Novel derivatives of 1,3,4-oxadiazoles are potent mitostatic agents featuring strong microtubule depolymerizing activity in the sea urchin embryo and cell culture assays

Article abstract of DOI:10.1016/j.ejmech.2009.12.072

A series of novel 1,3,4-oxadiazole derivatives based on structural and electronic overlap with combretastatins have been designed and synthesized. Initially, we tested all new compounds in vivo using the phenotypic sea urchin embryo assay to yield a number of agents with anti-proliferative, anti-mitotic, and microtubule destabilizing activities. The experimental data led to identification of 1,3,4-oxadiazole derivatives with isothiazole (5-8) and phenyl (9-12) pharmacophores featuring activity profiles comparable to that of combretastatins, podophyllotoxin and nocodazole. Cytotoxic effects of the two lead molecules, namely 6 and 12, were further confirmed and evaluated by conventional assays with the A549 human cancer cell line including cell proliferation, cell cycle arrest at the G2/M phase, cellular microtubule distribution, and finally in vitro microtubule assembly with purified tubulin. The modeling results using 3D similarity (ROCS) and docking (FRED) correlated well with the observed activity of the molecules. Docking data suggested that the most potent molecules are likely to target the colchicine binding site.

Full text of DOI:10.1016/j.ejmech.2009.12.072

European Journal of Medicinal Chemistry 45 (2010) 1683–1697
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Novel derivatives of 1,3,4-oxadiazoles are potent mitostatic agents featuring
strong microtubule depolymerizing activity in the sea urchin embryo and cell
culture assays
,
b
c
d
c
Alex S. Kiselyova , Marina N. Semenova , Natalya B. Chernyshova , Andrei Leitao , Alexandr V. Samet ,
*
c
c
d
e
e
´
Konstantine A. Kislyi , Mikhail M. Raihstat , Tudor Oprea , Heiko Lemcke , Margareta Lantow ,
Dieter G. Weiss e, Nazli N. Ikizalp f, Sergei A. Kuznetsov e, Victor V. Semenov c
,
f
a deCODE Chemistry, 2501 Davey Road, Woodridge, Chicago, IL 60616, USA
b Institute of Developmental Biology, RAS 26 Vavilov Str., Moscow 119334, Russia
c Zelinsky Institute of Organic Chemistry, RAS 47 Leninsky Prospect, Moscow 117913, Russia
d Department of Biocomputing, University of New Mexico, Albuquerque, NM, USA
e Institute of Biological Sciences University of Rostock, 3 Albert-Einstein-Str., D-18059 Rostock, Germany
f Chemical Block Ltd., 3 Kyriacou Matsi, Limassol 3723, Cyprus
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of novel 1,3,4-oxadiazole derivatives based on structural and electronic overlap with com-
bretastatins have been designed and synthesized. Initially, we tested all new compounds in vivo using the
phenotypic sea urchin embryo assay to yield a number of agents with anti-proliferative, anti-mitotic, and
microtubule destabilizing activities. The experimental data led to identification of 1,3,4-oxadiazole
derivatives with isothiazole (5–8) and phenyl (9–12) pharmacophores featuring activity profiles
comparable to that of combretastatins, podophyllotoxin and nocodazole. Cytotoxic effects of the two lead
molecules, namely 6 and 12, were further confirmed and evaluated by conventional assays with the A549
human cancer cell line including cell proliferation, cell cycle arrest at the G2/M phase, cellular micro-
tubule distribution, and finally in vitro microtubule assembly with purified tubulin. The modeling results
using 3D similarity (ROCS) and docking (FRED) correlated well with the observed activity of the mole-
cules. Docking data suggested that the most potent molecules are likely to target the colchicine binding
site.
Received 19 August 2009
Received in revised form
23 December 2009
Accepted 29 December 2009
Available online 14 January 2010
Keywords:
anti-mitotic agent
1,3,4-Oxadiazoles
Combretastatin
Tubulin
Microtubules
Sea urchin embryo
Cancer
Ó 2010 Elsevier Masson SAS. All rights reserved.
1. Introduction
these mitotic poisons [12–16] prompted scientists to expand search
for synthetic modulators of tubulin that a) bind to alternative site(s)
Targeting tubulin in rapidly dividing tumor cells has been a well-
validated strategy for cancer therapy [1–3]. Screening of natural
products for cytotoxic activity yielded numerous structurally diverse
classes of mitotic spindle poisons [4–7]. Current evidence suggests
that these natural products affect tubulin polymerization dynamics
by binding to microtubules (MT) at discrete regions [8], such as well-
characterized taxoid, vinca alkaloid and colchicine sites [9–11].
Paclitaxel (TaxolÒ) and Vinblastine binding to the former two sites
have been successfully used in clinics as chemotherapeutics to treat
various tumor types. However, high toxicity, non-linear pharma-
cology together with the development of drug resistance found for
[17]; b) mimic natural products, are synthetically feasible and drug-
like, c) result in the same anti-mitotic effect, and d) display better
therapeutic window. Many natural and synthetic compounds with
different molecular structure have been identified as colchicine site
ligands [7,18–20]. Podophyllotoxin (Scheme 1) is another example of
the natural product that binds to the colchicine (or adjacent) site on
tubulin [11,20]. These colchicine site binders share ‘‘homology’’ with
the A and C rings of colchicine described as a tilted ‘‘biaryl’’ system
connected by a hydrocarbon bridge of variable length (ex., B and E
rings of podophyllotoxin, Scheme 1), although different binding
modes for the two molecules have been introduced (reviewed in
[20]).
Synthetic derivatives of combretastatins, CA-4 phosphate,
AVE8062, and ZD-6126 are in clinical trials as potent antineoplastic
agents [3,18,20].
* Corresponding author. Tel.: þ1 630 783 4951.
E-mail address: akiselyov@decode.com (A.S. Kiselyov).
0223-5234/$ – see front matter Ó 2010 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2009.12.072

1684
A.S. Kiselyov et al. / European Journal of Medicinal Chemistry 45 (2010) 1683–1697
R
OH
H
MeO
MeO
MeO
O
A
O
NHAc
O
B
C
E
D O
A
H
MeO
O
C
OMe
1
RO
MeO
OMe
OMe
2
OR
OMe
Scheme 1. Colchicine R ¼ OH, R1 ¼ R2 ¼ Me, Combretastatin A-4 (CA-4) Podophyllotoxin R ¼ OH,R1, R2 ¼ –CH2–, Combretastatin A-2 (CA-2)R ¼ OPO3Na2, R1 ¼ R2 ¼ Me, CA4-P.
2. Results and discussion
anti-mitotic, and cytotoxic effects of the oxadiazoles 1–25 on a living
organism. Inaddition, theassayallowedfortheexpeditiousrankingof
the molecules based on their relative activity. The sea urchin embryo
model conveniently differentiates tubulin-independent anti-prolif-
erativeactivityand cytotoxicity/embryotoxicityof testedcompounds.
During cleavage stages, intensive synchronous cell divisions are
observed, whereas processes of differentiation and morphogenesis
become dominant from gastrulation. Tubulin-independent anti-
proliferative/anti-mitotic agents cause cleavage alteration along with
the inability to affect the embryo development after hatching at the
same concentration range. In contrast, toxicity implies embryo
abnormalities observed after treatment of both fertilized eggs and
hatched blastulae. The intrinsic mechanisms of action of tubulin-
In our search for the synthetic anti-mitotic agents binding to
tubulin, we selected the 1,3,4-oxadiazole template as a starting point
[21]. Earlier analogs from these series featured good potency in the in
vitro tubilin polymerization assay, caused mitotic arrest in cells
derived from multi-drug resistant tumors. In this paper, we report on
the developmentofa robust and high-yieldingsynthetic procedure to
access novel 1,3,4-oxadiazole derivatives 1–25 that specifically affect
tubulin dynamics (Table 1). Furthermore, in order to evaluate anti-
mitotic activity of these small molecules we introduced reliable
phenotypic assay based on the sea urchin embryo [22,23]. This in vivo
screening yielded a rapid ‘‘one pot’’ assessment of anti-proliferative,
Table 1
Structure Activity Relationship Studies of 1,3,4-Oxadiazoles 1–25 in the Sea Urchin Embryo Assay.
N
N
R2
R1
O
1-25
Cmpd ID
R1
R2
EC50, m
Ma
Cleavage Alteration
Cleavage Arrest
Spinning
CA-2
CA-4P
Podophyllotoxin
Nocodazole
0.002
0.005
0.02
0.01
0.01
0.05
0.01
0.05
1
0.5
0.1
0.005
H2N
H2N
O
1
2
0.5b
4
>5
N
N
O
NH
S
S
O
NT
NT
NT
NH
NH
O
H2N
Cl
Cl
3
4
0.2
0.5
0.5
0.5
N
S
H
N
NH
2
>2.5
N
S
N

A.S. Kiselyov et al. / European Journal of Medicinal Chemistry 45 (2010) 1683–1697
1685
Table 1 (continued)
Cmpd ID
R1
R2
EC50, m
Ma
Cleavage Alteration
Cleavage Arrest
0.005
Spinning
H
N
O
O
5
6
0.002
0.02
0.1
N
N
N
NH
S
S
S
N
H
N
O
0.002
0.005
NH
NH
O
N
H
N
O
7
8
0.005
0.005
0.05
0.01
0.05
0.1
O
N
9
0.001
0.005
0.005
0.002
0.002
0.005
0.005
0.005
0.005
10
11
12
0.001
0.0005
0.0005
13
14
0.001
0.001
0.01
0.02
1
0.1
15
16
0.025
0.05
0.1
0.2
0.2
1
(continued on next page)

1686
Table 1 (continued)
A.S. Kiselyov et al. / European Journal of Medicinal Chemistry 45 (2010) 1683–1697
Cmpd ID
R1
R2
EC50
Cleavage Alteration
, m
Ma
Cleavage Arrest
Spinning
0.2
17
0.05
0.5
18
0.4b,c
1
>5
19
20
4b
>4
>5
NH
O
O
S
N
N
2
5
5
NH
NH2
O
21
0.05
2
>2
O
NH
NH
O
O
N
22
23
1b
>10
>10
NH
NH
N
N
N
H
NH
O
N
N
1b
>10
>10
O
N
H
NH
N
O
O
N
24
0.2b
1
>2.5
NH
NH
Me
NH2
O
HN
25
>20
>20
>20
O
N
a
The sea urchin embryo assay was conducted as described in [22], duplicate measurements showed no differences in threshold concentration values.
Anti-proliferative activity that is non-tubulin related.
b
c
Molecule displayed post-hatching embryo toxicity at ꢀ1
mM when applied after hatching (see discussion below). NT – not tested.
independent anti-proliferative/anti-mitotic compounds can be
further elucidated in detail using the appropriate screening systems
[22]. Cytotoxic properties of selected compounds were further
evaluated by conventional cell-based assays with human cancer cells,
and by inhibition of purified tubulin polymerization in vitro. To
rationalize the tubulin activity of 1,3,4-oxadiazole derivatives, we

A.S. Kiselyov et al. / European Journal of Medicinal Chemistry 45 (2010) 1683–1697
1687
conducted a series of three-dimensional (3D) similarity comparisons.
Our computational insight suggested that the most active 1,3,4-oxa-
diazoles capture the overall configuration of combretastatins.
This simple in vivo model facilitates rapid ‘‘one pot’’ assessment of
anti-proliferative, anti-mitotic, and cytotoxic effects of small
molecules on a living organism. The assay includes i) fertilized egg
test for anti-mitotic activity displayed by the embryo’s cleavage
alteration/arrest, and ii) behavioral monitoring of a free-swimming
blastulae treated immediately after hatching (9–12 h after fertil-
ization). Changes to the embryo swimming pattern, i.e. lack of
forward movement, settlement to the bottom of the culture vessel
and rapid spinning around the animal–vegetal axis suggests
a tubulin destabilizing activity caused by a molecule [22]. Notably,
this experiment yields data on both anti-proliferative activity of
the test molecule as well as on its tubulin polymerization effect
(ex., destabilizing vs stabilizing) (Fig. 1). Data generated by the
assay for multiple marketed and experimental anti-mitotics
correlated well with their in vitro, ex vivo and in vivo potency and
efficacy [22,23].
We have evaluated synthetic molecules 1–25 using the sea
urchin embryo test system. Stock solutions of compounds were
prepared either in 95% EtOH at 5 mM or in DMSO at 5–10 mM
concentrations followed by their 10-fold dilution with 95% EtOH. In
our hands, this protocol enhanced solubility of the test compounds
in salt-containing medium (sea water), as evidenced by micro-
scopic examination of the samples. Further, we found that the
maximal tolerated concentrations of DMSO and EtOH in the in vivo
assay were 0.05% and 1% respectively. Higher concentrations of
either DMSO (ꢀ0.1%) or EtOH (>1%) caused non-specific alteration
and retardation of the sea urchin embryo development indepen-
dent of the treatment stage. The effects of tested molecules were
quantified by a threshold concentration resulting in cleavage
abnormalities and embryo death before hatching or full mitotic
arrest (Table 1). CA-2, CA-4P (OxiGene), podophyllotoxin (Sigma),
and nocodazole (Sigma) served as benchmark reference
compounds.
2.1. Chemical synthesis of 1,3,4-oxadiazoles
We used two related synthetic approaches to access the targeted
1,3,4-oxadiazoles, as shown on Scheme 2. Following route 1, the
isothiazole derivative was treated with neat hydrazine hydrate at
reflux for 4 h to yield the respective hydrazide in 70% yields. In order
to assemble the 1,3,4-oxadiazole ring, the hydrazide was reacted
with isothiocyanates Ar1-NCS followed by in situ cyclization of the
intermediate thiosemicarbazides with DCC to afford the key
precursor molecules 1–3 in 56–80% isolated yields. Pharmacophore
Ar2 was introduced into the amines 1–3 via the initial formation of
the respective Schiff bases followed by their NaBH4-mediated
reduction to furnish the targeted compounds 5, 7 and 8 (84–88%
yield). Notably, our attempts to accomplish a one-pot reductive
amination of 1–3 under a variety of experimental conditions
consistently afforded lower yields (by ca. 15–30%) of the final
heterocycles. Furthermore, we were unable to synthesize 6 following
the procedure described above (Scheme 2, Route 1). In order to
address this issue, we devised the alternative route 2 (Scheme 2).
It involved a reversed synthetic sequence starting with the
introduction of Ar2 followed by the installation of the 1,3,4-oxa-
diazole moiety to yield the targeted compounds 4 and 6 (75–80%
yield). Additional test molecules were synthesized using related
synthetic sequences involving the formation of the diverse
aromatic or heterocyclic ortho-amino hydrazides as described in
the Supporting Information.
2.2. Biological activity
As evidenced from Table 1, CA-2, CA-4P, podophyllotoxin, and
nocodazole displayed strong effects on the sea urchin embryos.
Effective concentrations of the reference compounds resulting in
embryo cleavage alteration/arrest were close to the reported IC50
values for a panel of human tumor cell lines [20,22,24]. It is worth
2.2.1. Evaluation of 1,3,4-oxadiazoles in the phenotypic sea urchin
embryo assay
We have recently introduced
a reliable and reproducible
phenotypic sea urchin embryo assay that allows for the selection of
potent mechanistically sound compounds targeting tubulin [22].
NH2
NH2
N
NH2
N
N
N
H
N
ii
i
CONHNH2
CONH2
O
(56-80%)
(70%)
Ar1
S
N
S
S
1, Ar1 = 3,4-(OCH2O)-C6H4-
2, Ar1 = 3,4-(OCH2CH2O)-C6H4-
3, Ar1 = 4-Cl-C6H4-
Route 1: used to synthesize 5, 7, 8
iii (59%)
iii
Ar2
NH
NH
N
N
N
H
N
Ar1
i, ii
CONHNH2
O
S
N
S
N
4, Ar1 = 4-Cl-C6H4-; Ar2 = 3-Py (80%)
5, Ar1 = 3,4-(OCH2O)-C6H3- ; Ar2 = 4-Py (84%)
6, Ar1 = 3,4-(OCH2CH2O)-C6H3- ; Ar2 = 4-Py (75%)
7, Ar1 = 3,4-(OCH2O)-C6H3- ; Ar2 = 3-Py (88%)
8, Ar1 = 3,4-(OCH2CH2O)-C6H3- ; Ar2 = 3-Py (84%)
Route 2: used to synthesize 4, 6
Scheme 2. Reagents and conditions: i) neat N2H4 $H2O, reflux, 4 h; ii) R1–C6H4–NCS, tBuOH, reflux 10 min; 50–60 ꢁC, 1h; DCC, MeCN, reflux, 3 h; iii) R2CHO (R ¼ 3-Py or
4-Py),MeCN, pTsOH, reflux, 3 h; NaBH4, iPrOH, reflux, 3h.

1688
A.S. Kiselyov et al. / European Journal of Medicinal Chemistry 45 (2010) 1683–1697
Fig. 1. Typical effects of tubulin-modulating compounds on early sea urchin embryo development. (A, B) Intact embryos. (A) 8-cell stage. (B) Early blastula. Fertilized eggs (10–
15 min after fertilization) were exposed continuously to compound 5 at a threshold concentration for cleavage alteration (2 nM; C, D); to compound 3 at threshold concentration for
cleavage arrest (0.5
a stabilized mitotic apparatus (light spots; G, H). The average embryo diameter is 115
mM; E, F); to paclitaxel (5 m
M; G, H). Time after fertilization (21 ꢁC): (A, C, E, G) - 2.5 h; (B, D, F, H) – 6 h. Note the formation of a tuberculous egg (F) and
mm.

A.S. Kiselyov et al. / European Journal of Medicinal Chemistry 45 (2010) 1683–1697
1689
noting that all tested molecules 1–25 exhibited comparable solu-
2.2.2. Cell growth inhibitory activity
bility in the assay system. Oxadiazole derivatives featuring both
isothiazole (5–8) and aryl (9–14) pharmacophores as R1 caused
significant cleavage alteration, arrest and spinning comparable to the
reference compounds (Table 1). Molecule 1 showed lower potency in
our assay system. Moreover, at low concentrations it featured non-
tubulin anti-proliferative activity whereas at higher concentrations
First the inhibitory effect of selected compounds on proliferation
of A549 lung carcinoma cells was determined with the MTT (3-(4,5-
dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) cytotox-
icity assay [26]. The IC50 values for 1, 6, 12–14, and nocodazole are
shown in Table 2.
Compounds containing benzene (12) and thiazole (6) rings
demonstrated strong cytotoxicity comparable with nocodazole
(Table 2). Analogues 13 and 14 lacking the –CH2–Py moiety showed
a markedly lower cell growth inhibitory effect. Reduced activity of
these molecules could be attributed to their poor cell permeability.
(ꢀ2
mM) it caused embryo abnormalities and embryotoxicity at later
developmental stages. 1,3,4-Oxadiazole 3 (R2 ¼ 4-Cl–C6H4–NH–)
afforded good activity (Fig. 1E, F). However, the product of respective
reductive amination with 3-pyridylaldehyde 4 was considerably less
active. Most active 1,3,4-oxadiazoles that affected tubulin polymer-
ization were consistently substituted with a piperonyl- or ethyl-
enedioxyphenyl moiety in the R1 substituent (ex. 5–12, Table 1).
Replacement of these oxoaryl groups caused a deleterious effect on
the activity of resulting compound (ex., molecules 7 and 8 vs 4). For
monosubstituted R2 ¼ 4-R–C6H4–NH–, where R ¼ F, Me, CF3, OCF3 or
disubstituted R2 ¼ 3,4-diF–C6H4–NH–, 3,4-diCl–C6H4–NH–, 3-Cl–4-
OMe–C6H4–NH– the respective 1,3,4-oxadiazole derivatives did not
Compound 1 was inactive up to 20 mM concentration. We were
unable to test higher concentrations of 1 due to its limited solubility
in the cell culture medium.
2.2.3. Cell cycle arrest in G2/M
In our second approach, the ability of compounds 1, 6, 12–14 to
alter A549 cell cycle progression was estimated by propidium
iodide staining followed by flow cytometry analysis. The EC50
values are shown in Table 2. Fig. 2 presents the detailed results of
the DNA-based flow cytometric analysis of A549 cells for two lead
molecules 6 and 12, and nocodazole as a positive control. All three
compounds induced concentration-dependent accumulation of
cells in the G2/M phase, with 12 showing the strongest effect, fol-
lowed by 6 and nocodazole (see also Table 2). Notably, both cell-
based assays ranked the test articles similarly: cytotoxicity of oxa-
diazoles and their ability to induce G2/M arrest decreased in the
following order: 12 > 6 > 14 > 13 > 1.In both tests compound 1 did
not produce an effect on A549 cell proliferation.
show any activity in the assay (% ITP ¼ 20–30% at 10
mM) (data not
shown [25]).
In general, a variety of substituents at R1 were tolerated. For
example, both unsubstituted anilines and their derivatives featured
strong potency in vivo (Table 1, 9–12 and 13–14). Isothiazole
derivatives 5-8 displayed activities comparable to those of the
respective phenyl derivatives 9–12. For the ‘‘linear’’ derivatives of
1,3,4-oxadiazoles (15–18), substituents at R1 reduced overall
activity of the molecules. In addition, embryo abnormalities were
observed for the compound 18 at concentrations ꢀ 1
mM. An
introduction of 3-(2-aminoquinoline) group at R1 also resulted in
a molecule 21 with reduced anti-tubulin activity.
2.2.4. The effect on microtubule distribution in cells
Derivatives featuring pyrazole and imidazole rings at R1 (22–24)
showed no MT-based anti-proliferative activity, as concluded from
both phenotype of exposed embryos and lack of embryo spinning.
Molecule 21 was classified as a relatively weak tubulin modulator.
Specifically, as evidenced from the detailed microscopy analysis it
caused cleavage alterations and arrest with formation of tubercu-
lous eggs typical of tubulin destabilizers. An unsubstituted pyrazole
pharmacophore in R1 position as in 25 led to a complete loss of
anti-proliferative activity. Oxadiazole 19 lacking amine bridge was
In our third assay, we evaluated the effect of selected
compounds on microtubule (MT) stability and distribution in
cultured A549 cells. Cells were incubated with 1, 6, 12–14, and
nocodazole as a positive control. Then MTs were labelled by the
indirect immunofluorescence method and analysed by fluores-
cence microscopy. 0.6% DMSO in water served as a control (Fig. 3)
Molecules 12 and 6 displayed the most pronounced MT dis-
rupting effect of all molecules tested (Fig. 3C, D). Specifically, cells
incubated with 12 did not feature microtubules as compared with
nocodazole. Similarly, 6 ehxibited strong tubulin destabilizing
effect, however small MT fragments were still discernable. Both 13
and 14 showed little or no effect on MT network at concentration of
only marginally active in the sea urchin embryo assay (4
mM).
Additional molecules featuring R2 ¼ 4-F–C6H4–NH–, 4-Cl–C6H4–
NH–, 4-MeO–C6H4–NH– and 3,4-(OCH2O)–C6H4–NH–, also dis-
played very low activity (data not shown).
5
m
M (Fig. 3E, F). At higher concentrations (25
did affect the MT network (Fig. 3G, H). Compound 1 did not affect
the MTs in A549 cells even at the 20 M concentration (not shown).
mM) these molecules
Based on our in vivo data, we have selected the most active
tubulin destabilizers represented by thiazole (6) and benzene (12)
derivatives as well as their precursors 13 and 14, and the non-
tubulin anti-mitotic compound 1 as a comparator for further
evaluations using in vitro tubulin polymerization and cell-based
assays. Our rationale was to correlate the effects of lead candidates
identified via the sea urchin embryo assay with the data generated
by conventional studies of anti-tubulin activity. Biological activi-
ties of these agents were tested using four experimental
approaches.
m
2.2.5. Inhibition of purified tubulin polymerization
Finally, we tested molecules prioritized by the in vivo experi-
ments using in vitro tubulin polymerization assay. Specifically,
purified bovine brain tubulin was polymerized in the presence of
10% DMSO followed by incubation with 10
mM of 1, 6, 12–14,
nocodazole and additional 0.6% DMSO in water. Within 1 h post
treatment, the resulting preparations were analysed by video-
Table 2
Cytotoxicity of Selected Oxadiazoles and their Effect on the Cell Cycle.
DMSOa
Nocodazolea
1b
6
12
13
14
Cytotoxicity IC50 ꢂ SD, nMc
G2/M arrest EC50 ꢂ SD, nMd
No effect
No effect
36.6 ꢂ 5.8
>20,000
>20,000
53.3 ꢂ 11.5
63.0 ꢂ 4.24
15.0 ꢂ 0.0
44.5 ꢂ 0.7
3000 ꢂ 866
8550 ꢂ 353
1566.6 ꢂ 115.5
2825 ꢂ 318.2
106.5 ꢂ 9.2
a
DMSO (0.4%) was used as a negative control, and nocodazole as a positive control.
b
c
At concentrations higher than 20
Average of three experiments.
Average of two experiments.
mM compound 1 formed crystals.
d

1690
A.S. Kiselyov et al. / European Journal of Medicinal Chemistry 45 (2010) 1683–1697
Fig. 2. The effect of selected oxadiazoles on A549 cell cycle. (A) DNA-based flow cytometry histograms of nocodazole (a), 12 (b), and 6 (c). (B) A concentration-dependent cell cycle
arrest in the G2/M phase caused by nocodazole (a), 12 (b), and 6 (c). 0.4% DMSO in water was used as a control.
enhanced contrast differential interference contrast (VEC-DIC)
microscopy (Fig. 4).
and 14, in these series the absence of –CH2–Py group (ex., 1) yielded
complete loss of the MT destabilizing effect. Respective pyrazole or
imidazole derivatives (22–24) showed significant non-tubulin anti-
mitotic activity.
As shown on Fig. 4, molecule 12 completely inhibited the MT
assembly, this effect being more profound than that of nocodazole.
Specifically, samples treated with 6 displayed more aggregates or
short MT’s than nocodazole. For the compound 14, a number of
depolymerized MT’s have been detected, albeit the effect was
moderate compared to both 12 and nocodazole. Molecule 13
exerted only marginal effect on MT dynamics.
2.2.6. Computational studies of 1,3,4-oxadiazole derivatives
To rationalize the tubulin activity of 1,3,4-oxadiazole deriva-
tives, we conducted a series of three-dimensional (3D) similarity
comparisons [27]. These studies utilized reference compounds that
were docked into the colchicine binding site at the interface of
Both cell-based and in vitro MT assays converged on the mole-
cule 12. This agent consistently featured the highest activity when
compared with nocodazole. Compound 6 demonstrated strong
tubulin destabilizing effect as well. Analogs 13 and 14 were less
active, while compound 1 characterized as non-tubulin agent by
sea urchin embryo assay failed to depolymerize MT’s both in whole
cells and in vitro. Data generated by the cell culture assays corre-
lated well with the results of the sea urchin embryo studies.
Specifically, derivatives of oxadiazoles endowed with benzene
substituents (ex., 12) were the most potent. Replacement of the
–CH2–Py moiety (13 and 14) led to both decreased anti-prolifera-
tive activity in A549 cells as well as to a weaker tubulin destabi-
lizing properties in vivo and in vitro. The ability of 13 and 14 to
produce cleavage alteration/arrest at nanomolar concentrations
could be explained by the high frequency of cell divisions that occur
every 35–40 min at early stages of the sea urchin embryo devel-
opment. Oxadiazoles containing isothiazole moiety (ex., 6) dis-
played comparable activity across the assay panel. In contrast to 13
a/b tubulin subunits. The reference set comprised: i) known tubulin
binders colchicine and its sulfur analogue DAMA-colchicine [11],
podophyllotoxin [11], nocodazole [5,6], CA-2, CA-4 [28], and taxol
[10]; ii) compounds with unknown binding site, namely indibulin
[29] and benzoquinolizidine [30] (chemical structures provided in
the Supporting Information, Fig. S1). A pool of conformers from the
reference set was generated using Omega [31] and ROCS [32] (both
from OpenEye Scientific Software Ltd). It was further used to select
poses matching the reported co-crystal structure of colchicine and
tubulin. This procedure was repeated for benzoquinolizidine,
indibulin and oxadiazole series although there were no data sug-
gesting that these molecules bind tubulin at the colchicine binding
site. Previous work showed that combretastatins compete for
tubulin binding with colchicines. This data suggest overlapping
binding sites for these molecules [28,33]. ROCS results implied that
two compounds (13 and 14) had higher similarity to CA-2 and CA-4
(Table 3).

A.S. Kiselyov et al. / European Journal of Medicinal Chemistry 45 (2010) 1683–1697
1691
Fig. 3. Effect of selected compounds on the MT networks in A549 lung carcinoma cells analysed by indirect immunofluorescence microscopy. Cells were treated with 0.6% DMSO in
water: (A) as a negative control, with 5 M of nocodazole as positive control (B), and following test compounds for 8 h: 12 (5 M, C), 6 (5 M, D), 14 (5 M, E; 25 M, G), 13 (5 M, F;
25 M, H). Bar: 20 m.
m
m
m
m
m
m
m
m
Molecules 6 and 12 were also potent MT dynamics inhibitors in
the biological assays. However, their similarity score was affected
by bulky substituents absent in combretastatins (Fig. 5).
colors: combretastatin A-2 (green), docked compounds (gray).
Oxygen, nitrogen, sulfur and hydrogen are red, blue, yellow and
white, respectively. Non-polar hydrogens were omitted for the sake
of clarity.
Because ROCS is a ligand-based tool that does not take into
account protein constraints, compound 12 appeared to be mis-
aligned in relation to other molecules featuring 4-Py–CH2 substit-
uent. Several studies suggested this substitution region to be
unfavorable due to its steric clash with amino acid residues of the
tubulin binding pocket (reviewed in [19]). We concluded that the
inverted alignment of 12 could yield better fitting yielding high
ROCS similarity scores (0.99 vs CA-2, and 0.95 vs CA-4, respec-
tively). Black arrows in Fig. 5 indicate potential H-bond donors. The
aromatic rings were out-of-plane (twisted) due to the mismatch
between amine and vinyl spacer groups (orange arrow in Fig. 5)
reducing the overall similarity score of this dataset. Considering
tubulin activity of the studied molecules, we concluded that this
slight twisting of the aromatic rings was tolerated. The binding
mode for 6 and 12 matched the consensus pharmacophore derived
from a set of tubulin inhibitors that bind the colchicine site [34].
Blue and black arrows show superposition of hydrophobic and
hydrogen bonding interactions, respectively. Orange arrow indi-
cates mismatch of the spacer group. Labels A and B in 11 illustrate
overlapping rings of CA-2 and this oxadiazole derivative. Carbon
Following the reported SAR data for combretastatin derivatives
[19,35] we propose that the aniline linker bridge affects the align-
ment of phenyl rings and keeps them oriented in a twisted manner
similar to other tubulin inhibitors (Fig. 6) [36].
In order to rationalize the observed SAR, we further docked
several representative molecules (1, 4, 7, 12, Fig. 6) into the tubulin
cavity (PDB ID 1SA0) using FRED [37]. These molecules appeared to
bind in a specific manner, fitting R2 moiety (Table 1) into the
deepest region of tubulin cavity, while the oxadiazole and R1
substituents were oriented towards the cavity entrance.
In our hands, data for cytotoxicity, G2/M arrest, MT polymeri-
zation and the sea urchin embryo assays converged on 12 as
a molecule with the activity profile similar to nocodazole. It was
followed by compounds 6, 14 and 13 in the decreasing order of
potency. Compound 1 was inactive in most of the assays, as it did
not display anti-tubulin properties. The molecule did cause
cleavage arrest in sea urchin embryo. However, this activity was
likely driven by a non-tubulin mechanism as suggested in the
literature for several CA analogues with low anti-tubulin activity

1692
A.S. Kiselyov et al. / European Journal of Medicinal Chemistry 45 (2010) 1683–1697
Fig. 4. Effect of selected compounds on purified tubulin assembly in vitro analysed by VEC-DIC microscopy. Samples treated with additional 0.6% DMSO in water as a negative
control (A) showed single MTs. Samples treated with 10 M of nocodazole (B), 12 (C), 6 (D), 14 (E) or 13 (F) showed lack of intact MTs of MT bundles. Bar: 17 m.
m
m
[28]. Lack of the dioxolane moiety in R2 was detrimental to the anti-
mitotic activity of molecules (compare 4, 7 and 8, Table 1). In
analyzing results of the docking studies, we reasoned that the
dioxolane group of 7 and 12 potentially overlapped with the tri-
methoxy substituents of both colchicine and CA-2. The binding
mode for 4 was likely inverted. None of the identified positions
provided reasonable overlap of the para-chlorophenyl group with
the trimethoxyphenyl substituent of CA-4 (Fig. 6). Direct linking of
the phenyl rings in CA-4 derivatives reduced the activity of these
molecules. This finding suggests potential steric hindrance at the
inner part of colchicine binding site of tubulin, where the R2 group
is likely to bind. Moreover, our earlier report showed that the
hydrogen bond acceptor at para position was critical to the anti-
mitotic activity [36].
Docking poses further suggested a potential hydrogen bond
between the Cys238 and the dioxolane ring (Fig. 6). As opposed to
the strict steric requirements for R2, the substitution pattern of R1
pharmacophore was less restrictive. As seen in Table 1, all potent
compounds featured diverse substituents at the R1 position. Both
isothiazole as well as phenyl moieties were well tolerated. Notably,
pyrrole or imidazole analogues 22, 23, 24 displayed marginal
activity in our assays, suggesting unfavorable interaction(s) with
the colchicine binding site. Ortho-anilines with or without N-
substituent were consistently active as well. Methoxylation of the
core ring reduced overall potency (compare compounds 13, 16 and
17; 14 and 15) presumably due to the steric hindrance between R2
and amino acid residues of tubulin. Replacement of pyridine group
with isothiazole in R1 reduced potency of the resultant compound
20, probably because of a mismatch between this group and the
hydrophobic pocket in tubulin. Compound 18 was a weak tubulin
inhibitor despite of its favorable substitution with the 3,4,5-tri-
methoxyphenyl group in R1 (ex., A-ring in combretastatins). This
outcome is likely due to the lack of spacer between R1 and oxa-
diazole core causing steric clash between the molecule and the
inner part of the pocket [35].
ROCS data further suggested that the inactive compound 1 was
in the same range of similarity with other molecules from the set.
Docking experiments with 1 placed its oxadiazole group in the
same orientation as A-ring of CA-4 and CA-2. This is an unfavorable
binding mode as the oxadiazole moiety is deeply buried into the
binding site (Fig. 6). We have observed similar behavior for the
compounds 21 and 25 both featuring unsubstituted amines.
Unfortunately, we were unable to identify a low energy minimum
conformation (<ꢃ75 kcal molꢃ1) for the active compounds 13 and
14 to rationalize their binding mode to tubulin.
Table 3
3D similarity scores for selected ligands to CA–2 and CA–4.a
Compound
CA-2
CA-4
13
14
17
1
6
12
1.39
1.35
1.28
1.28
1.01
0.99
1.32
1.32
1.15
1.21
0.98
0.98
a
ROCS results were based on the shape and color (electrostatics) similarity score.
See Supporting Information, Fig. S2.

A.S. Kiselyov et al. / European Journal of Medicinal Chemistry 45 (2010) 1683–1697
1693
Fig. 5. Superposition of selected compounds to CA-2 by ROCS.
Fig. 6. Docking results for binding of selected 1,3,4-oxadiazoles (1, 4, 6, 7, 12 and 21) to
a- and b-subunits of tubulin. View from the pocket entrance. Tubulin structure is represented
as cyan ( -subunit) and light green ( -subunit) ribbons. Some amino acid side chains are shown. Dashed lines represent the key hydrogen bonds.
a
b

1694
A.S. Kiselyov et al. / European Journal of Medicinal Chemistry 45 (2010) 1683–1697
3. Conclusions
4.2.2. 2-(4-Aminoisothiazolyl-3)-5-((3,4-
methylenedioxy)phenyl)amino-1,3,4-oxadiazole 1
A series of novel 1,3,4-oxadiazole derivatives based on struc-
tural and electronic overlap with combretastatin A-2 have been
designed, synthesized and tested in vivo using the sea urchin
embryo development assay. We monitored the effects of these
agents on two specific developmental stages of the embryo,
namely i) fertilized egg to assess anti-mitotic activity; ii) free-
swimming blastulae to detect behavioral changes in the embryo
It is prepared from hydrazide of 4-aminoisothiazole carboxylic-
3-acid and 3,4-methylenedioxyphenylisothio cyanate as described
in procedure 1: 80% yield, mp 268–270 ꢁC. 1H NMR (DMSO-d6,
d):
5.81 (s, 2H, NH2), 6. 00 (s, 2H, OCH2O), 6.93 (d,J ¼ 8.7 Hz, 1H), 7.04
(d, J ¼ 8.7 Hz,1H), 7.29 (s,1H), 7.80 (s,1H),10.78 (s,1H, NH). 13C NMR
(DMSO-d6, d): 97.5, 101.4, 101.7, 110.3, 125.2, 133.7, 138.8, 142.2,
148.0, 148.2, 158.1, 166.7. Mass-spectrum: EI-MS (70 eV, 20 ꢁC) (m/z,
I, %): 303 [M]þ (100), 162 (54), 125 (52). Anal. Calcd for C12H9N5O3S:
C, 47.52; H, 2.99; N, 23.09; S 10.57. Found: C, 47.22; H, 3.03; N,
22.87; S 10.70.
swimming pattern. These were quantified by
a threshold
concentration resulting in respective abnormalities allowing for
both rapid prioritization of the tubulin inhibitors and identifi-
cation of non-specific cytotoxic substances. Following the sea
urchin embryo studies, the prioritized actives were further
profiled in a panel of cell-based and in vitro assays. Further
evaluation of microtubule and cell cycle effects yielded lead
molecules 6 and 12 with activity profiles similar to that of
nocodazole. Specifically, these compounds featured considerable
anti-mitotic effect with EC values of 0.5–5 nM in the sea urchin
embryo based on the tubulin destabilizing effect. Molecules 6
and 12 were further confirmed to be equipotent to nocodazole in
their ability to depolymerize MT both in vitro and in the intact
cells. Similar to nocodazole, compounds 6 and 12 afforded
pronounced cytotoxicity (IC50 ¼ 53 and 15 nM, respectively) and
cell cycle arrest in G2/M phase (EC50 ¼ 63 and 45 nM, respec-
tively) in cultured A549 human lung carcinoma cells. A combi-
nation of the sea urchin embryo assay with the tests of tubulin
dynamics in cultured human tumor cells allowed for the rapid
identification and prioritization of promising and specific MT
inhibitors from a novel chemical class. SAR data for 1,3,4-oxadi-
zaoles were further aligned with results of computational
studies. Several identified actives were superimposed to CA-4
and CA-2 molecules and docked into the colchicine binding site
of tubulin. Docking results were used to propose the putative
mode of binding for these novel molecules.
4.2.3. 2-(4-Aminoisothiazolyl-3)-5-((3,4-
ethylenedioxy)phenyl)amino-1,3,4-oxadiazole 2
It is prepared from the hydrazide of 4-aminoisothiazolecarboxylic-
3-acid and 3,4-ethylenedioxyphenylisothiocyanate according to
procedure A: 69% yield, mp 265–267 ꢁC. 1H NMR (DMSO-d6,
d): 4.23
(m, 4H, OCH2CH2O), 5.83 (s, 2H, NH2), 6.86 (d, J ¼ 8.7 Hz,1H), 7.00 (dd,
J1 ¼ 8.7 Hz, J2 ¼ 2.5 Hz, 1H), 7.24 (d, J ¼ 2.5 Hz, 1H), 7.79 (s, 1H), 10.80
(br.s, exch. D2O, 1H, NH). EI-MS (m/z, I, %): 317 [M]þ. Anal. Calcd for
C13H11N5O3S: C, 49.20; H, 3.49; N, 22.07; S 10.11. Found C, 48.94; H,
3.51; N, 21.95; S 9.75.
4.2.4. 2-(4-Aminoisothiazolyl-3)-5-(4-chloro)phenylamino-1,3,4-
oxadiazole 3
It is prepared from hydrazide of 4-aminoisothiazolylcarboxylic-
3 acid and 4-chlorophenylisothiocianate according to the proce-
dure 1, yield 56%, mp 211–213 ꢁC. 1H NMR (DMSO-d6,
d): 5.85 (s, 2H,
NH2), 7.44 (d, J ¼ 8.6 Hz, 2H), 7.63 (d, J ¼ 8.6 Hz, 2H), 7.8 (s, 1H), 11.1
(br.s, exch. D2O, 1H, NH). EI-MS (m/z, I, %): 295 [M]þ (27), 293 [M]þ
(63), 127 (49), 125 (100). Anal. Calcd for for C11H8ClN5OS: C, 44.98;
H, 2.75; N, 23.84; Cl 12.07; S 10.92. Found: C, 45.30; H, 2.96; N,
23.51; Cl 12.02; S 11.06.
4.2.5. N-(4-chlorophenyl)-5-{4-[(pyridin-3-
ylmethyl)amino]isothiazol-3-yl}-1,3,4-oxadiazol-2-amine 4
Solution of of 4-aminoisothiazole-3-carboxamide (0.715 g;
5 mM), 3-pyridylaldehyde (0.696 g; 6.5 mM) and pTsOH (20 mg) in
a mixture of MeOH (1.5 mL) and MeCN (6 mL) was stirred for 48 h at
RT. The mixture was concentrated in vacuo and treated with
tBuOMe (5 mL). The precipitate was filtered, washed with a mixture
of tBuOMe-hexanes (1:1) and dried to furnish a pale yellow crystals
of the Schiff base (0.901 g, 78%). EI-MS (m/z, I, %): 232 [M]þ The
material was used in the next step without further purification.
NaBH4 (0.28 g; 7.5 mM) was added portion-wise (15 min) to the
solution of the Schiff base (0.9 g, 3.9 mM) in MeOH (6 mL) under
vigorous stirring at RT. After the addition was completed, the
reaction mixture was stirred for additional 30 min and concen-
trated in vacuo. The residue was triturated with water, pH was
adjusted to 9 with AcOH, the mixture was extracted with tBuOMe
(5 ꢄ 20 mL), the organic phase was collected and concentrated in
vacuo and evaporated. The solid material was dried with heating in
vacuo (P2O5) to afford the target carboxamide as a pale yellow
crystals (0.90 g, 98.7% yield), m.p. 135–137ꢁ. EI-MS (m/z, I, %):
234 [M]þ.
4. Experimental section
4.1. Chemistry
NMR data were collected on a Bruker DR-500 instrument
(working frequencies of 500.13 MHz (1H)). Mass spectra were
collected on a Finningan MAT/INCOS 50 instrument (70 eV) using
direct probe injection.
4.2. 4-Aminoisothiazolecarboxylic-3-acid hydrazide
A mixture of amide of 4-aminoisothiazole-3-carboxylic acid [38]
(5.39 g; 38 mM) and 28 mL of 85% hydrazine hydrate was refluxed
for 3 h and cooled to RT. The resulting precipitate was filtered off,
washed with ice water (3 ꢄ 5 mL) and dried in vacuo to furnish acyl
hydrazide (4.17 g, 70% yield), mp 170–172 ꢁC. 1H NMR (DMSO-d6,
d):
4.45 (s, 2H, NH2), 5.72 (s, 2H, NH2), 7.68 (s, 1H), 9.7 (br.s, 1H, NH).
4.2.1. Preparation of 5-(arylamino)-1,3,4-oxadiazoles from acid
hydrazides and arylisothiocyanates (general procedure 1)
A
solution (or suspension) of acyl hydrazide and
4.2.6. 4-[(pyridin-3-ylmethyl)amino]isothiazole-3-carbohydrazide
A mixture of 4-((3-pyridyl)methyl)4-aminoisothiazole-3-car-
boxamide (0.89 g; 3.8 mM) and 85% hydrazine hydrate (6 mL) was
refluxed for 2 h. Excess of hydrazine was removed in vacuo followed
by co-evaporation with iPrOH. The resulting oil was further dried in
vacuo with heating under P2O5 to furnish 4-((3-pyr-
idyl)methyl)aminoisothiazol-3-carbohydrazide (0.92g, 97%yield) as
a viscous oil (Rf 0.22, EtOAc–MeOH ¼ 19:1). This material was used
arylisothiocyanate(1 mM each) in tBuOH (3 mL) was refluxed for
10 min followed by a vigorous stirring for 1 h at 50–60 ꢁC. The
reaction mixture was concentrated in vacuo and the residue was
treated with MeCN (5 mL). DCC (1.5 mM) was added to the resulting
suspension, the mixture was stirred at reflux for 3 h and cooled to
RT. The precipitate was filtered off, washed with hot benzene or hot
MeCN (2 ꢄ 2 mL) and dried.

A.S. Kiselyov et al. / European Journal of Medicinal Chemistry 45 (2010) 1683–1697
1695
in the next step without further purification. The target compound
4 was prepared from the 4-[(pyridin-3-ylmethyl)amino]isothia-
OCH2O), 6.75 (t, J ¼ 6.1 Hz, 1H, NH), 6.92 (d, J ¼ 8.7 Hz, 1H), 7.10
(d, J ¼ 8.7 Hz, 1H), 7.30 (s, 1H), 7.37 (dd, J ¼ 7.8 Hz, J ¼ 5.0 Hz, 1H),
7.75 (s, 1H), 7.80 (d, J ¼ 7.8 Hz, 1H), 8.47 (d, J ¼ 5.0 Hz, 1H), 8.63
zole-3-carbohydrazide (0.25 g;
isothiocyanate (1.2 mM) as described in the procedure 1, yield
0.163 g (50%), mp 244–246ꢁ. 1H NMR (DMSO-d6,
1
mM) and 4-Cl-phenyl-
(s, 1H), 10.80 (br.s, 1H, NH). 13C NMR (DMSO-d6,
d): 49.0, 97.7,
d): 4.5 (d, J ¼ 6 Hz,
100.6, 101.8, 111.3, 123.3, 123.4, 126.4, 133.9, 135.2, 138.0, 141.9,
147.8, 148.2, 148.6, 148.9, 158.4, 166.6. EI-MS (m/z, I, %): 394 [M]þ
(42), 162 (79), 137 (71), 92 (100). Anal. Calcd for C18H14N6O3S: C,
54.81; H, 3.58; N, 21.31; S 8.13. Found: C, 55.00; H, 3.69; N, 21.05;
S 7.96.
2H, CH2), 6.75 (t, J ¼ 6 Hz,1H, NH), 7.35 (dd, J1 ¼7.5 Hz, J2 ¼ 5 Hz,1H,
3-Py), 7.44 (d, J ¼ 8.3 Hz, 2H, Ar), 7.64 (d, J ¼ 8.3 Hz, 2H, Ar), 7.78 (s,
1H, isothiazole), 7.8 (d, J ¼ 5 Hz, 1H, 4-Py), 8.48 (d, J ¼ 7.5 Hz, 1H, 2-
Py), 8,62 (s, 1H, 2-Py), 11.13 (br.s., exch. D2O, 1H, NH). 13C NMR
(DMSO-d6, d): 48.3, 118.0, 123.9, 124.2, 126.7, 129.0, 133.7, 134.8,
135.6, 138.3, 141.5, 148.6, 148.9, 158.0, 166.8. EI-MS (m/z, I, %): 384
[M]þ, 386 [M]þ. Anal. calcd for C17H13ClN6OS: C, 53.06; H, 3.40; N,
21.84; S 8.33. Found: C, 53.17; H, 3.55; N, 21.95; S 8.56.
4.2.10. 2-((4-((Pyridyl-3)methyl)amino)isothiazolyl-3)-5-(3,4-
ethylenedioxy)phenylamino-1,3,4-oxadiazole 8
It is prepared from 2-((4-amino)isothiazolyl-3)-5-(3,4-ethyl-
enedioxy)phenylamino-1,3,4-oxadiazole 2 and 3-pyridinecarbal-
dehyde according to procedure 2, 84% yield, mp 226–228 ꢁC.1H NMR
4.2.7. 2-((4-((Pyridyl-4)methyl)amino)isothiazolyl-3)-5-(3,4-
ethylenedioxy)phenylamino-1,3,4-oxadiazole 6
(DMSO-d6,
d
): 4.24 (m, 4H, OCH2CH2O), 4.53 (d, J ¼ 6.0 Hz, 2H), 6.75
A mixture of 4-aminoisothiazole-3-carboxamide [38] (3.44 g;
24 mM), 4-pyridylaldehyde (3.08 g; 28.8 mM) in MeOH (10 mL),
MeCN (25 mL) and pTsOH (100 mg) was stirred 5 h at RT. The
precipitate of Schiff base was filtered off, washed with MeCN (5 mL)
dried, and reduced with NaBH4 without further purification as
follows. To the suspension of Schiff base (4.9 g; 21.1 mM) in iPrOH
(50 mL) was added NaBH4 (2.28 g; 60 mM) in three portions with
stirring and reflux for 7 h. The reaction mixture was concentrated in
vacuo to 1/4 of the original volume. The residue was triturated with
water, AcOH was added to pH 9, the precipitate was filtered off,
washed with ice water and dried. The resulting amine (2.37 g) and
NH2NH2$H2O (17 mL) were refluxed for 3.5 h, the mixture was cooled
toRT, the precipitatewasfilteredoff, washedwithicewater(2ꢄ 3mL)
and dried to afford 4-((4-pyridyl)methyl)aminoisothiazol-3-carbo-
hydrazide in a 59% total yield (3 steps starting from 4-amino-
isothiazole-3-carboxamide), mp 146–148 ꢁC. The material was used
in the next step without further purification. The hydrazide (1.5 g;
6 mM) and 3,4-ethylenedioxyphenylisothiocyanate were reacted as
described in the procedure 1 to yield 6, 75%, mp 255–258 ꢁC.1H NMR
(t, J ¼ 6.0 Hz, 1H, NH), 6.85 (d, J ¼ 8.7 Hz,1H), 7.00 (d, J ¼ 8.7 Hz, 1H),
7.22 (s, 1H), 7.36 (dd, J ¼ 7.8 Hz, J ¼ 5.0 Hz, 1H), 7.75 (s, 1H), 7.80 (d,
J ¼ 7.8 Hz,1H), 8.45 (d, J ¼ 5.0 Hz,1H), 8.64 (s,1H),10.72 (br.s,1H, NH).
EI-MS (m/z, I, %): 408 [M]þ (31), 176 (37), 151 (89), 92 (100). Anal.
Calcd for C19H16N6O3S: C, 55.87; H, 3.95; N, 20.58; S 7.85. Found: C,
56.20; H, 4.14; N, 20.32; S 7.99.
4.2.11. 2-((4-((Pyridyl-4)methyl)amino)isothiazolyl-3)-5-(3,4-
methylenedioxy)phenylamino-1,3,4-oxadiazole 5
It is prepared from 2-((4-amino)isothiazolyl-3)-5-(3,4-methyl-
enedioxy)phenylamino-1,3,4-oxadiazole 1 and 4-pyridinecarbal-
dehyde according to procedure 2, 84% yield, mp 252–255 ꢁC. 1H
NMR (DMSO-d6,
d
): 4.51 (d, J ¼ 6.0 Hz, 2H), 6.00 (s, 2H, OCH2O), 6.82
(t, J ¼ 6.1 Hz, 1H, NH), 6.93 (d, J ¼ 8.7 Hz, 1H), 7.05 (d, J ¼ 8.7 Hz, 1H),
7.31 (s, 1H), 7.38 (d, J ¼ 5.5 Hz, 2H), 7.64 (s, 1H), 8.50 (d, J ¼ 5.5 Hz,
2H), 10.81 (s, 1H, NH). EI-MS (m/z, I, %): 394 [M]þ (100), 216
(38), 162 (50), 161 (47), 92 (46). Anal. Calcd for C18H14N6O3S: C,
54.81; H, 3.58; N, 21.31; S 8.13. Found: C, 54.71; H, 3.70; N, 21.15; S
8.05.
(DMSO-d6,
d
): 4.24 (m, 4H, OCH2CH2O), 4.53(d, J ¼ 6.1 Hz, 2H), 6.83 (t,
J ¼ 6.1 Hz, 1H, NH), 6.84 (d, J ¼ 8.7 Hz, 1H), 7.00 (dd, J ¼ 8.7 Hz,
5. Both synthetic protocols and analytical data for
compounds 9–25 are summarized in the supporting
information
J ¼ 2.5 Hz, 1H), 7.25 (d, J ¼ 2.5 Hz, 1H), 7.37 (d, J ¼ 5.5 Hz, 2H), 7.62 (s,
1H), 8.52 (d, J ¼ 5.5 Hz, 2H), 10.72 (s, 1H, NH). 13C NMR (DMSO-d6,
d):
48.7, 69.4, 107.1, 112.0, 115.3, 126.5, 129.4, 133.6, 138.4, 138.8, 141.7,
144.6, 148.2, 149.1, 158.2, 166.5. EI-MS (m/z, I, %): 408 [M]þ (63), 233
(66), 216 (49), 176 (84), 151 (77), 120 (85), 92 (100). Anal. Calcd for
C19H16N6O3S: C, 55.87; H, 3.95; N, 20.58; S 7.85. Found: C, 56.09; H,
3.91; N, 20.45; S 8.12.
5.1. Biology
5.1.1. Sea urchin embryo assay
Adult sea urchins Paracentrotus lividus were collected from
Mediterranean Sea at Cyprus coast and kept in an aerated seawater
tank. Gametes were obtained by intracoelomic injection of 0.5 M
KCl. Eggs were washed with filtered sea water and fertilized by
adding drops of a diluted sperm. Embryos were cultured at room
temperature under gentle agitation with a motor-driven plastic
paddle (60 rpm) in filtered sea water. The embryos were observed
with a light microscope Biolam (LOMO, S.-Peterburg, Russia). For
treatment with the test compounds, 5 mL aliquots of embryo
suspension were transferred to 6-well plates and incubated as
a monolayer at a concentration up to 3000 embryos/mL. Stock
solutions of compounds were prepared either in 95% EtOH at 5 mM
or in DMSO at 5–10 mM concentrations followed by a 10-fold
dilution with 95% EtOH. The anti-proliferative activity was assessed
by exposing fertilized eggs (10–25 min after fertilization, 45–
60 min before the first mitotic cycle completion) to 2-fold
decreasing concentrations of the compound. Cleavage alteration
and arrest were clearly detected at 2.5–6 h after fertilization (Fig.1).
The effects were quantitatively estimated as a threshold concen-
tration resulting in cleavage alteration and embryo death before
hatching or full mitotic arrest. For tubulin destabilizing activity, the
compounds were tested on free-swimming blastulae just after
4.2.8. Reductive amination of heterocyclic amines with aldehydes
(general procedure 2)
A suspension of amine (1 mM), pyridyl aldehyde and pTsOH
(0.1 mM) in dry MeCN (5 mL) was placed in a flask equipped with
addition funnels and a pressure equalizing sleeve filled with 4 Å
molecular sieves. The resulting mixture was refluxed with stirring
for 3 h and cooled to RT. The precipitate of a Schiff base was filtered,
dried and reduced with NaBH4 without further purification. NaBH4
(3 mM) was added to the suspension of Schiff base in iPrOH (5 mL)
and the resulting mixture was refluxed for 3h with stirring.
Subsequently, iPrOH was removed in vacuo, the residue was treated
with water. The resulting crystals were filtered off, washed with
aqueous 5% AcOH, water and dried.
4.2.9. 2-((4-((Pyridyl-3)methyl)amino)isothiazolyl-3)-5-(3,4-
methylenedioxy)phenylamino-1,3,4-oxadiazole 7
It is prepared from 2-((4-amino)isothiazolyl-3)-5-(3,4-meth-
ylenedioxyphenyl)amino-1,3,4-oxadiazole
idinecarbaldehyde according to procedure 2, 88% yield, mp 230–
232 ꢁC. 1H NMR (DMSO-d6,
): 4.50 (d, J ¼ 6.1 Hz, 2H), 6.00 (s, 2H,
(3)
and
3-pyr-
d

1696
A.S. Kiselyov et al. / European Journal of Medicinal Chemistry 45 (2010) 1683–1697
hatching (9–12 h after fertilization), originated from the same
embryo culture. Embryo spinning was observed after 0.5–20 h of
treatment, depending on the nature and concentration of the
compound. Both spinning and lack of forward movement were
interpreted to be the result of the tubulin destabilizing activity of
a molecule according to previous studies [21,22].
acquired with a Nikon Diaphot 300 inverted microscope (Nikon
GmbH, Du¨sseldorf, Germany) equipped with a cooled charge-
couple device camera system (SenSys; Photometrics, Munich,
Germany).
5.1.6. Tubulin polymerization assay
Purified tubulin was isolated from bovine brain in BRB80 buffer
(80mM PIPES, pH 6.8, 1 mM MgCl2, 1 mM EGTA) [40]. Polymeri-
zation was initialized by adding of 10% DMSO and GTP (0.5 mM) to
1.2 mg/mL tubulin solution. Tubulin samples were incubated for 1 h
at 37 ꢁC with test compounds and with nocodazole as a positive
5.1.2. Cell culture
A549 human lung epithelial carcinoma cells (CCL-185Ô) were
cultured with Dulbecco’s Modified Eagle medium (DMEM) con-
taining 10% fetal bovine serum and 1% antibiotic penicillin/strep-
tomycin at 37 ꢁC under a 5% CO2 humidified atmosphere.
control at concentration of 10 mM, and 0.6% DMSO as a negative
control. The MT assembly was analysed using VEC-DIC microscopy
using a Nikon Diaphot 300 inverted microscope (Nikon GmbH,
Du¨sseldorf, Germany) equipped with aHamamatsu C2400-07
Newvicon and ARGUS 20 image processor (Hamamatsu Photonics
Deutschland GmbH, Herrsching, Germany) [41].
5.1.3. MTT cytotoxicity assay
Cells were seeded in 96-well plates at a density of 3 ꢄ 103 cells
per well. Stock solutions of test compounds were prepared in
dimethylsulfoxide (DMSO). Cells were treated for 24 h with 6 and
12 at 0.001–0.3
mM, 13 and 14 at 0.5–25
mM, 1 at 1–20mM, or with
nocodazole at 0.001–0.3
mM as a positive control (8 wells per each
5.2. Computational studies
concentration value). DMSO (0.4%) served as a control. The number
of surviving cells was determined by the colorimetric MTT assay
[26]. MTT (3-(4,5-dimethylthiazolyl-2)-2,5-diphenyl-2H-tetrazo-
liumbromid, Roth GmbH, Karlsruhe, Germany) was prepared at
5 mg/mL in phosphate buffer saline (PBS) and filtered through
Known tubulin inhibitors were used as a reference set (Fig. S1).
Smiles codes and 3D structures were gathered from PubChem
database and the respective co-crystal structures with tubulin from
PDB: taxolÔ (1JFF), DAMA-colchicine (1SA0), podophyllotoxin
(1SA1). Other compounds from this report (including CA-2 and CA-
4) were modeled in Sybyl 8.1 (Tripos, Saint Louis, MO). Omega, ROCS,
FRED and VIDA software were acquired from OpenEye Scientific
Software Inc. (Santa Fe, NM). Omega software (v. 2.2.1) was chosen to
sample the 3D conformations for all molecules using RMS ¼ 0.3;
energy window (ewindow) ¼ 15 Kcal molꢃ1; maximum number of
conformations to be saved (maxconfs) ¼ 1000; maximum number of
conformations generated (maxconfgen) ¼ 100,000; (fixrms) ¼ 0.1;
maximum amount of time (maxtime) ¼ 360 s; other parameters
were set as default. The coordinates of the original X-ray structures
of podophyllotoxin and DAMA-colchicine were modified in order to
include these conformations.
3D similarity testing was performed by Rapid Overlay of
Chemical Structures (ROCS v. 2.3.1) with the default parameters and
10 chosen conformations (maxconfs). Fast Rigid Exhaustive Dock-
ing (FRED v. 2.2.3) [37] was the chosen docking program and all
calculations were based on the default parameters, number of
poses (num poses) ¼ 200 and number of alternative poses
(num_alt_poses) ¼ 19. A higher than usual number of poses was
selected in order to try to cover different poses and interactions
between the compound and the protein, which may be captured in
one of the 20 poses saved at the end of the docking procedure. The
X-ray crystallographic structures of tubulin with DAMA-colchicine
(1SA0) were chosen to be the reference for docking. Podophyllo-
toxin (1SA1 X-rays structure) was used as a reference to check the
pose of this compound. Taxol does not interact with tubulin in the
a 0.22
0.5 mg/mL) was added to each well 2 h before the end of compound
exposure. Then the supernatant was removed, and 100 L DMSO
mm filter. 20 mL sterile MTT solution (final concentration
m
containing 10% SDS and 0.6% acetic acid was added to each well.
Resulting formazan crystals were solubilised by thorough mixing
on a plate shaker. Optical density was measured at 590 nm with
690 nm reference filter using EL808 Ultra Microplate Reader (Bio-
Tek Instruments, Winooski, USA). Experiments for all compounds
were repeated 3 times and IC50 values were determined by
sigmoidal curve fitting using Excel-based software.
5.1.4. Cell cycle analysis
A549 cells were seeded at a density of 5 ꢄ 105 cells in 50 mm
culture dishes, treated with test compounds at concentrations of
0.001–25 mM for 24 h and harvested by trypsinization. Nocoda-
zole and 0.4% DMSO served as positive and negative controls,
respectively. After centrifugation (1000 rpm, 25 ꢁC, 5 min) and
resuspension in HBS (14 mM HEPES, pH 7.4, 0.9% NaCl2) cells
were fixed by dropwise adding to 70% ethanol (ꢃ20 ꢁC). Fixed
cells were centrifuged (1300 rpm, 4 ꢁC, 10 min) and treated with
1 mg/mL RNase in HBS (Sigma, St. Louis, USA) for 30 min at 37 ꢁC,
followed by incubation with 50 mg/mL propidium iodide in PBS
(Roth, Karlsruhe, Germany) for 30 min at 37 ꢁC. The DNA content
of a total of 10000 cells in each sample was analysed using an
Epics Altra flow cytometer (Beckman Coulter, Germany). Experi-
ments for all compounds were repeated 2 times, and EC50 values
were determined by sigmoidal curve fitting using Excel-based
software.
pocket formed at the interface of a/b subunits like colchicine and
podophyllotoxin [42]. The X-ray structures of taxol (1JFF) and
podophyllotoxin (1SA1) were superimposed to 1SA0 to see the
5.1.5. Immunofluorescence staining of cellular microtubules
partial occlusion of the colchicine binding site by the
tubulin (Fig. S3).
a-helix of
For MT staining A549 cells were cultured in 12-well plates on
small glass coverslips (11 mm diameter) at a density of 2 ꢄ 104
cells per coverslip. Cells were incubated with selected compounds
or nocodazole as a positive control at concentrations of 0.025, 5,
Two cages were built to dock the reference compounds into the
1SA0 structure (Fig. S4 and Table S1): 1). The entire interface was
considered when building the lager cage, but the docking results
did not support the idea that some compounds displace GTP from
its binding site (not shown), so that it was not further considered in
the study; 2) a smaller cage centered on the colchicine binding site
provided the best results. The addition of two-point pharmaco-
phores further improved the overall results for the dataset in study
analysed in VIDA (v. 3.0.0). Two hydrophobic pharmacophores used
to guide docking experiments were chosen from the set presented
and 25
m
M at 37 ꢁC and 5% CO2 for 8 h 0.6% DMSO served as
a negative control. The whole process of cell fixation and staining
was described previously [39]. Fixed cells were labelled with
mouse monoclonal antibody against alpha tubulin at a dilution of
1:400 (Sigma, St. Louis. USA), followed by incubation of Alexa
Flour488 labelled goat anti-mouse IgG at a dilution of 1:200
(Molecular Probes, Eugene, USA). Images of all samples were

A.S. Kiselyov et al. / European Journal of Medicinal Chemistry 45 (2010) 1683–1697
1697
[21] X. Ouyang, E.L. Piatnitski, V. Pattaropong, X. Chen, H.-Y. He, A.S. Kiselyov,
A. Velankar, J. Kawakami, M. Labelle, L. Smith II, J. Lohman, S.P. Lee,
A. Malikzay, J. Fleming, J. Gerlak, Y. Wang, R.L. Rosler, K. Zhou, S. Mitelman,
M. Camara, D. Surguladze, J.F. Doody, M.C. Tuma, Oxadiazole derivatives as
a novel class of antimitotic agents: synthesis, inhibition of tubulin polymeri-
zation, and activity in tumor cell lines. Bioorg. Med. Chem. Lett. 16 (2006)
1191–1196.
[22] M. Semenova, A.S. Kiselyov, V.V. Semenov, Sea urchin embryo as a model
organism for the rapid functional screening of tubulin modulators. Bio-
Techniques 40 (2006) 765–774 video illustrations are available at. http://
www.chemblock.com.
[23] M.N. Semenova, A.S. Kiselyov, I.Y. Titov, M.M. Raihstat, M. Molodtsov,
E. Grishchuk, I. Spiridonov, V.V. Semenov, In Vivo evaluation of indolyl
glyoxamides in the sea urchin embryo model: correlation with in vitro tubulin
dynamics effects. Chem. Biol. Drug Design 70 (2007) 485–490.
[24] G.R. Pettit, S.B. Singh, M.R. Boyd, E. Hamel, R.K. Pettit, J.M. Schmidt, F. Hogan,
Antineoplastic agents. 291. Isolation and synthesis of combretastatins A-4,
A-5, and A-6. J. Med. Chem. 38 (1995) 1666–1672.
in [30] Both their positioning as well as the results for a training set
are displayed in Fig. S5.
Acknowledgments
This work was supported by the German organization DAAD
(German Academic Exchange Service).
Supplementary data
Supplementary data associated with this article can be found in
the online version, at doi:10.1016/j.ejmech.2009.12.072.
References
[25] X. Ouyang, A. Kiselyov, X. Chen, H.-Y. He, J. Kawakami, V. Pattaropong,
E. Piatnitski, M.C. Tuma, J. Kincaid, Heterocyclic compounds and their use as
anticancer agents, Patent WO 2005/004818 A2, (2005).
[1] D. Bray, Cell Movements: From Molecules to Motility, second ed. Garland
Publishing, New York, 2001.
[26] T. Mosmann, Rapid colorimetric assay for cellular growth and survival:
application to proliferation and cytotoxicity assays. J. Immunol. Methods 65
(1983) 55–63.
[27] P.C.D. Hawkins, A.G. Skillman, A. Nicholls, Comparison of shape-matching and
docking as virtual screening tools. J. Med. Chem. 50 (2007) 74–82.
[28] C.M. Lin, S.B. Singh, P.S. Chu, R.O. Dempcy, J.M. Schmidt, G.R. Pettit, E. Hamel,
Interactions of tubulin with potent natural and synthetic analogs of the
antimitotic agent combretastatin: a structure-activity study. Mol. Pharmacol.
34 (1988) 200–208.
[29] G. Bacher, B. Nickel, P. Emig, U. Vanhoefer, S. Seeber, A. Shandra, T. Klenner,
T. Beckers, D-24851, a novel synthetic microtubule inhibitor, exerts curative
antitumoral activity in vivo, shows efficacy toward multidrug-resistant tumor
cells, and lacks neurotoxicity. Cancer Res. 61 (2001) 392–399.
[30] A. Couture, E. Deniau, P. Grandclaudon, S. Lebrun, S. Le´once, P. Renard,
B. Pfeiffer, First synthesis and pharmacological evaluation of benzoindolizi-
[2] M.A. Jordan, L. Wilson, Microtubules as a target for anticancer drugs. Nat. Rev.
Cancer 4 (2004) 253–265.
[3] A.S. Kiselyov, K. Balakin, S.E. Tkachenko, A.V. Ivachtchenko, Recent develop-
ments in discovery and development of antimitotic agents. Anti-Cancer
Agents Med. Chem. 7 (2007) 189–208.
[4] L. Wordeman, T.J. Mitchison, Dynamics of microtubule assembly in vivo. in:
J.S. Hyams, C.W. Lloyd (Eds.), Microtubules. Wiley-Liss Inc., New York, 1994, pp.
287–301.
[5] E. Hamel, Antimitotic natural products and their interactions with tubulin.
Med. Res. Rev. 16 (1996) 207–231.
[6] A. Jordan, J.A. Hadfield, N.J. Laurence, A.T. McGown, Tubulin as a target for
anticancer drugs: agents which interact with the mitotic spindle. Med. Res.
Rev. 18 (1998) 259–296.
[7] D.G.I. Kingston, Tubulin-interactive natural products as anticancer agents.
J. Nat. Prod. (2009) Jan 6 [Epub ahead of print].
[8] L. Wilson, D. Panda, M.A. Jordan, Modulation of microtubule dynamics by
drugs: a paradigm for the actions of cellular regulators. Cell Struct. Funct. 24
(1999) 329–335.
[9] E. Nogales, M. Whittaker, R.A. Milligan, K.H. Downing, High-resolution model
of the microtubule. Cell 96 (1999) 79–88.
dine and benzoquinolizidine analogues of
Chem. 8 (2000) 2113–2125.
a- and b-peltatin. Bioorg. Med.
[31] http://eyesopen.com/products/applications/omega.html.
[32] http://eyesopen.com/products/applications/rocs.html.
[33] S. Fortin, E. Moreau, J. Lacroix, J.-C. Teulade, A. Patenaudea, R. Gaudreaulta,
N-Phenyl- N0-(2-chloroethyl)urea analogues of combretastatin A-4: is the
N-phenyl-N0-(2-chloroethyl)urea pharmacophore mimicking the trimethoxy
phenyl moiety. Bioorg. Med. Chem. Lett. 17 (2007) 2000–2004.
[34] T.L. Nguyen, C. McGrath, A.R. Hermone, J.C. Burnett, D.W. Zaharevitz, B.W. Day,
P. Wipf, E. Hamel, R. Gussio, A common pharmacophore for a diverse set of
colchicine site inhibitors using a structure-based approach. J. Med. Chem. 48
(2005) 6107–6116.
[10] J.F. Diaz, R. Strobe, Y. Engelborghs, A.A. Souto, J.M. Andreu, Molecular recog-
nition of Taxol by microtubules. Kinetics and thermodynamics of binding of
fluorescent Taxol derivatives to an exposed site. J. Biol. Chem. 275 (2000)
26265–26276.
[11] R.B.G. Ravelli, B. Gigant, P.A. Curmi, I. Jourdain, S. Lachkar, A. Sobel,
M. Knossow, Insight into tubulin regulation from a complex with colchicine
and a stathmin-like domain. Nature 428 (2004) 198–202.
[35] A. Chaudhary, S.N. Pandeya, P. Kumar, P.P. Sharma, S. Gupta, N. Soni,
K.K. Verma, G. Bhardwaj, Combretastatin A-4 analogs as anticancer agents.
Mini Rev. Med. Chem. 7 (2007) 1186–1205.
[36] X. Ouyang, X. Chen, E.L. Piatnitskii, A.S. Kiselyov, H.-Y. He, Y. Mao,
V. Pattaropong, Y. Yu, K.H. Kim, J. Kincaid, L. Smith II, W.C. Wong, S.P. Lee,
D.L. Milligan, A. Malikzay, J. Fleming, J. Gerlak, D. Dhanvanthri, J.F. Doody,
H.-H. Chiang, S.N. Patel, Y. Wang, R.L. Rosler, P. Kussie, M. Labelle, M.C. Tuma,
Synthesis and structure-activity relationships of 1,2,4-triazoles as a novel class
of tubulin polymerization inhibitors. Bioorg. Med. Chem. Lett. 15 (2005)
5154–5159.
[37] M.R. McGann, H.R. Almond, A. Nicholls, J.A. Grant, F.K. Brown, Gaussian
docking functions. Biopolymers 68 (2003) 76–90.
[38] K. Gewald, P. Belimann, Syntese und reactionen von 4-aminoisothiazolen.
Liebigs Ann. Chem. (1979) 1534–1546.
[39] A. Al-Haddad, M.A. Shonn, B. Redlich, A. Blocker, J.K. Burkhardt, H.
Yu, J.A. Hammer 3rd, D.G. Weiss, W. Steffen, G. Griffiths, S.A. Kuznetsov,
Myosin Va bound to phagosomes binds to F-actin and delays microtubule-
dependent motility. Mol. Biol. Cell. 12 (2001) 2742–2755.
[40] S.A. Kuznetsov, V.I. Rodionov, A.D. Bershadsky, V.I. Gelfand, V.A. Rosenblat,
High molecular weight protein MAP2 promoting microtubule assembly in
vitro is associated with microtubules in cells. Cell Biol. Int. Rep. 4 (1980)
1017–1024.
[12] E.K. Rowinsky, V. Chaudhry, D.R. Cornblath, R.C. Donehower, Neurotoxicity of
taxol. J. Natl. Cancer Inst. 15 (1993) 107–115.
[13] S.P. Cole, K.E. Sparks, K. Fraser, D.W. Loe, C.E. Grant, G.M. Wilson, R.G. Deeley,
Pharmacological characterization of multidrug resistant MRP-transfected
human tumor cells. Cancer Res. 54 (1994) 5902–5910.
[14] A. Lorico, G. Rappa, R.A. Flavell, A.C. Sartorelli, Double knockout of the mrp gene
leads to increased drug sensitivity in vitro. Cancer Res. 56 (1996) 5351–5355.
[15] P. Giannakakou, D.L. Sackett, Y.K. Kang, Z. Zhan, J.T. Buters, T. Fojo,
M.S. Poruchynsky, Paclitaxel-resistant human ovarian cancer cells have
mutant beta-tubulin that exhibit impaired paclitaxel-driven polymerization.
J. Biol. Chem. 272 (1997) 17118–17125.
[16] M. Kavallaris, D.Y.S. Kuo, C.A. Burkhart, D.L. Regl, M.D. Norris, M. Haber,
S.B. Horwitz, Taxol-resistant epithelial ovarian tumors are associated with
altered expression of specific beta-tubulin isotypes. J. Clin. Invest. 100 (1997)
1282–1293.
[17] H.C. Pitot, E.A. McElroy Jr., J.M. Reid, A.J. Windebank, J.A. Sloan, C. Erlichman,
P.G. Bagniewski, D.L. Walker, J. Rubin, R.M. Goldberg, A.A. Adjei, M.M. Ames,
Phase I trial of Dolastatin-10 (NSC 376128) in patients with advanced solid
tumors. Clin. Cancer Res. 5 (1999) 525–531.
[18] N.-H. Nam, Combretastatin A-4 analogues as antimitotic antitumor agents.
Curr. Med. Chem. 10 (2003) 1697–1722.
[19] G.C. Tron, T. Pirali, G. Sorba, F. Pagliai, S. Busacca, A.A. Genazzani, Medicinal
chemistry of combretastatin A-4: present and future directions. J. Med. Chem.
49 (2006) 3033–3044.
[20] S. Desbene, S. Giorgi-Renault, Drugs that inhibit tubulin polymerization: the
particular case of podophyllotoxin and analogues. Curr. Med. Chem. – Anti-
Cancer Agents 2 (2002) 71–90.
[41] D.G. Weiss, W. Maile, R.A. Wick, W. Steffen, Video microscopy. in: A.J. Lacey
(Ed.), Light Microscopy in Biology. A Practical Approach, second ed. Oxford
University Press, USA, 1999, pp. 73–105.
´ ˆ
[42] J. Lo¨we, H. Li, K.H. Downing, E. Nogales, Refined structure of aa-tubulin at
3.5 Å. J. Mol. Biol. 313 (2001) 1045–1057.