Regioselective Suzuki coupling of dihaloheteroaromatic compounds as a rapid strategy to synthesize potent rigid combretastatin analogues

Article abstract of DOI:10.1021/jm200555r

Combretastatin A-4 (CA-4) is a potent tubulin depolymerizing agent able to inhibit tumor growth and with antivascular effects. Although it is in clinical trials, the search for novel analogues that may display better/different features is still ongoing. I

Full text of DOI:10.1021/jm200555r

ARTICLE
pubs.acs.org/jmc
Regioselective Suzuki Coupling of Dihaloheteroaromatic
Compounds as a Rapid Strategy To Synthesize Potent Rigid
Combretastatin Analogues
Sewan Theeramunkong,† Antonio Caldarelli,† Alberto Massarotti, Silvio Aprile, Diego Caprioglio,
Roberta Zaninetti, Alessia Teruggi, Tracey Pirali, Giorgio Grosa, Gian Cesare Tron,*
and Armando A. Genazzani
Dipartimento di Scienze Chimiche, Alimentari, Farmaceutiche e Farmacologiche, Universitꢀa degli Studi del Piemonte Orientale
“A. Avogadro”, Via Bovio 6, 28100 Novara, Italy
S Supporting Information
b
ABSTRACT: Combretastatin A-4 (CA-4) is a potent tubulin depolymeriz-
ing agent able to inhibit tumor growth and with antivascular effects. Although
it is in clinical trials, the search for novel analogues that may display better/
different features is still ongoing. In this manuscript we describe the synthesis
of novel constrained analogues of CA-4 obtained in only two synthetic steps
exploiting a regioselective Suzuki coupling of dihalogenated heteroaromatic
and alicyclic compounds. All the compounds synthesized have been eval-
uated for cytotoxicity and for their ability to inhibit tubulin assembly. One of
them, 38, displayed low nanomolar cytotoxicity and proved to have a
pharmacodynamic profile similar to that of CA-4 and a better pharmacoki-
netic profile, but most important of all, this synthetic strategy may pave
the way for the easy and rapid generation of novel rigid analogues of
combretastatins.
’ INTRODUCTION
Combretastatin A-4 (CA-4, 1) is a natural product derived
from the African bush willow Combretum caffrum and has been
first described over 20 years ago.1 It is a strong tubulin depoly-
merizing agent and therefore inhibits tumor growth and has
antivascular effects.2 Its prodrug (disodium salt water-soluble
phosphate derivative) has now entered clinical trials for both solid
and liquid tumors.3 Similarly, a close derivative of combretastatin
Figure 1. Rationale for the constrainment of combretastatin A-4.
A4 (AVE8062)4 is also undergoing clinical trials. A number of
trials have been initiated with these drugs (www.clinicaltrials.gov),
and both compounds have demonstrated a sufficient safety profile
in phase I trials, suggesting that this may be a viable therapeutic
strategy.
It is important to note that (i) the cis olefinic bridge is able to
undergo rapid cisÀtrans isomerization under the influence of
heat, light, and protic media5 and that (ii) the olefinic bridge
represents a weak point for metabolic stability.7 It is therefore
not surprising that a number of studies have attempted to
replace the olefinic bridge, in particular with more rigid and
metabolically stable structures able to maintain the correct
conformation of the two adjacent rings.5,8 To this end, one of
the strategies consists of the replacement of the olefinic bond
with five-membered heterocyclic rings (Figure 1). Many of these
novel compounds have been shown to be active, and indeed,
some of the new compounds have been shown to display an
increased potency. A second reason to replace the olefinic bridge
is given by the difficulty in adding moieties to the two ring
Despite its low molecular weight and simple molecular
structure, CA-4 (Figure 1) is one of the most powerful inhibitors
of tubulin polymerization known to date. The structure can be
divided into three separate components, the two rings (usually
termed A and B) and the olefinic bridge. The structureÀactivity
relationship (SAR) of CA-4 has been investigated thoroughly by
a number of groups and is reasonably well-understood,5 although
full SAR comprehension may be in part hampered by synthetic
limitations and by literature bias. In brief, it is thought that the cis-
configuration of the olefinic bridge, the presence of the 3,4,5-
trimethoxy group on ring A, and the para-methoxy group on ring
B are all fundamental for antitubulin activity. Nonetheless, active
compounds not obeying this general rule have been described.6
Received: January 26, 2011
Published: June 22, 2011
r
2011 American Chemical Society
4977
dx.doi.org/10.1021/jm200555r J. Med. Chem. 2011, 54, 4977–4986
|

Journal of Medicinal Chemistry
ARTICLE
Figure 2. Selected examples of regioselective couplings of polyhaloheteroaromatic compounds.
Scheme 1. Synthesis of Combretathiophenes 6À31
structures, as both appear to have direct interactions with tubulin.
For example, to generate dual-acting drugs or drugs that bear
additional functionalities, it may be envisaged that a ring repla-
cing the olefinic bring may be the best harbor. Indeed, this has
been shown to be possible.9
impact changes on the two rings, as it is difficult to compare data
from different laboratories and different rigidified series.
Following this line of thought, we were intrigued by the
possibility of generating heterocyclic analogues of combretasta-
tin through a regioselective Suzuki coupling of polyhalogenated
heteroaromatics. In the past decade this strategy has gained a
prominent role in the synthesis of pharmaceuticals and natural
products thanks to the possibility of discriminating between the
reactivities of the carbonÀhalogen bonds in the oxidative addi-
tion step (Figure 2).12
Although exceptions are well documented,10 one possible con-
cern regarding this strategy is due to an increase of the number of
synthetic steps required (compared to the synthesis of CA-4)11
which can hamper a full exploitation of SAR studies. Second, it
is difficult to unravel how modifications in the olefinic bridge
4978
dx.doi.org/10.1021/jm200555r |J. Med. Chem. 2011, 54, 4977–4986

Journal of Medicinal Chemistry
ARTICLE
Scheme 2. Synthesis of Combretafurans 33 and 34
Indeed, this strategy would enable us to generate, also
through parallel synthesis with a minimum number of synthetic
steps, a series of rigidified CA-4 analogues bearing a number of
substituents. If this strategy were to work, it would be possible to
evaluate methodically an increased number of ring A and B
substitutions to generate a complete SAR of CA-4. The limit
on the number of analogues generated would be only given
by the number of polyhalogenated heteroaromatics available or
synthesizable.
In this manuscript, we describe the synthesis and the
biological evaluation of new furan, thiophene, and cyclopentene
rigid analogues of CA-4 prepared in only two synthetic steps
using a regioselective Suzuki coupling strategy. Following this
strategy, we now report two rigidified analogues that display
IC50 for cytotoxicity in the low nanomolar range and, more
importantly, a strategy to develop limitless CA-4 analogues in
few synthetic steps.
to its analogues, usually leads to an increased number of synthetic
steps, with the use of protecting groups. It is important to stress
that the Suzuki reaction does not require the protection of
the ÀOH group, which confers potency to CA-4, being involved
in a key hydrogen-bond interaction with tubulin.13 Indeed, the
experience in the combretastatin field suggests that the presence
of a phenolic group in the meta position imparts an increased
number of synthetic steps, a more complicated synthetic plan and
a reduced overall yield.14
For the synthesis of combretafurans starting from 3, we
capitalized on the different reactivity in the oxidative coupling
step between the two carbonÀbromine bonds. In this way, it was
possible to insert the first substituted phenyl ring at the 2-position
(3,4,5-trimethoxyphenyl, 4-methoxyphenyl, and 3-hydroxy-4-
methoxyphenyl) to give 32, 35, and 37, respectively. Also in this
case, we did not observe the formation of bis-adduct byproducts.
These intermediates were then coupled with the appropriate
boronic acid (18, 30, 31) to give the final combretafurans 33, 34,
36, and 38 (Schemes 2 and 3).
’ CHEMISTRY
Finally, in order to demonstrate the generality of this concept,
we prepared the true rigid analogue of combretastatin, replacing
the olefinic double bond with a cyclopentene. Starting with the
commercially available 1,2-dibromocyclopent-1-ene and using
our same reaction conditions for the symmetrical dihaloheter-
oaromatics, we obtained the monosubstituted analogue 39 in
23% yield. This low yield was probably due to the thermal
instability of the dibromo analogue and not to the formation of
the bis-adduct byproduct. Compound 39 was then coupled with
(3-hydroxy-4-methoxyphenyl)boronic acid (31) to give the final
compound 40 in 44% yield (Scheme 4).
To accomplish our task, we decided to use both symmetrical
and nonsymmetrical dihalogenated heterocyclic rings, namely,
the commercially available 3,4-dibromothiophene (2) and the
2,3-dibromofuran (3).
Starting with 2, we decided to insert the 3,4,5-trimethoxy-
phenyl group in the first step, as this moiety can behave as an
anchor in the tubulin binding site and to examine several
substitutions on the other side that correspond to the B-ring of
CA-4. Several experimental conditions were tested in order to
reduce the amount of the symmetrical disubstituted analogue
obtainable during the first step. Under these optimized conditions
(3:1 thiophene/boronic acid, sodium carbonate, DMF, 50 °C) we
were able to obtain the intermediate 5 in 62% yield and to recover
the unreacted dibromothiophene without observing the forma-
tion of the bis-adduct. With this intermediate in our hands, we
selected a library of boronic acid derivatives (A6ÀA31) bearing
different substituents and coupled them to 5 to give a library of
25 combretathiophenes (6À31) (Scheme 1).
’ BIOLOGICAL RESULTS
To evaluate the cytotoxic nature of the synthesized com-
pounds, we used a neuroblastoma cell line (SH-SY5Y) that we
had previously characterized as sensitive to combretastatin A-4.15
As an initial screening, we incubated all compounds at 1 μM, as
we aimed at identifying only compounds with nanomolar
potency (Table 1). Among combretathiophenes, compounds
6, 8, 24, 30, and 31 were able to reduce viability/cell growth by at
least 70% as measured by the MTT colorimetric method. On the
The experience in the combretastatin field suggests that the
presence of a hydroxy group in the meta position is crucial. Yet
the presence of this moiety, which confers potency to CA-4 and
4979
dx.doi.org/10.1021/jm200555r |J. Med. Chem. 2011, 54, 4977–4986

Journal of Medicinal Chemistry
ARTICLE
Scheme 3. Synthesis of Combretafurans 36 and 38
Scheme 4. Synthesis of the Cyclopentene Analogue of CA-4 40
other hand, all combretafurans, independent of the ring orienta-
tion, appeared efficacious at this concentration (33, 34, 36, 38).
Similarly, the cyclopentene analogue of combretastatin was also
efficacious (40).
Last, we directly investigated tubulin polymerization, visually
(Figure 5a), on Western blots (Figure 5b) and via a tubulin
polymerization in vitro assay. When cells were treated for 16 h
with 3 times the IC50 concentration, the tubulin cytoskeleton
(recognized with a specific antibody), clearly observable in
control cells, disappeared in all treated samples.
To confirm this finding, cells were grown in the presence of 1,
31, 34, 38, or 40 and proteins were extracted in the presence
of paclitaxel, which prevents further tubulin rearrangements
(Figure 5b). Western blotting of the pelletable fraction and the
soluble fraction then allows distinction between the polymerized
and free form of tubulin.15 Indeed, while in control cells tubulin
was present in similar amounts in the polymerized (pellet) and
free form (supernatant). The equilibrium was significantly
shifted toward the free form when cells were treated with 1 or
with any of the compounds characterized. This is yet further
evidence that inhibition of tubulin polymerization represents the
primary mechanism of action of these compounds.
Last, we performed an in vitro tubulin polymerization assay.
Compound 1 displayed an IC50 of 1.2 ( 0.05. Compound 31
displayed an IC50 of 3.5 ( 0.2. Compound 38 displayed an IC50
of 2.2 ( 1.0, and compound 40 displayed an IC50 of 1.5 ( 0.1.
Compounds 1, 31, and 40 inhibited tubulin polymerization fully,
while compound 38 inhibited about 50% of tubulin polymeriza-
tion. The latter data might in part explain the different cell cycle
pattern (Figure 4) elicited by this molecule.
We then performed full concentrationÀresponse curves of the
compounds selected (see Supporting Information). In this set of
experiments, CA-4 displayed an apparent IC50 of 2.3 ( 1.0 nM.
ConcentrationÀresponse curves supported the initial data on
cytotoxicity, but most compounds had IC50s in the high nano-
molar range. The most potent combretathiophene compound
was 31, with an IC50 of 79 ( 67 nM. Although compounds 6 and
8 were not remarkably potent (IC50 over 300 nM), their activity
is of interest because the moieties present on ring B display some
degree of novelty. To confirm their potential activity on tubulin,
we then performed FACS analysis to investigate their effect on
the cell cycle (Figure 3). Both of these compounds were able to
induce a G2/M block which might be indicative of their
antitubulin activity. Combretafurans and the cyclopentane ana-
logue were, on the other hand, rather potent. The most potent
compound was represented by 34, which displayed an IC50 of
approximately 2.4 ( 1.3 nM.
Analysis of the most potent compounds revealed that 31, 34,
38, and 40 all bear the 4-methoxy-3-hydroxy substitution.
Indeed, this is the archetypical ring substitution that has been
shown by many as being the most efficacious.5
Wethenproceeded to confirm the mechanism of action of these
drugs. As these drugs displayed similar cytotoxicity in HeLa cells
(Table 1), we then proceeded with this characterization in this
latter cell line. First, we performed FACS analysis. From previous
experiences, cells are grown for 16 h in the presence of 3 times the
IC50 concentration and the DNA content is monitored with
propidium iodide. All compounds were able to induce a severe
G2/M block, which eventually leads to cell death (Figure 4).
’ IN VITRO METABOLISM
Human liver microsomes and the S9 system, which contains
both liver cytosol and microsomes, were used to assess the in
vitro metabolic stability of compounds 31, 38, and 40 in
comparison to CA-4. In particular, phase I metabolism was
4980
dx.doi.org/10.1021/jm200555r |J. Med. Chem. 2011, 54, 4977–4986

Journal of Medicinal Chemistry
ARTICLE
Table 1. Cytotoxicity in SH-SY5Y and HeLa Cells of the
Synthesized Compoundsa
of quinone species were observed in the metabolic fate of CA-4,
while this obviously did not occur for the rigidified analogues.
The in vitro human hepatic microsomal metabolism of CA-4
involves three main metabolic pathways: O-demethylation, aro-
matic hydroxylation, and glucuronidation.7 Compounds 31, 38,
and 40 shared a similar metabolic pattern. Indeed, as reported
in the Table 2, the metabolites arising from O-demethylation,
aromatic hydroxylation, and conjugation were observed in LCÀ
MS2 chromatograms (see Supporting Information). Similarly the
formation of quinonic species was also established through the
identification of their GSH adducts. Indeed GSH adducts can be
easily detected by LCÀMS using full-scan and product ion scan
modes (see Supporting Information). However, compound 31
also showed an additional metabolic activation pathway, possibly
arising from the oxidation of thiophene ring. Again, the presence
in the LCÀMS2 traces (m/z = 696, Supporting Information) of
another series of GSH adducts, generated by both S-oxidation
or epoxidation on ring C, could be considered diagnostic. In
contrast, compounds 38 and 40 did not show any additional
metabolic activation on ring C. All compounds tested in the
phase I microsomal incubations showed an extent of transforma-
tion that did not exceed 5% in the first hour of incubation. As
previously reported for CA-4, the glucuronidation pathway was
prominent. This was the case also for compounds 31, 38, and 40.
Yet from a quantitative viewpoint, the glucoronidation was
slower for 38 and 40 compared to CA-4.
SHSY-5Y
HeLa
compd
viability (1 μm), %
IC50, nM
IC50, nM
1
23 ( 2
27 ( 4
80 ( 2
30 ( 1
94 ( 8
52 ( 3
97 ( 4
93 ( 8
84 ( 7
100 ( 16
108 ( 11
45 ( 3
91 ( 5
91 ( 3
49 ( 3
83 ( 2
90 ( 4
90 ( 5
91 ( 5
28 ( 1
82 ( 14
65 ( 9
93 ( 6
51 ( 3
109 ( 8
31 ( 1
29 ( 3
23 ( 1
23 ( 1
24 ( 1
20 ( 2
33 ( 6
2.3 ( 1.0
350 ( 70
1.7 ( 4.1
420 ( 67
6
7
8
310 ( 46
420 ( 21
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
33
34
36
38
40
’ MOLECULAR MODELING
330 ( 120
430 ( 17
Molecular modeling studies were performed to investigate the
potential binding ability of these compounds to the colchicine
binding site of R,β-tubulin. Docking studies showed that 34,
the most active compound, occupies the colchicine binding site
of tubulin in agreement with the X-ray structure complex of
DAMAÀcolchicineÀtubulin.17 The trimethoxyphenyl moiety in
ring A of 34 was positioned in the binding cavity buried in the
β-subunit. The thiol group of βC241 formed a hydrogen bond
with the oxygen atom of one of the methoxy groups, and several
amino acids of β-tubulin formed hydrophobic interactions with
the trimethoxyphenyl moiety of 34. The hydroxyl group of ring B
formed a hydrogen bond with the main chain nitrogen atom of
RV181 (Figure 6). Also in the case of compounds 31, 38, and 40
the docked pose is similar to the colchicine binding mode. The
missing hydrogen bonding interaction with RV181 possibly
explains the reduced biological activity of compounds 30, 33,
and 36 (see Supporting Information).
480 ( 107
79 ( 67
28 ( 1.0
2.4 ( 1.3
34 ( 12
9.4 ( 1.7
37 ( 20
420 ( 34
200 ( 79
17 ( 5.2
3.1 ( 0.8
170 ( 55
20.0 ( 11.5
300 ( 180
a Values in the second column represent % of viable cells after 48 h of
treatment with 1 μM compound. Values in the last column represent the
calculated IC50 values for those compounds in which a full concentration
response curve was performed. Values are the mean ( SD of 12
determinations from 3 separate experiments. IC50 values were deter-
mined only for the most potent compounds.
’ CONCLUSIONS
performed in microsomal incubations in the presence of a
NADP(H)-regenerating system, allowing evaluation of the me-
tabolites arising from oxidative metabolic pathways. Phase II
metabolism was investigated in microsomes and S9 fractions by
using the appropriate cofactor to evaluate the glucuronide and
sulfate conjugates formation. Moreover, microsomal incubations
were also carried out in the presence of the nucleophilic trapping
agent GSH capable of reacting with electrophilic intermediate
species. Indeed, the formation of reactive intermediates in the
metabolism represents a potential liability in drug discovery and
development potentially leading to adverse reactions and/or
toxicity via covalent binding to cellular proteins and enzymes.16
First, the ZÀE isomerization of the olefin bond and the formation
The present manuscript describes the use of a two-step
regioselective Suzuki to generate potent rigid analogues of
combretastatin. As a proof of principle, three different scaffolds
were prepared. All scaffolds prepared were novel, and therefore,
direct comparisons with alternative synthetic strategies are not
possible. Nonetheless, literature/patent data can be used to
support the easiness of this strategy.9,18 For example, we have
previously synthesized 3,4-diphenyl substituted furan compounds,9
which can therefore be directly compared to the 2,3-diphenyl
substituted furans described here. The furan with the 3,4,5-
trimethoxyphenyl ring A and the 4-methoxy-3-hydroxyphenyl
ring B was prepared in five synthetic steps using a complicated
intramolecular aldol condensation as a key step.
4981
dx.doi.org/10.1021/jm200555r |J. Med. Chem. 2011, 54, 4977–4986

Journal of Medicinal Chemistry
ARTICLE
Figure 3. Analogues bearing atypical substitutions on ring A induce cell cycle arrest. Cell cycle analysis of Hela cells treated for 16 h with vehicle (ctrl) or
3 Â IC50 values of indicated compounds. Data are representative of three separate experiments. Similar results were obtained in separate experiments
when drugs were incubated for 8 or 24 h. The Y-axis represents cell number, and the X-axis represents fluorescence on a linear scale.
CA-4 has already entered clinical trials. Possible improve-
ments over this molecule might be metabolic stability, the
absence of cisÀtrans double bond isomerization (which might
lead to off-target effects), and a longer half-life. It could therefore
be envisaged that analogues that retain the potency and efficacy
of CA-4 but that are different in the pharmacokinetic profile
might be useful. Indeed, it should also be considered that drug
interactions are becoming a sensible issue in cancer therapy,20
and therefore, having more than one option could be desirable in
the different therapeutic schemes. In this respect, the synthetic
strategy described here might be helpful and the improved
pharmacokinetic stability of two of the five-membered rings
described (38 and 40) might be perceived as an improvement.
As tubulin isoforms are different in different tumor types,21 it
could also be envisaged that this strategy could be employed to
generate a battery of novel analogues (bearing a five-membered
ring or any other kind) with increased selectivity. Nonetheless, a
number of rigidified analogues have already been synthesized,
and therefore, it is time to seek such differences.
Melting points were determined in open glass capillary with a Stuart
scientific SMP3 apparatus and are uncorrected. All the compounds were
1
checked by IR (FT-IR THERMO-NICOLET AVATAR), H and 13C
APT (JEOL ECP 300 MHz), and mass spectrometry (Thermo Finningan
LCQ-deca XP-plus) equippedwith anESI sourceand anion trapdetector.
Chemical shifts are reported in parts per million (ppm). Column chro-
matography was performed on silica gel Merck Kieselgel 70À230 mesh
ASTM or silica gel Merck Kieselgel 230À400 ASTM when using Biotage
Isolera. Thin layer chromatography (TLC) was carried out on 5 cm Â
20 cm plates with a layer thickness of 0.25 mm (Merck silica gel 60 F254).
When necessary, they were developed with KMnO4. The purity of the
target compounds (>95%) was analyzed via elemental analysis and was
within (0.4% of the calculated value. All molecular modeling calculations
were performed using the software Autodock 419 running on a Tesla
workstation equipped with two Intel Xenon X5650, 2.67 GHz.
The geometry of the colchicine was taken from the structure of the
tubulin filed in the Brookhaven Protein Data Bank (entry code 1SA0).16
For the docking a grid spacing of 0.375 Å and 56 Â 60 Â 50 number of
points were used. The grid was centered on the mass center of the
colchicine coordinates. The GA-LS method was adopted using the
default settings. Charges were assigned to the protein using the program
ADT (AutoDock Tools). Autodock generated 100 possible binding
conformations for ligands. To validate the use of the Autodock program,
the docking studies were performed on the reference compounds
colchicine and combretastatin A-4. Autodock successfully reproduced
the binding conformations reported in the literature with acceptable
root-mean-square deviation (rmsd < 1 Å) of atom coordinates.
’ EXPERIMENTAL SECTION
Chemistry. Commercially available reagents and solvents were used
without further purification. Dimethylformamide (DMF) was purified
by distillation at reduced pressure, collecting the fraction having bp
76 °C at 39 mmHg and storing on activated molecular sieves (4 Å).
When needed, the reactions were performed in flame- or oven-dried
glassware under a positive pressure of dry N2 or in a Schlenk tube.
3-Bromo-4-(3,4,5-trimethoxyphenyl)thiophene (5). To a so-
lution of commercially available 3,4-dibromothiophene (1.92 g, 7.97 mmol)
4982
dx.doi.org/10.1021/jm200555r |J. Med. Chem. 2011, 54, 4977–4986

Journal of Medicinal Chemistry
ARTICLE
Figure 4. Combretastatin and the most potent analogues synthesized induce cell cycle arrest. Cell cycle analysis of Hela cells treated for 16 h with vehicle
(ctrl) or 3 Â IC50 values of indicated compounds. Data are representative of three separate experiments. Similar results were obtained in separate
experiments when drugs were incubated for 8 or 24 h. The Y-axis represents cell number, and the X-axis represents fluorescence on a linear scale.
and 3,4,5-trimethoxybenzene boronic acid (0.56 g, 2.66 mmol) in 43 mL of
DMF, the tetrakis(triphenylphosphine)palladium (0.15 g, 0.13 mmol) and
an aqueous solution of sodium carbonate monohydrate (0.99 g 7.97 mmol)
were added, and the mixture was stirred. The mixture was then warmed at
50 °C for 4.5 h. The mixture was diluted with water and extracted with ethyl
acetate (Â3). The combined organic fractions were washed with brine
(Â1), dried over sodium sulfate, and concentrated under vacuum. The
crude product was purified by flash column chromatography by using
petroleum ether/ethyl acetate (8:2) as eluent to give 0.52 g of 5 as a white
solid (59%).
dissolved in DMF, and 1 equiv of the corresponding boronic acid (A6-31)
was added. The tetrakis(triphenylphosphine)palladium (0.05 equiv) and
an aqueous solution of sodium carbonate monohydrate (3 equiv) were
added and stirred. The mixture was then heated at 80 °C until the
reaction was finished. The mixture was diluted with water and
extracted with ethyl acetate (Â3). The combined organic fractions
were washed with brine (Â1) and dried over anhydrous sodium sulfate
and concentrated in vacuo. The crude product was purified by column
chromatography.
3-Bromo-2-(3,4,5-trimethoxyphenyl)furan (32). To a solu-
tion of commercially available 2,3-dibromofuran (1 g, 4.42 mmol)
and 3,4,5-trimethoxybenzene boronic acid (0.39 g, 1.84 mmol) in
General Procedure fortheSynthesisofThiophenes(6À31).
The 3-bromo-4-(3,4,5-trimethoxyphenyl)thiophene (5) (1.2 equiv) was
4983
dx.doi.org/10.1021/jm200555r |J. Med. Chem. 2011, 54, 4977–4986

Journal of Medicinal Chemistry
ARTICLE
Figure 6. Superimposition of the presumptive conformation of active
34 (stick model) on top of the X-ray structure of DAMAÀcolchicine
(wire model). The structures are docked into the colchicine binding site
of tubulin (PDB entry 1SA0). The backbone of tubulin is shown as
ribbon representation (R-tubulin, yellow; β-tubulin, blue). The amino
acids of tubulin within 4.0 Å from colchicine are shown as wire models.
Hydrogen bonds (distance of <3 Å) are shown as dotted lines.
sulfate, and concentrated under vacuum. The crude product was
purified by flash column chromatography by using petroleum ether/
ethyl acetate (8:2) as eluent to give 0.38 g of a white solid (66%).
3-(4-Methoxyphenyl)-2-(3,4,5-trimethoxyphenyl)furan (33).
The previously synthesized 3-bromo-2-(3,4,5-trimethoxyphenyl)furan
(33) (100 mg, 0.32 mmol) was dissolved in DMF (2.5 mL) and stirred.
The commercially available 4-methoxybenzene boronic acid (40 mg,
0.27 mmol), tetrakis(triphenylphosphine)palladium (15 mg, 0.013 mmol),
and an aqueous solution of sodium carbonate monohydrate (99 mg,
0.80 mmol) were added, and the mixture was stirred. The mixture was
then heated at 70 °C for 3 h. The mixture was diluted with water and
extracted with ethyl acetate (Â3). The combined organic fractions were
washed with brine (Â1), dried over sodium sulfate, and concentrated
under vacuum. The crude product was purified by flash column
chromatography by using petroleum ether/ethyl acetate (8:2) as eluent
to give 53 mg of a light yellow oil (59%).
Figure 5. Combretastatin and the most potent compounds synthesized
affect tubulin polymerization. (a) Immunofluorescence using an anti-R-
tubulin antibody (green) and DRAQ5 (blue) for nuclear staining of
control cells (ctrl) or cells treated for 24 h with 3 Â IC50 value of the
indicated compounds. Images have been obtained with a 63Â oil
immersion objective. (b) Western blot of R-tubulin extracted in the
presence of paclitaxel from Hela cells treated with the indicated
compounds at 3 Â IC50 value. Results are representative of three
separate experiments. P = pelletable fraction. S = soluble fraction.
2-Methoxy-5-[2-(3,4,5-trimethoxyphenyl)-3-furyl]phenol
(34). To a solution of the previously synthesized 3-bromo-2-(3,4,
5-trimethoxyphenyl)furan (32), 60 mg, 0.19 mmol) and (3-hydroxy-
4-methoxyphenyl)boronic acid (29 mg, 0.16 mmol) in 2.5 mL of DMF,
the tetrakis(triphenylphosphine) palladium (9 mg, 0.01 mmol) and an
aqueous solution of sodium carbonate monohydrate (59 mg 0.80 mmol)
were added, and the mixture was stirred. The mixture was then heated at
70 °C for 2 h. The mixture was diluted with water and extracted with
ethyl acetate (Â3). The combined organic fractions were washed with
brine (Â1) and dried over sodium sulfate and concentrated under
vacuum. The crude product was purified by flash column chromatog-
raphy by using petroleum ether/ethyl acetate (8:2) as eluent to give 31
mg of an amorphous yellow solid (54%).
3-Bromo-2-(4-methoxyphenyl)furan (35). To a solution of
commercially available 2,3-dibromofuran (0.8 g, 3.54 mmol) and 4-meth-
oxybenzene boronic acid (0.224 g, 1.48 mmol) in 8 mL of DMF, the
tetrakis(triphenylphosphine)palladium (0.085 g, 0.07 mmol) and an aqu-
eous solution of sodium carbonate monohydrate (0.55 g, 4.43 mmol) were
added, and the mixture was stirred. The mixture was then warmed at 50 °C
for 3 h. The mixture was diluted with water and extracted with ethyl acetate
(Â3). The combined organic fractions were washed with brine (Â1) and
dried over sodium sulfate and concentrated under vacuum. The crude
product was purified by flash column chromatography by using petroleum
ether/ethyl acetate (9:1) as eluent to give 0.24 g of a yellow oil (65%).
Table 2. Metabolic Profile for CA-4, 31, 38, and 40
metabolic pathway
CA-4 31
38
40
O-demethylation
+
+
+
+
+
aromatic hydroxylation (ring B)
metabolic activation
+
+
+
quinones
ring C
+
+
+
+
À
e5
+
À
e5
+
À
e5
+
phase I transformation(%)a
conjugation
e5
+
glucuronides +
sulfates
+
+
+
+
phase II transformation (%)a
a Expressed as peak area percent obtained from LCÀUV analyses of
incubation extracts (t = 60 min).
∼40 ∼40 ∼20 ∼30
10 mL of DMF, the tetrakis(triphenylphosphine)palladium (0.11 g,
0.09 mmol) and an aqueous solution of sodium carbonate monohy-
drate (0.68 g 5.51 mmol) were added, and the mixture was stirred. The
mixture was then warmed at 50 °C for 5 h. The mixture was diluted
with water and extracted with ethyl acetate (Â3). The combined
organic fractions were washed with brine (Â1), dried over sodium
4984
dx.doi.org/10.1021/jm200555r |J. Med. Chem. 2011, 54, 4977–4986

Journal of Medicinal Chemistry
ARTICLE
2-(4-Methoxyphenyl)-3-(3,4,5-trimethoxyphenyl)furan (36).
The previously synthesized 3-bromo-2-(4-methoxyphenyl)furan (35)
(100 mg, 0.40 mmol) was dissolved in DMF (2.5 mL), and the mixture was
stirred. The commercially available 3,4,5-trimethoxybenzene boronic acid
(70 mg, 0.33 mmol), tetrakis(triphenylphosphine)palladium (19 mg,
0.016 mmol), and an aqueous solution of sodium carbonate monohydrate
(123 mg, 0.99 mmol) were added, and the mixture was stirred. The mixture
was then heated at 70 °C for 3 h. The mixture was diluted with water and
extracted with ethyl acetate (Â3). The combined organic fractions were
washed with brine (Â1) and dried over sodium sulfate and concentrated
under vacuum. The crude product was purified by flash column chroma-
tography by using petroleum ether/ethyl acetate (8:2) as eluent to give
54.1 mg of an amorphous yellow solid (48%).
5-(3-Bromo-2-furyl)-2-methoxyphenol (37). To a solution of
the commercially available 2,3-dibromofuran (1.1 g, 4.43 mmol) and
(3-hydroxy-4-methoxyphenyl)boronic acid (0.33 g, 1.84 mmol) in
10 mL of DMF, the tetrakis(triphenylphosphine)palladium (0.11 g,
0.09 mmol) and an aqueous solution of sodium carbonate monohydrate
(0.69 g 5.53 mmol) were added, and the mixture was stirred. The mixture
was then warmed at 50 °C for 5 h. The mixture was diluted with water and
extracted with ethyl acetate (Â3). The combined organic fractions were
washed with brine (Â1) and dried over sodium sulfate and concentrated
under vacuum. The crude product was purified by flash column chroma-
tography by using petroleum ether/ethyl acetate (8:2) as eluent to give
0.42 g of an amorphous reddish solid (84%).
2-Methoxy-5-[3-(3,4,5-trimethoxyphenyl)-2-furyl]phenol
(38). To a solution of the previously synthesized 5-(3-bromo-2-furyl)-2-
methoxyphenol (37) (110 mg, 0.41 mmol) and the commercially available
3,4,5-trimethoxybenzene boronic acid (72 mg, 0.34 mmol) in 2.5 mL of
DMF, tetrakis(triphenylphosphine)palladium (20 mg, 0.02 mmol) and an
aqueous solution of sodium carbonate monohydrate (127 mg, 1.02 mmol)
were added, and the mixture was stirred. The mixture was then heated at
70 °C for 2 h. The mixture was diluted with water and extracted with ethyl
acetate (Â3). The combined organic fractions were washed with brine
(Â1) and dried over sodium sulfate and concentrated under vacuum.
The crude product was purified by flash column chromatography by
using petroleum ether/ethyl acetate (8:2) as eluent to give 0.08 g of an
amorphous yellow solid (66%).
5-(2-Bromocyclopent-1-en-1-yl)-1,2,3-trimethoxybenzene
(39). To a solution of the commercially available 1,2-dibrocyclopentene
(0.73 g, 3.22 mmol) and 3,4,5-trimethoxybenzene boronic acid
(0.20 g, 0.92 mmol) in 10 mL of DMF, tetrakis(triphenylphosphine)
palladium (0.05 g, 0.05 mmol) and an aqueous solution of sodium
carbonate monohydrate (0.34 g, 2.76 mmol) were added, and the
mixture was stirred. The mixture was then warmed at 50 °C for 2 h.
The mixture was diluted with water and extracted with ethyl acetate
(Â3). The combined organic fractions were washed with brine (Â1)
and dried over sodium sulfate and concentrated under vacuum. The
crude product was purified by flash column chromatography by using
petroleum ether/ethyl acetate (8:2) as eluent to give 0.07 g of a yellow
oil (23%).
Cell Culture and Cytotoxicity Assay. The SH-SY5Y human
neuroblastoma cell line was obtained from ATCC (LGC Promo-
chem Teddington, U.K.) and cultured in 50% DMEM and 50% F-12
supplemented with 10% fetal bovine serum, 2 mM L-glutamine,
penicillin (100 μg/mL), and streptomycin (100 μg/mL). For cytotoxi-
city assays, cells were plated on 24-well plates and grown for 48 h in the
presence or absence of combretastatin A-4 or the synthesized com-
pounds. On the experimental day, cells were washed twice in Locke’s
solution (134 mM NaCl, 5 mM KCl, 4 mM NaHCO3, 10 mM HEPES
[pH 7.6], 2.3 mM CaCl2, 1 mM MgCl2, 5 mM sucrose) and incubated
for 30 min with MTT (250 μg/mL in Locke’s solution) at 37 °C.
Reactions were then stopped and the crystals solubilized in isopropyl
alcohol/HCl before being read at 570 nm in a spectrophotometer. To
determine IC50 values, data were plotted and fitted using the Kaleida-
graph software (Synergy Software, Reading, PA).
Immunocytochemistry. SH-SY5Y were plated on glass coverslips
and grown to subconfluency in the presence or absence of drugs for 24 h.
Coverslips were then fixed in paraformaldehyde (3.7%) for 20 min at
room temperature. Cells were then washed once in phosphate buffered
saline (PBS) and permeabilized in ice cold 80% MeOH for 10 min in ice.
Cells were then washed twice in PBS and blocked with 0.2% cold fish
gelatine. Anti-R-tubulin (mouse) primary antibody (Sigma-Aldrich,
1:1000) was incubated for 20 min at room temperature, and after three
further washes, coverslips were incubated with a goat anti-mouse Alexa
Fluo 488 (Molecular Probes, 1:2000) and Draq 5 (Cell Signaling
1:1000). Slides were then visualized in a Leica SP2 confocal microscopy.
In Vivo and in Vitro Tubulin Polymerization Assay. To
measure the degree of tubulin polymerization in vivo, we used an
adaptation of the methods described by Minotti et al.15 In brief, cells were
grown in 75 cm2 flasks in the presence or absence of drugs (3 Â IC50) for
16 h. Cells were then tripsinized and centrifuged at 600g for 3 min and
washed once by PBS. Cells were then resuspended in 70 μL of hypotonic
buffer (20 mM Tris-HCl, pH 6.8, 1 mM MgCl2, 2 mM EGTA, protease
inhibitors, 0.5% NP-40) containing 4 μg/mL paclitaxel. Lysates were
incubated for 10 min at room temperature. Lysates were then corrected for
protein amounts (Bradford assay, Sigma-Aldrich), and 50 μg (corre-
sponding to 50 μL) was centrifuged at 13 000 rpm for 15 min at room
temperature. Supernatant and pellet were then resuspended in equal
volumes of SDS loading buffer and run on a 10% SDSÀpolyacrylamide
gel. After transferof proteins to nitrocellulose (blockedin 5% milk), tubulin
was identified with an anti-R-tubulin primary antibody (1:1000, Sigma-
Aldrich) and anti-mouse secondary antibody peroxidase-conjugated
(1:8000, Amersham Bioscience) and visualized by chemiluminescence
(Supersignal WestPico, Pierce). In vitro polymerization was performed
using a commercially available kit (Cytoskeleton Inc.).
Flow-Cytometric Analysis of Cell-Cycle Status. SH-SY5Y cells
grown in the presence or absence of compounds for 16 h were washed
once in PBS and resuspended in 1 mL of 30:70 ice cold PBS/EtOH and
stored at À20 °C. Cells were then washed twice in PBS and resuspended
in PBS containing RNase (100 μg/mL) for 30 min at room temperature.
DNA was then stained with a PBS solution containing 5 mM EDTA and
100 μg/mL propidium iodide for 30 min at room temperature in the dark.
Cell cycle analysis was determined with a FACSVantage SE DiVa (Becton
Dickinson).
2-Methoxy-5-[2-(3,4,5-trimethoxyphenyl)cyclopent-1-en-1-
yl]phenol (40). To a solution of the previously synthesized 5-
(2-bromocyclopent-1-en-1-yl)-1,2,3-trimethoxybenzene (39) (60 mg,
0.19 mmol) and (3-hydroxy-4-methoxyphenyl)boronic acid (27 mg,
0.16 mmol) in 2.5 mL of DMF, tetrakis(triphenylphosphine) palladium
(9 mg, 0.008 mmol) and an aqueous solution of sodium carbonate
monohydrate (59 mg, 0.48 mmol) were added, and the mixture was
stirred. The mixture was then heated at 70 °C for 1 h. The mixture was
diluted with water and extracted with ethyl acetate (Â3). The combined
organic fractions were washed with brine (Â1) and dried over sodium
sulfate and concentrated under vacuum. The crude product was purified
by flash column chromatography by using petroleum ether/ethyl acetate
(8:2) as eluent to give 0.014 g of a white solid (44%).
In Vitro Metabolism and Modeling. The procedure to inves-
tigate in vitro metabolism and the modeling protocol are detailed in the
Supporting Information.
’ ASSOCIATED CONTENT
S
Supporting Information. Characterization of all com-
b
pounds synthesized (mp, IR, MS, and 1H and 13C NMR data),
elemental analysis results of all target compounds, molecular
modeling binding modes of 30, 31, 33, 36, 38, and 40, and
4985
dx.doi.org/10.1021/jm200555r |J. Med. Chem. 2011, 54, 4977–4986

Journal of Medicinal Chemistry
ARTICLE
metabolism procedures and results for 1, 31, 38, and 40. This
material is available free of charge via the Internet at http://pubs.
acs.org.
(9) Pirali, T.; Busacca, S.; Beltrami, L.; Imovilli, D.; Pagliai, F.;
Miglio, G.; Massarotti, A.; Verotta, L.; Tron, G. C.; Sorba, G.; Genazzani,
A. A. Synthesis and cytotoxic evaluation of combretafurans, potential
scaffolds for dual-action antitumoral agents. J. Med. Chem. 2006, 49,
5372–5376.
’ AUTHOR INFORMATION
(10) (a) Cafici, L.; Pirali, T.; Condorelli, F.; Del Grosso, E.;
Massarotti, A.; Sorba, G.; Canonico, P. L.; Tron, G. C.; Genazzani,
A. A. Solution-phase parallel synthesis and biological evaluation of
combretatriazoles. J. Comb. Chem. 2008, 10, 732–740. (b) Romagnoli,
R.; Baraldi, P. G.; Cruz-Lopez, O.; Lopez Cara, C.; Carrion, M. D.;
Brancale, A.; Hamel, E.; Chen, L.; Bortolozzi, R.; Basso, G.; Viola, G.
Synthesis and antitumor activity of 1,5-disubstituted 1,2,4-triazoles
as cis-restricted combretastatin analogues. J. Med. Chem. 2010, 53,
4248–4258. Wang, L.; Woods, K. W.; Li, Q.; Barr, K. J.; McCroskey,
R. W.; Hannick, S. M.; Gherke, L.; Credo, R. B.; Hui, Y.-H.; Marsh, K.;
Warner, R.; Lee, J. Y.; Zielinski-Mozng, N.; Frost, D.; Rosemberg, S. H.;
Sharm, H. L. Potent orally active heterocycle-based combretastatin A-4
analogues: synthesis, structureÀactivity relationship, pharmacokinetics,
and in vivo antitumor activity. J. Med. Chem. 2002, 45, 1697–1711.
Ohsumi, K.; Hatanaka, T.; Fujita, K.; Nakagawa, R.; Fukuda, Y.; Nihei,
Y.; Suga, Y.; Morinaga, Y.; Akiyama, Y.; Tsuji, T. Bioorg. Med. Chem. Lett.
1998, 8, 3153–3158.
Corresponding Author
*Phone: +39-0321-375857. Fax: +39-0321-375821. E-mail: tron@
pharm.unipmn.it.
Author Contributions
†These authors equally contributed to the present work.
’ ACKNOWLEDGMENT
Financial support from M.I.U.R. PRIN 2008 Italy and Asso-
ciazione per la Ricerca sul Cancro (AIRC) Italy are gratefully
acknowledged. S.T. thanks Italian Ministry of Foreign Affairs for
the Ph.D. fellowship.
’ ABBREVIATIONS USED
(11) (a) Cushman, M.; Nagarathnam, D.; Gopal, D.; He, H.-M.; Lin,
C. M.; Hamel, E. J. Med. Chem. 1992, 35, 2293–2306. (b) Gaukroger, K.;
Hadfield, J. A.; Hepworth, L. A.; Lawrence, N. J.; McGown, A. T. Novel
synthesis of cis and trans isomers of combretastatin A-4. J. Org. Chem.
2001, 66, 8135–8138.
CA-4, combretastatin A-4; CA-4P, combretastatin A-4 phosphate;
MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bro-
mide; GSH, glutathione reduced
(12) Handy, S. T.; Zhang, Y. A simple guide for predicting regios-
electivity in the coupling of polyhaloheteroaromatics. Chem. Commun.
2006, 299–301.
’ REFERENCES
(1) Pettit, G. R.; Singh, S. B.; Hamel, E.; Lin, C. M.; Alberts, D. S.; et al.
Isolation and structure of the strong cell growth and tubulin inhibitor
combretastatin A-4. Experientia 1989, 45, 209–211.
(13) (a) Nguyen, T. L.; McGrath, C.; Hermone, A. R.; Burnett, J. C.;
Zaharevitz, D. W.; Day, B. W.; Wipf, P.; Hamel, E.; Gussio, R. A common
pharmacophore for a diverse set of colchicine site inhibitors using
a structure-based approach. J. Med. Chem. 2005, 48, 6107–6116.
(b) Kong, Y.; Grembecka, J.; Edler, M. C.; Hamel, E.; Mooberry, S.;
Sabat, M.; Rieger, J.; Brown, M. L. Structure-based discovery of a boronic
acid bioisostere of combretastatin A-4. Chem. Biol. 2005, 12, 1007–1014.
(14) Tron, G. C.; Pagliai, F.; Del Grosso, E.; Genazzani, A. A.; Sorba,
G. Synthesis and cytotoxic evaluation of combretafurazans. J. Med. Chem.
2005, 48, 3260–3268.
(15) Minotti, A. M.; Barlow, S. B.; Cabral, F. Resistance to anti-
mitotic drugs in Chinese hamster ovary cells correlates with changes in
the level of polymerized tubulin. J. Biol. Chem. 1991, 266, 3987–3994.
(16) Evans, D. C.; Watt, A. P.; Nicoll-Griffith, D. A.; Baillie, T. A.
DrugÀprotein adducts: an industry perspective on minimizing the
potential for drug bioactivation in drug discovery and development.
Chem. Res. Toxicol. 2004, 17, 3–16.
(17) Ravelli, R. B.; Gigant, B.; Curmi, P. A.; Jourdain, I.; Lachkar, S.;
Sobel, A.; Knossow, M. Insight into tubulin regulation from a complex
with colchicines and a stathmin-like domain. Nature 2004, 428, 198–202.
(18) Zhang, W.; Yang, Q.; Wu, Y.; Wu, L.; Li, W.; Qiao, F.; Bao, K.;
Zhang, L. Preparation of 2,3-Diarylthiophene Derivatives As Antitumor
Agents. CN Patent 101429189, 2009. At least nine synthetic steps are
required to prepare the thiophene analogues of combretastatin reported
in this patent.
(19) Morris, G. M.; Goodsell, D. S.; Halliday, R. S.; Huey, R.; Hart,
W. E.; Belew, R. K.; Olson, A. Automated docking using a Lamarckian
genetic algorithm and empirical binding free energy function. J. Comput.
Chem. 1998, 19, 1639–1662.
(2) (a) Bibby, M. C. Combretastatin anticancer drugs. Drugs Future
2002, 27, 475–480.(b) Pinney, K. G.; Jelinek, C.; Edvardsen, K.; Chaplin,
D. J.; Pettit, G. R. The Discovery and Development of the Combretas-
tatins. In Antitumor Agents from Natural Products; Kingston, D., Newman,
D., Cragg, G., Eds.; CRC Press: Boca Raton, FL, 2005; pp 23À46.
(3) (a) Kanthou, C.; Tozer, G. M. The tumor vascular targeting agent
combretastatin A-4 phosphate induces reorganization of the actin cytos-
keleton and early membrane blebbing in human endothelial cells. Blood
2002, 99, 2060–2069. (b) Chaplin, D. J.; Hill, S. A. The devel-
opment of combretastatin A-4 phosphate as a vascular targeting agent.
Int. J. Radiat. Oncol., Biol., Phys. 2002, 54, 1491–1496. (c) West, C. M. L.;
Price, P. Combretastatin A4 phosphate. Anti-Cancer Drugs 2004, 15,
179–186. (d) Siemann, D. W.; Chaplin, D. J.; Walicke, P. A. A review
and update of the current status of the vasculature-disabling agent
combretastatin-A4 phosphate (CA4P). Expert Opin. Invest. Drugs 2009,
18, 189–197.
(4) Delmonte, A.; Sessa, C. AVE8062: a new combretastatin derivative
vascular disrupting agent. Expert Opin. Invest. Drugs 2009, 18, 1541–1548.
(5) 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.
(6) (a) Gaukroger, K.; Hadfield, J. A.; Lawrence, N. J.; Nlan, S.;
McGown, A. T. Structural requirements for the interaction of combre-
tastatins with tubulin: How important is the trimethoxy unit? Org.
Biomol. Chem. 2003, 1, 3033–3037. (b) Beale, T. M.; Myers, R. M.;
Shearman, J. W.; Charnock-Jones, D. S.; Brenton, J. D.; Gergely, F. V.;
Ley, S. V. Med. Chem. Commun. 2010, 1, 202–208.
(20) Riechelmann, R. P.; Del Giglio, A. Drug interactions in
oncology: how common are they? Ann. Oncol. 2009, 12, 1907–1912.
(21) Leandro-Garcia, L. J.; Leskela, S.; Landa, I.; Montero-Conde,
C.; Lopez-Jimꢁenez, E.; Leton, R.; Cascon, A.; Robledo, M.; Rodriguez-
Antona, C. Tumoral and tissue-specific expression of the major human
β-tubulin isotyoes. Cytoskeleton 2010, 67, 214–223.
(7) Aprile, S.; Del Grosso, E.; Tron, G. C.; Grosa, G. In vitro
metabolism study of combretastatin A-4 in rat and human liver micro-
somes. Drug Metab. Dispos. 2007, 35, 2252–2261.
(8) (a) Nam, N.-H. Combretastatin A-4 analogues as antimitotic
antitumor agents. Curr. Med. Chem. 2003, 10, 1697–1722. (b) Hsieh,
H. P.; Liou, J. P.; Mahindroo, N. Curr. Pharm. Des. 2005, 11, 1655–1677.
Chaudari, A.; Pandeya, S. N.; Kumar, P.; Sharma, P. P.; Gupta, S.; Soni,
N.; Verma, K. K.; Bhardwaj, G. Combretastatin A-4 analogues as
anticancer agents. Mini-Rev. Med. Chem. 2007, 12, 1186–1205.
4986
dx.doi.org/10.1021/jm200555r |J. Med. Chem. 2011, 54, 4977–4986