P.; Nicolaou, K. C.; Sasmal, P. K.; Ritzen, A.; Namoto, K. interaction
of epothilone analogs with the paclitaxel binding site; relationship
between binding affinity, microtubule stabilization, and cytotoxicity.
Chem. Biol. 2004, 11, 225-236.
(25) Buey, R. M.; Barasoain, I.; Jackson, E.; Meyer, A.; Giannakakou,
P.; Paterson, I.; Mooberry, S.; Andreu, J. M.; Diaz, J. F. Microtubule
interactions with chemically diverse stabilizing agents: Thermody-
namics of binding to the paclitaxel site predicts cytotoxicity. Chem.
Biol. 2005, 12, 1269-1279.
(26) Medrano, F. J.; Andreu, J. M.; Gorbunoff, M. J.; Timasheff, S. N.
Roles of colchicine rings B and C in the binding process to tubulin.
Biochemistry 1989, 28, 5589-5599.
(27) Medrano, F. J.; Andreu, J. M.; Gorbunoff, M. J.; Timasheff, S. N.
Roles of ring C oxygens in the binding of colchicine to tubulin.
Biochemistry 1991, 30, 3770-3777.
(28) Andreu, J. M.; Gorbunoff, M. J.; Lee, J. C.; Timasheff, S. N.
Interaction of tubulin with bifunctional colchicine analogues: An
equilibrium study. Biochemistry 1984, 23, 1742-1752.
(29) Andreu, J. M.; Timasheff, S. N. Conformational states of tubulin
liganded to colchicine, tropolone methyl ether, and podophyllotoxin.
Biochemistry 1982, 21, 6465-6476.
(30) Tripos, Sybyl 7.2; Tripos Inc.: 1699 South Hanley Rd, St. Louis,
Missouri 63144, http://www.tripos.com.
(31) Hamel, E. Evaluation of antimitotic agents by quantitative compari-
sons of their effects on the polymerization of purified tubulin. Cell
Biochem. Biophys. 2003, 38, 1-21.
(32) Verdier-Pinard, P.; Lai, J.-Y.; Yoo, H.-D.; Yu, J.; Marquez, B.; Nagle,
D. G.; Nambu, M.; White, J. D.; Falck, J. R.; Gerwick, W. H.; Day,
B. W.; Hamel, E. Structure-activity analysis of the interaction of
curacin A, the potent colchicine site antimitotic agent, with tubulin
and effects of analogs on the growth of MCF-7 breast cancer cells.
Mol. Pharmacol. 1998, 53, 62-76.
(33) Donzelli, M.; Bernardi, R.; Negri, C.; Prosperi, E.; Padovan, L.;
Lavialle, C.; Brison, O.; Scovassi, A. I. Apoptosis-prone phenotype
of human colon carcinoma cells with a high level amplification of
the c-myc gene. Oncogene 1999, 18, 439-448.
(34) Molecular Operating Environment (MOE 2006.08); Chemical Com-
puting Group, Inc.: Montreal, Quebec, Canada, http://www.chem-
comp.com.
(35) Code “scoring.svl” obtained from the SLV Exchange website http://
svl.chemcomp.com., Chemical Computing Group, Inc., Montreal,
Canada.
JM061479U