Synthesis and biological activity of fluorinated combretastatin analogues

Article abstract of DOI:10.1021/jm701362m

With the aim of understanding the influence of fluorine on the double bond of the cis-stilbene moiety of combretastatin derivatives and encouraged by a preliminary molecular modeling study showing a different biological environment on the interaction site

Full text of DOI:10.1021/jm701362m

2708
J. Med. Chem. 2008, 51, 2708–2721
Synthesis and Biological Activity of Fluorinated Combretastatin Analogues
Domenico Alloatti, Giuseppe Giannini,* Walter Cabri, Isabella Lustrati, Mauro Marzi, Andrea Ciacci, Grazia Gallo,
M. Ornella Tinti, Marcella Marcellini, Teresa Riccioni, Mario B. Guglielmi, Paolo Carminati, and Claudio Pisano
R&D, Sigma-Tau Industrie Farmaceutiche Riunite S.p.A., Via Pontina Km 30.400, 00040 Pomezia, Italy
ReceiVed October 31, 2007
With the aim of understanding the influence of fluorine on the double bond of the cis-stilbene moiety of
combretastatin derivatives and encouraged by a preliminary molecular modeling study showing a different
biological environment on the interaction site with tubulin, we prepared, through various synthetic approaches,
a small library of compounds in which one or both of the olefinic hydrogens were replaced with fluorine.
X-ray analysis on the difluoro-CA-4 analogue demonstrated that the spatial arrangement of the molecule
was not modified, compared to its nonfluorinated counterpart. SAR analysis confirmed the importance of
the cis-stereochemistry of the stilbene scaffold. Nevertheless, some unpredicted results were observed on a
few trans-fluorinated derivatives. The position of a fluorine atom on the double bond may affect the inhibition
of tubulin polymerization and cytotoxic activity of these compounds.
Introduction
Combretastatin A-4 (CA-4a; compound 1 in Figure 1), a
natural product first isolated in 1989 by G.R.Pettit et al.1 from
the bark of the South African bush willow, Combretum caffrum,
is one of the most potent antimitotic agents. CA-4 shows strong
cytotoxicity against a variety of cancer cells, including multi-
drug-resistant cell lines.2 Several studies3a,b describe its ability
to induce widespread necrosis of solid tumors.
It also exerts highly selective effects in proliferating endot-
Figure 1. Natural and synthetic combretastatins with their correspond-
ing water-soluble prodrug.
helial cells and, as a consequence, demonstrates strong sup-
pressive activity on tumor blood flow (TBF).3c Three derivatives
are currently in clinical trials: CA-4 disodium phosphate (CA-
number of analogues with a modified bridge were synthesized.
4P; 4), a water-soluble prodrug of CA-4 (1);4,5 Oxi-4503, a
water-soluble combretastatin A-1 prodrug (CA-1diP; 5);6,7 and
AC7700 (6), also known as AVE8062 (Figure 1), an ami-
nocombretastatin prodrug developed in Japan in 1998.8,9
As a result of the interesting activity of CA-4 and its
analogues and the simplicity of its structure, a large number of
derivatives were synthesized with the aim of extending
structure-activity relationship (SAR) information to this class
of molecules. As a consequence of these SAR analyses, the cis-
stilbene scaffold of combretastatin appears distinct in three parts:
(a) the trimethoxyphenyl moiety (A-ring), which influences
cytotoxic activity; (b) the double bond that affects both
cytotoxicity and tubulin binding; and (c) the B ring (phenol),
which influences binding to tubulin.2
Saturation of the olefinic bridge (alkane derivatives),2 as well
as an increased unsaturation (alkyne derivatives),10 reduce the
cytotoxic and antitubulin activity of combretastatin. A loss of
activity also occurs with the introduction of groups as large as
-COOH, -CHO, CN, or larger groups, such as heterocyclic
structures,9,11,12,13a with only a few exceptions.13b
The introduction of fluorine atoms into molecules of biologi-
cal interest, such as steroids, carbohydrates, and others, proved
beneficial. Antitumor agents, such as fluorouracil (EFUDEX),
floxuridin (FUDR), tegafur (FENTAL), the more recently
discovered gefitinib (IRESSA), camptothecin DX-8951f (Exa-
tecan mesilate), and homocamptothecin (DN80915), are all
examples of compounds harboring a fluorine atom. This growing
interest is due to the unusual and unique chemical properties
that fluorine can exert on the parent compound without altering
its steric bulk.14,15 Electron withdrawal by fluorine results in a
strong polarization of the C-F bond, and the consequent
pronounced electronic effects can have implications for reactions
in adjacent centers and for drug-target interactions.16 In
addition, the incorporation of fluorine into a drug increases its
lipophilicity, thus enhancing absorption into biological mem-
branes, while the small covalent radius of fluorine can facilitate
the docking of drugs with their receptor(s). The hydrophobic
character of this covalent fluorine-carbon assembly is associated
with a high dipole moment. This effect was recently described
as polar hydrophobicity.17 The increase in lipophilicity obtained
by fluoroalkylation can be applied further to increase the binding
properties of drug molecules to the hydrophobic regions of its
target molecule. A lipophilic group attached to the ligand, which
With respect to the double bond, the spatial relationship
between the two aromatic rings of combretastatin, as well as
that of colchicine and similar drugs, is an important structural
feature that determines their ability to bind to tubulin. A large
* To whom correspondence should be addressed. Phone: +39-06-
91393640. Fax: +39-06-91393638. E-mail: giuseppe.giannini@sigma-tau.it.
a Abbreviations: CA-4, combretastatin A-4; BMEC, bovine microvascular
endothelial cells; MDR, multidrug resistance; DAMA-colchicine, N-deacetyl-
N-(2-mercaptoacetyl)-colchicine; GDP, guanosine-5′-diphosphate; GTP,
guanosine-5′-triphosphate; Ala, alanine; Asn, asparagine; Gln, glutamine;
His, histidine; Val, valine; Thr, threonine; rms, root-mean-square; rmsd,
root-mean-square deviation; H-bond, hydrogen bond; OPLS-AA, optimized
potential for liquid simulations-all atom; TBDMSi, tert-butyl-dimethyl-silyl;
DAST, diethylaminosulfur trifluoride; TBAF, tetra-n-butylammonium fluo-
ride; ADME, absorption, distribution, metabolism and excretion; DMEM,
Dulbecco’s modified eagle medium; EDTA, ethylenediaminetetraacetic acid;
PBS, phosphate-buffered saline; CCD, charge-coupled device.
10.1021/jm701362m CCC: $40.75
2008 American Chemical Society
Published on Web 04/09/2008

Fluorinated Combretastatin Analogues
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 9 2709
further tested for their antiproliferative activity on tumor cell
lines and in the inhibition of tubulin polymerization.
Computational Chemistry. As a preparatory study of
molecular modeling, we chose to evaluate, in silico, the influence
of the fluorinated atom on the cis-stilbene double bond of the
combretastatin scaffold. With this aim in mind, we started from
known information about the colchicine binding site on a tubulin
dimer.28 The crystal structure shows five distinct protein
subunits: two Rꢀ-tubulin heterodimers and the stathmin-like
domain of the regulatory protein, RB3. The colchicine binding
site inhibitor, cocrystallized with each heterodimer, is N-deacetyl-
N-(2-mercaptoacetyl)-colchicine (DAMA-colchicine).
To begin with, the stathmin-like domain and subunits A and
B were removed. Subunits C and D were chosen because the
1-methoxy group of DAMA-colchicine of subunits A and B
presented close contact with the side chain of Ala ꢀ316. The
full Rꢀ-tubulin complex, consisting of GDP, GTP-Mg2+, and
DAMA-colchicine, was used in the study. Bond order and
formal charges were assigned to heteroatom groups. Hydrogen
atoms were added to the entire system and the orientation of
hydroxyl groups, amide groups of Asn and Gln, and the charge
state of His residues were optimized. Subsequently, restrained
minimization of the complex structure was performed until
reaching an rms gradient of 0.1. The original X-ray structure
and the refined complex had a heavy atom rms deviation (rmsd)
of 0.28 Å, thus showing good stability despite an X-ray
resolution of 3.58 Å. Indeed, the internal ligand docking showed
a deviation from the X-ray coordinates of only 0.44 Å.
Figure 2. Fluorinated combretastatins.
The molecular volume and electrostatic properties of the
colchicine site impose severe constraints on the conformation
of colchicine site inhibitors. Thus, we calculated a hydrophobic
and hydrophilic map of the internal tubulin surface and
molecular electrostatic potential surface for compounds 1, 3a,
3b, 7, 8a, 9a, and 11a–16a to determine what would be the
main contribution to binding affinity.
Figure 3. Fluorinated heterocombretastatins.
does not interfere with the active site of the enzyme to be
inhibited, enhances its bioactivity upon binding to hydrophobic
residues.
However, the introduction of fluorine atoms into a molecule
is not always advantageous, as in the case of anthracyclines,
where the introduction of fluorine atoms in different scaffold
positions did not produce analogues in clinical development.18
Prompted by the interesting results obtained with fluorinated
colchicinoids against MDR cell lines,19 the first fluorocombre-
tastatins were synthesized by replacement of the hydroxyl20,21
and/or methoxyl22,23 on the B-ring, as well as by replacement
of the methoxy group(s)24 on the A-ring. There are also
examples where fluorine was introduced into both rings.21
In addition, a fluorine on a stilbene-like double bond could
influence the metabolic oxidation of this kind of molecular
scaffold both on double bond and on phenolic ring (B-ring).25
The introduction of a strong electron-withdrawing group, such
as fluorine, could increase the metabolic survival of potentially
bioactive molecules.
To explore the influence of fluorine substitution in the
combretastatin series, we studied, in silico, and synthesized a
small library of compounds in which one or both of the olefinic
hydrogens were replaced with fluorine; to evaluate the influence
of a different halogen, a monobromo monofluoro analogue was
also synthesized (Figure 2).26
In addition, the same modifications were done on amino and
nitro combretastatins, as well as on a new class of combretasta-
tins recently synthesized in our laboratory,27 in which the B-ring
was replaced with benzofuran and benzothiophen heterocycles
(Figure 3).
The cocrystallized inhibitor, DAMA-colchicine, was removed,
and the accessible space in the active site was partitioned into
three types of regions: hydrophobic (favorable to occupancy
by hydrophobic ligand groups), hydrophilic (favorable to
occupancy by hydrophilic ligand groups), and neither hydro-
phobic nor hydrophilic (of mixed character where at first
approximation any type of group could reside with little effect
on binding affinity). Hydrophobic and hydrophilic areas are
365.1 and 1413.4 Å2, respectively. The middle portion of the
active site is mainly hydrophobic (in magenta), while Figure
4a shows the peripheral section with a preferable hydrophilic
(in light blue) characteristic.
Next, docking studies on the same compounds were carried
out to investigate their binding mode. The trimethoxyphenyl
group of OH and NH2 derivatives move slightly from X-ray
coordinates of the colchicine site (rmsd 1.7 Å), while the NO2
derivative is located in close proximity (rmsd 1.2 Å). These
shifts are necessary to accommodate the B-ring in the correct
position to interact with Thr R179 or Val R181. As expected,
fluorine introduction on the vinyl moiety is well-tolerated,
promoting only little adjustment on the molecules. Figure 4b
shows the docking-derived superposition of compounds 1, 3a,
and 3b (cyan, green, and orange, respectively). Comparing the
hydrophobic and hydrophilic maps of the active site to the ligand
docking solution, we can see that both phenyl groups of
inhibitors are buried in the hydrophobic area. In addition, the
CR of the vinyl bridge is also completely embedded in the
hydrophobic portion in contrast with the Cꢀ. Hence, the fluorine
The fluorinated analogues, as well as reference compounds,
were screened for their antiproliferative activity on bovine
microvascular endothelial cells (BMEC); some of them were

2710 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 9
Alloatti et al.
As a result of this study, we suggest that it might be
interesting to explore how the different contributions (hydro-
phobic and electrostatic) of a fluorine atom on the vinyl double
bond can influence tubulin binding.
Chemistry. Synthesis of Difluorocombretatastin. The di-
fluorinated analogue of the naturally occurring CA-4 (7),
together with its amino derivate (14), via the corresponding nitro
(11), were synthesized using a route previously disclosed by
Shimizu et al. for the preparation of difluorinated stilbenes.29
Addition of the carbenoid obtained in situ from tribromo-
fluoromethane and BuLi in diethyl ether/THF 1:1 at –130 °C
to the TBDMSi-protected isovanillin 22a and to its nitro
analogue 22b gave the halogenated ethanols 23a and 23b,
respectively. A DAST-mediated fluorination in dichloromethane
afforded the difluoroethanes 24a and 24b.
Dehydrobromination-promoted lithium tetramethylpiperidine
in diethyl ether/THF led us to the cis intermediates 25a and
25b, which were coupled under Suzuki conditions with com-
mercially available 3,4,5-trimethoxybenzenboronic acid in re-
fluxing toluene to give, after silyl deprotection or zinc reduction,
final stilbenes 7 and 14 in good overall yields (Scheme 1).
Synthesis of Monofluorocombretastatin. As a first approach
to the monofluorinated analogues, we chose a semisynthetic
route which, after either a direct or bromohydrine-mediated (29;
Scheme 2) bromofluorination, followed by a base-promoted
elimination, would have led us to the desired monofluoro
stilbenes.
As Scheme 2 describes, we used (Z)-4-methoxy-stilbene as
our model compound and were able to obtain the desired
fluorinated stilbenes in acceptable overall yields. Encouraged
by this result, we applied the same protocol to CA-4 but obtained
only low yields of the desired product, the major reaction
observed being the bromination of the extremely activated
A-ring. As a result, we adopted more reliable approaches for
the two different regioisomers.
Figure 4. (a) CA-4 docked atop the colchicine-binding site. Hydro-
phobic (magenta) and hydrophilic (light blue) regions are shown. (b)
Docking-derived superposition of 1 (cyan), 3a (green), and 3b (orange).
Table 1. Electrostatic Potential Values of Reference and Fluorinated
Compounds
electrostatic potential (kcal/mol)
Synthesis of r-Fluorocombretastatin. The monofluorinated
regioisomers of CA-4, 8, and its amino derivative 15, in which
the fluorine atom is on the CR (benzylic position with respect
to the trimethoxyphenyl moiety), were prepared by using a
fluorinated ylide strategy already explored by Burton et al.30
The key step in this approach was the synthesis of the
halogenated styrenes 31a and 31b from the reaction of iso-
vanillin and aldehyde 22b with the proper fluorinated ylide,
generated in situ from triphenylphosphine and tribromofluo-
romethane in dichloromethane.
cmpd
min
max
2-MeO
1
7
8a
9a
-250.8
-232.4
-243.3
-245.9
-302.4
-291.2
-293.3
-293.4
-256.9
-227.5
-235.3
-216.7
143.4
153
-247
-225
147.4
146.4
298.3
308.1
294.3
293.2
122.5
139.7
129
-206
-243
3a
-230
11a
12a
13a
3b
14a
15a
16a
-216
-224
-229
-256.9
-227.5
-235.3
-216.7
We obtained the desired styrenes in moderate yield as a 4:1
mixture of the geometric E- and Z-isomers, which were coupled
with 3,4,5-trimethoxybenzenboronic acid. While for nitro
derivative 12 we had no problem purifying the isomers, a zinc
reduction in acetic acid gave 15 in good yields and isolation of
the E-isomer of 32 needed a further silyl-protection step (33)
to allow good chromatographic separation (Scheme 3).
128.7
position on the vinyl moiety could influence binding affinity in
various ways.
To evaluate the electronic contribution of fluorine atom ligand
substitution, we compared the electrostatic potential surfaces
of different fluoro derivatives with the same substituent on the
B-ring (see Supporting Information). In particular, we investi-
gated the H-bond capability of OH and NH2 as donor groups
and the 2-methoxy group as acceptor. Electrostatic potential
maps were calculated on the minimized docking-derived
conformations. Table 1 reports some representative electrostatic
potential values, where the H-bond donor capability (positive
values) improves with respect to the number of fluorine atoms.
On the contrary, the H-bond acceptor property (negative values)
is negatively influenced by fluorine substitution, depending on
its position and the B-ring substituent.
Synthesis of ꢀ-Fluorocombretastatin. The addition of the
carbenoid, obtained in situ from tetrabromomethane and BuLi
in diethyl ether/THF 1:1 at –130 °C, to TBDMSi-protected
isovanillin 22a gave tribrominated alcohol 34 (Scheme 4), which
after DAST-mediated fluorination and DBU-promoted elimina-
tion was converted to halogenated styrene 36 in good yield.
This intermediate was converted to stilbene 38 following two
alternative routes consisting of the same synthetic steps, with
radical debromination and a Suzuki-coupling, but in different
order.

Fluorinated Combretastatin Analogues
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 9 2711
Scheme 1. Synthesis of the Difluorinated Analogue of CA-4 and its Corresponding Nitro and Aminoa
a Reagents and conditions: (i) CFBr3, BuLi, Et2O/THF, -130 °C, 2 h, 65%; (ii) DAST, CH2Cl2, -78°C, 2 h, 60%; (iii) tetramethylpiperidine, BuLi,THF/
Et2O, -100°C, 4 h, 59% (R ) TBDMSiO), 30% (R ) NO2); (iv) 3,4,5-trimethoxybenzenboronic acid, 2 M Na2CO3, Pd(PPh3)4, refluxing toluene, 16 h,
83%; (v) TBAF, THF, 0 °C-RT, 2 h, 82%; (vi) Zn, AcOH, RT, 3 h, 80%.
Scheme 2. Study of Monofluoruration Using a
(Z)-4-Methoxy-stilbenea
for the synthesis of ꢀ-regioisomer of aminofluorocombretastatin
16, starting from the commercially available 3,4,5-trimethoxy-
benzaldehyde 40 (Scheme 5).
The synthetic approach used for compound 8 allowed us to
obtain the heterocyclic derivatives 18 and 19 (Scheme 6).
The use of a differently substituted boronic acid allowed us
to obtain, directly from intermediate 41, compound 17 (Scheme
7), which could be confronted with the reported analogue
bearing no fluorine on the double bond.21
E/Z-Configuration of the synthesized compounds was assessed
measuring the F-F coupling constants (19F-NMR) for di-
fluorinated compounds and the H-F coupling constants (1H
NMR) for monofluoro derivatives.
For selected fluorinated derivatives only, disodium phosphate
prodrugs were synthesized according to the reported procedure,31
and we report compound 43 as an example (Scheme 8) in the
Experimental Section. These prodrugs are currently evaluated
in in vivo models.
A crystal of 7, suitable for X-ray analysis (Figure 5), was
grown from a mixture of dichloromethane and hexane (1/2, v/v).
The X-ray crystal structure revealed that the planes of the two
phenyl rings were inclined toward each other, according to the
nonfluorinated analogue structure.32 Replacement of two olefinic
hydrogens with fluorine did not modify the natural conformation
of the CA-4.
Biological Results. The reference compound, as well as all
synthesized compounds, was preliminarily tested for activity
on bovine microvascular endothelial cells (BMEC).
The results in Table 2 show that all trans-stereoisomers were
at least three times less-active than their cis-forms on BMEC.
The cis-isomers were then evaluated for their interference with
tubulin polymerization and tested on NCI-H460 non-small cell
lung carcinoma and on HT29 colon carcinoma cells.
a Reagents and conditions: (i) NBS, HF ·Pyr, CH2Cl2, Et2O, -20 °C-RT,
1 h, 60%; (ii) DBU, DMSO, 80 °C, 3 h, 85%; (iii) NBS, CH2Cl2, -15°C,
1 h then 5% NaHCO3, 40% (incomplete conversion); (iv) DAST, CH2Cl2,
-78 °C-RT, 2 h, 78%.
The two routes proved to be almost comparable in efficiency,
but the Pd-coupling on compound 36 showed a higher selectivity
for cis-stilbene 37, whose geometry was assessed by NOE
experiments, compared to its trans-isomer (cis/trans ratio is 3:1).
As a consequence, this synthetic route is slightly preferred. By
deprotection of the silyl group of intermediate 37, it is possible
to obtain the R-bromo-ꢀ-fluorocombretastatin 10, which was
studied to explore the influence of a larger halogen on the double
bond.
Removal of the TBDMSi-group with TBAF in THF gave
the desired product 9 in good yield (Scheme 4).
Synthesis of ꢀ-Fluoroaminocombretastatin. An approach
analogous to that used for the R-fluorocombretastatin was used
CA-4 (1) inhibited tubulin aggregation with an IC50 of 4.9
µM and has potent antiproliferative effect on endothelial cells

2712 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 9
Scheme 3. Synthesis of CR-Fluoro Analoguesa
Alloatti et al.
a Reagents and conditions: (i) CFBr3, Ph3P, CH2Cl2, RT, 2 h, 31%, E/Z ) 4:1; (ii) 3,4,5-trimethoxybenzenboronic acid, Pd(Ph3P)4, 2 M Na2CO3, toluene,
reflux, 44%, E/Z ) 1:4; (iii) TBDMSiCl, imidazole, CH2Cl2, RT, 60%, chromatographic separation, E/Z ) 1:4; (iv) 1 M TBAF, THF, 0 °C-RT, 2 h, 90%;
(v) Zn, AcOH, RT, 16 h, 50%.
Scheme 4. Synthesis of Cꢀ-Fluoro Analoguesa
a Reagents and conditions: (i) CFBr4, BuLi, THF, -78 °C, 2 h, 40%; (ii) DAST, CH2Cl2, -78 °C to RT, 2 h, 75%; (iii) DBU, RT, 4 h, 83%; (iv)
3,4,5-trimethoxybenzenboronic acid, 2 M Na2CO3, Pd(PPh3)4, refluxing toluene, 16 h, 31%; (v) TBAF, THF, 0 °C-RT, 2 h, 89%; (vi) AIBN, BuSn3H, 80%;
(vii) 3,4,5-trimethoxybenzenboronic acid, 2 M Na2CO3, Pd(PPh3)4, refluxing toluene, 16 h, 20%.
(IC50 ) 2.6 nM), according to literature data.27 Fluorine
substitution into the double bond site in CR position (8a) led to
a 3-fold increment of tubulin disruption compared to that
produced by one substitution in Cꢀ (9a) or in both positions
(7), while this analogue (8a) produced a modest improvement
in the inhibition of endothelial proliferation compared to CA-4
or its fluorine parent compounds. Moreover, it is interesting to
note that the substitution of a fluorine with a bromine atom (10)
resulted in a complete loss of activity compared to difluoro
analogue 7.
In comparison with NCI-H460 cells, HT29 colon carcinoma
cells exhibited a certain degree of resistance to most of the
fluorinated compounds with the 3-hydroxyl substitution on the
B-ring, except for the 9a derivative.
Replacement of the OH group with an NO2 on the 3-position
of the B-ring generated compounds such as 3a, 11a, and 13a,
which displayed similar potencies on tubulin polymerization
compared to previous molecules, but a dramatic loss of
antiproliferative activity in endothelial and tumor cells was
observed. In this series, it was surprising that the presence of a

Fluorinated Combretastatin Analogues
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 9 2713
Scheme 5. Synthesis of Cꢀ-Fluoroaminocombretastatina
a Reagents and conditions: (i) CFBr3, Ph3P, CH2Cl2, RT, 2 h, 31%, E/Z ) 1:1; (ii) 4-methoxy-3-nitrophenylboronic acid, Pd(Ph3P)4, 2 M Na2CO3,
toluene, reflux, 56%, E/Z ) 1:1.1; (iii) Zn, AcOH, RT, 2 h, 81% (starting from 13a).
Scheme 6. Synthesis of CR-Fluoro Heterocombretastatina
Scheme 7. Synthesis of Cꢀ-Fluoro 3′-Deoxy-3′-fluoro CA-4a
a Reagents and conditions: (i) 4-methoxy-3-fluorophenylboronic acid,
Pd(Ph3P)4, 2 M Na2CO3, toluene, reflux, 62%, E/Z ) 1.8:1. (*) The same
synthetic sequence led to the Z-isomer.
a Reagents and conditions: (i) CFBr3, Ph3P, CH2Cl2, RT, 2 h, 31%, E/Z
) 4:1; (ii) 3,4,5-trimethoxybenzenboronic acid, Pd(Ph3P)4, 2 M Na2CO3,
toluene, reflux, 44%, E/Z ) 1:4.
Scheme 8. Synthesis of Disodium Phosphate Prodrug of the
Difluorinated Analogue of CA-4a
fluorine atom in the CR position of derivative 12a produced an
enhanced effect on tubulin aggregation, which correlated with
a 4-fold increment in potency against endothelial and tumor
cells compared to its reference 3a.
Next, we examined the reference compound 3b, which
presented an NH2 group on the 3-position of the B-ring and its
fluorinated derivatives, 14a, 15a, and 16a. Once again, fluorina-
tion on the CR into the olefin site (15a) incremented the
inhibition of tubulin polymerization three times, in comparison
with the compounds produced by the other fluorinated amino
analogues (14a and 16a) but this constraint produced weak
improvement of antiproliferative effect on endothelial and cancer
cell lines, as compared to reference compound 3b. On the
contrary, although fluorination on the amino derivatives 14a
and 16a resulted in a 3- to 5-fold loss in tubulin inhibition, the
antiproliferative potency against endothelial and tumor cells was
similar to that of 3b. These last results confirm and extend
previous observations reported in literature, in which a nonlinear
relationship between proliferation inhibition and the effect on
tubulin polymerization by compounds with the amino group in
position 3′ were observed.9
a Reagents and conditions: (i) (BnO)2P(O)H, DIEA, CCl4, CH3CN, -25
°C, 88%; (ii) TMSCl, NaBr, CH3CN, 2 h, RT; (iii) NaOH, H2O, 15 h, 92%
(two steps).
Among the derivatives selected for further investigation, we
chose the fluorinated nitro-analogue 12a, which showed unex-
pected tubulin inhibitory activity and antiproliferative properties.
To evaluate the biological properties of this compound on
BMEC cells, it was studied for its activity on the microtubule
network and for its effect on cell cycle of BMEC, in comparison
with CA-4. Cells were treated for 24 h at a concentration
corresponding to IC80 to maximize the biological effects of
compounds in asynchronously growing BMEC cells. Tubulin
immunostaining showed that the 12a molecule strongly affected
the cell shape and microtubule arrangement in a manner
comparable to CA-4 (Figure 6).
According to the effect on the microtubule cytoskeleton, flow
cytometric data showed that 12a had the same pattern as the
reference compound, leading to an effective arrest of the cell
cycle in the G2/M phase and induction, to a significant extent,
of apoptosis (Table 3).
Figure 5. X-ray crystal structure of 7.
addition, this strategy was justified by the preparatory data of
the molecular modeling studies which suggested: a different
environment for the two protons on double bond, more
hydrophobic for H on CR and more hydrophilic for H on Cꢀ,
as well as the contribution of the fluorine in CR position of the
double bond, to H-bond acceptor and donor properties of the
molecules.
Usually, to introduce fluorine atoms on the stilbene double
bond in a regioselective manner, a semisynthetic approach is
possible. However, this synthetic route is not feasible on
combretastatin-like scaffolds, probably because of unfavorable
electronic effects. Our studies provided a versatile synthetic
approach, enabling us to obtain all the possible mono and
difluoroderivatives of both the natural and the nitro/amino series.
In addition, monofluorinated analogues of a new class of
heterocombretastatins were obtainable via the same approach.
Conclusions
The aims of this work were to devise a synthetic strategy to
obtain mono and difluoro- derivatives of CA-4 and to verify
the influence of these substitutions on antitumor activity. In

2714 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 9
Alloatti et al.
Table 2. IC50 on Tubulin Polymerization and Cytotoxic Effect on BMEC, H460, and HT29 Cell Linesg
code number
double bond geometrya
X
R1, R2
inhibition of tubulin aggregationb
BMECc
H-460d
HT29e
1
7
8a
8b
9a
9b
10
3a
11a
11b
12a
12b
13a
13b
3b
14a
14b
15a
15b
16a
16b
17a
17b
CA-4
A
A
A
B
A
B
A
A
A
B
A
B
A
B
A
A
B
A
B
A
B
A
B
OH
OH
OH
OH
OH
OH
OH
NO2
NO2
NO2
NO2
NO2
NO2
NO2
NH2
NH2
NH2
NH2
NH2
NH2
NH2
F
H, H
F, F
4.9 ( 0.2
7.3 ( 0.3
2.5 ( 0.1
2.6 ( 0.21
4.6 ( 0.5
1.7 ( 0.6
13 ( 2.7
4.7 ( 0.1
26.6 ( 2.4
>200
135 ( 1.3
749 ( 75
2300 ( 53
38.4 ( 2.5
2800 ( 72
329 ( 2.5
>3000
3.1 ( 0.4
3.6 ( 0.2
204 ( 5.4
2.0 ( 0.8
86 ( 1.3
2.1 ( 0.2
111 ( 15
23 ( 1.3f
610 ( 15
1.24 ( 0.02
3.0 ( 0.05
1.1 ( 0.02
118 ( 0.03
110 ( 0.01
100 ( 11
ST2303
ST2966
ST2967
ST3036
ST3037
ST3000
ST3098
ST2998
ST2999
ST2969
ST2970
ST3038
ST3039
ST3100
ST2996
ST2997
ST2971
ST2972
ST3040
ST3041
ST3001
ST3002
F, H
F, H
H, F
H, F
Br, F
H, H
F, F
8.8 ( 0.1
4 ( 0.2
28.9 ( 3.7
6.8 ( 0.1
8.2 ( 0.2
128 ( 0.6
1166 ( 75
113 ( 1
1307 ( 26
F, F
F, H
F, H
H, F
H, F
H, H
F, F
2.3 ( 0.3
42.6 ( 6
32 ( 6.5
202 ( 19
10.9 ( 1.5
379 ( 0.2
3.5 ( 0.2
17.5 ( 1.4
4.9 ( 0.1
7.3 ( 0.9
5 ( 0.5
4.5 ( 0.1
F, F
F, H
F, H
H, F
H, F
H, F
H, F
5.5 ( 0.1
2.4 ( 0.2
2.9 ( 0.02
5 ( 0.3
13.7 ( 0.6
11.5 ( 0.9
F
a For general structure, see Figure 2. b A: Z-stilbene scaffold; B: E-stilbene scaffold. c IC50 [IC50 ( SD; µM]. d Bovine microvascular endothelium [IC50
( SD; nM]. e Lung carcinoma [IC50 ( SD; nM]. f Colon carcinoma [IC50 ( SD; nM]. Data represent the means of at least three experiments. g For the
dehalogenated compound, ref 21 reports IC50 ) 10 nM and IC50 ) 4 nM on K562 and HUVEC cell lines, respectively.
vitro properties overlapping those of the parent compounds.
Nevertheless, small differences can be observed, which always
remain in the low nanomolar range. The introduction of a
fluorine atom, both in CR and in Cꢀ, (7), or only in Cꢀ (9a),
results in a slight reduction of tubulin aggregation. On the
contrary, when the fluorine atom was introduced only on CR
(8a) an improvement of activity was observed when compared
to the parent compound (1). And the same trend, more potent
with fluorine on CR and less potent with fluorine on Cꢀ or
CR-Cꢀ, was observed for nitro (11a, 12a, 13a) and amino (14a,
15a, 16a) derivatives.
The same effect was observed as cytotoxic activity, both on
the endothelial (BMEC) and tumor cell lines (H-460).
On the HT29 tumor cell line, which is naturally less
responsive to combretastatin (1) and to nitrocombretastatin (3a)
but sensitive to aminocombretastatin (3b), introduction of
fluorine showed an equivalent trend (more potent with fluorine
on CR) limited to nitro-(11a, 12a, 13a) and amino-(14a, 15a,
16a) derivatives. An exception to this trend was observed for
the natural scaffold where the most efficacious substitution was
in Cꢀ (9a).
The biological results seem to be in agreement with the
Figure 6. Immunostaining experiments on 1 and 12a. R-Tubulin
(green) and nuclear staining (blue) of cells after 24 h of drug treatment.
Magnification: 100×.
previous observation, in terms of different biological environ-
ments regarding the interaction of CR and Cꢀ (see Figure 4a).
The substitution of fluorine with bromine on CR (compound
10) resulted in a 40-fold reduction in antiproliferative activity
on the BMEC cell line with respect to the parent compounds 7
and 9a. This data confirm the negative influence of steric factors
on the double bond of the stilbene scaffold.
A reduction in antiproliferative activity against BMEC was
observed when the phenolic OH of 9a/b was substituted with a
second fluorine atom to give compound 17a/b.
The trans-stilbene scaffold is known to be much less potent,
compared to the corresponding cis. The same trend was observed
for our fluorinated derivatives, with some exceptions. Indeed,
derivatives 8b, 9b, and to a lesser extent 14b,15b, and 16b,
still possessed a nanomolar activity although endowed with a
minor antiproliferative potency with respect to 1. To explain
these biological data, one hypothesis considers the influence of
Table 3. Flow Cytometric Analysis of Apoptosis and Cell Cycle of
NCI-H460 Cellsa
G0/1
%
S%
G2/M%
apoptosis (subG1)%
untreated
cmpd 1 (CA-4)
cmpd 12a
44.2
4.9
3.2
39.9
8.6
9.8
16.0
86.4
87.1
0.8
20.5
18.4
a NCI-H460 cells were exposed for 24 h to the compounds (IC80 dose),
stained with propidium iodide, and processed on a FACScan flow cytometer.
Cell cycle distribution and apoptosis were assessed as described in
Experimental Section.
Taken together, this study demonstrates that it is possible to
introduce one or two fluorine atoms on the stilbene double bond
of combretastatins without altering their biological properties.
In fact, the monofluoro- and difluoro- derivatives showed in

Fluorinated Combretastatin Analogues
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 9 2715
silica gel F-254 plates. Flash chromatography was done using Merck
silica gel 60 (0.063–0.200 mm). Solvents were dried according to
standard procedures and reactions requiring anhydrous conditions
were performed under argon. Solutions containing the final products
were dried with Na2SO4, filtered, and concentrated under reduced
pressure using a rotatory evaporator. All the final products
undergoing biological testing were analyzed in a Jasco HPLC
system consisting of two PU-2080 pumps and an MD-2010 detector,
with purity above 99%. Microanalysis of all new synthesized
compounds was within (0.4% of calculated values.
tert-Butyl-dimethyl-silyl Isovanillin (22a). To a solution of
6.09 g (40 mmol) of isovanillin in 50 mL of CH2Cl2 were added
6.64 g of TBDMSiCl (44 mmol, 1.1 equiv) and 2.95 g (44 mmol,
1.1 equiv) of imidazole. The solution was stirred at room temper-
ature for three h and then washed with 0.5 M HCl. The crude
product was purified on a silica gel column using hexane/ethyl
acetate 9:1, to give 9 g (33 mmol, 83%) of a colorless oil. Rf )
0.27 (hex./ethyl acetate 95:5). MS (IS): [MH]+ ) 267.2 [M + Na]+
) 289.2 (main peak). 1H NMR (200 MHz, CDCl3, δ): 0.20 (s, 6H,
2xCH3); 1.03 (s, 9H, tBu); 3.92 (s, 3H, OCH3); 6.98 (d, J ) 8.4
Hz, 1H, CH); 7.39 (d, J ) 2.2 Hz, 1H, CH); 7.50 (dd, J ) 8.4 Hz,
J ) 2.2 Hz, 1H, CH); 9.84 (s, 1H, CHO). Anal. (C14H22O3Si) C,
H.
2,2-Dibromo-2-fluoro-1-(4-methoxy-3-tert-butyl-dimethyl-
silyloxyphenyl) Ethanol (23a). A mixture of 2.66 g (10 mmol) of
TBDMS-isovanillin 22a and 2.98 g (11 mmol, 1.1 equiv) of CFBr3
in 80 mL of Et2O/THF (1:1) was brought to T ) -130 °C; 4.4 mL
(11mmol, 1.1 equiv) of a 2.5 M BuLi solution in hexane was added
to the mixture in ten min. After two h at T ) -70 °C, it was
necessary to add 1.3 mL of BuLi solution and 0.3 mL of CFBr3 to
drive the reaction to completion. The reaction was quenched with
60 mL of NH4Cl saturated solution and diluted with 20 mL of
diethyl ether. The aqueous phase was back-extracted with 2 × 20
mL of diethyl ether, and the organic fractions were collected, dried
over anhydrous sodium sulfate, and then purified on a silica gel
column using hexane/ethyl acetate 95:5 to give 3.1 g (6.8 mmol,
68%) of a waxy solid. Rf ) 0.5 (hex/AcOEt 85:15). MS (IS): [M
+ Na]+ ) 479.1; 481.1; 483.1 (1:2:1). 1H NMR (200 MHz, CDCl3,
δ): 0.18 (s, 6H, 2 × CH3); 1.01 (s, 9H, tBu); 3.84 (s, 3H, OCH3);
5.03 (d, J ) 9.5 Hz, 1H, CH,); 6.87 (d, J ) 8.4 Hz, 1H, CHar);
7.09 (m, 2H, 2 × CH). Anal. (C15H23Br2FO3Si) C, H.
fluorine on trans-double bond stability and on the possible
facilitation of reduction of the nitro group with respect to the
corresponding amino derivative. Thus, these unpredictable
results encourage further investigation.
One of the synthetic procedures exploited in this paper was
used to synthesize the fluorinated (on CR) analogues (18 and
19) of a novel class of heterocombretastatins previously
developed in our laboratories.27 Preliminary unpublished results
seem to confirm, even for these compounds, that the introduction
of fluorine does not significantly affect biological activity (see
Supporting Information).
The biological data collected up to now show that the
introduction of one/two fluorines on the double bond of
combretastatin, unlike the introduction of bulkier groups, does
not appreciably influence the biological response of most of
the synthesized compounds. Therefore, according to what is
already described for other fluorinated drugs,33 it may be
reasonable to observe differences in pharmacokinetic and
metabolic properties with respect to the nonfluorinated ana-
logues. An ADME (absorption, distribution, metabolism, and
excretion) analysis is currently under investigation and will be
reported in due course.
Experimental Section
Computational Chemistry. Compounds 1, 3a, 3b, 7, 8a, 9a,
and 11a-16a, structural analogues of CA-4, were selected for this
study. The X-ray structure of the Rꢀ tubulin-DAMA-colchicine
complex (pdb code 1SA0, chains C and D)28 was used; said
complex was subjected to the standard protein preparation procedure
implemented in Maestro.34 The hydrophobic and hydrophilic
surfaces of the colchicine binding site were calculated with Maestro
hydrophobic/philic facility. The active-site mapping procedure
operates in a manner analogous to Goodford’s GRID algorithm.35
“Hydrophilic” and “hydrophobic” regions are defined in a way that
considers both spatial proximity to the receptor and suitability for
occupancy by solvent–water. The first step in the computing grid
is to define a rectilinear box that contains an active site and to define
grid points with a typical grid spacing of 1 Å within the box. Next,
van der Waals energies and x, y, and z electrostatic field components
are computed at each of the grid points. Receptor atom partial
charges and van der Waals parameters are taken from the OPLS-
AA force field.36 A probe is represented by a van der Waals sphere
of radius 1.5 Å and well depth of 0.2 kcal/mol and has a point
dipole moment of 2.3 debye.37,38 The 3D structure of the ligands
was energy minimized using Macromodel (500-step truncated
Newton conjugated gradient with convergence threshold 0.01).39
Semiempirical PM3 charges and electrostatic potential maps were
calculated with the Spartan ’06 software.40 As reported in litera-
ture,41 we correlated the hydrogen bond capability of OH, NH2,
and the 2-methoxy groups to the maximum and minimum values,
respectively, of electrostatic potential.
1,1-Dibromo-1,2-difluoro-2-(4-methoxy-3-tert-butyl-dimethyl-
silyloxyphenyl)ethane (24a). (Diethylamino)sulfur trifluoride at 1.5
mL (11.2 mmol; 1.8 equiv) in 10 mL CH2Cl2 was added to a
solution of 2.84 g (6.2 mmol) of alcohol 23a in 14 mL of CH2Cl2
at -78 °C. The reaction mixture was allowed to warm up to 0 °C
over a period of two h, quenched with 25 mL of NaHCO3 saturated
solution and diluted with 20 mL of diethyl ether. The organic phase
was dried over anhydrous sodium sulfate and purified on preparative
TLC using hexane/ethyl Acetate 98:2 to give 1.8 g (4 mmol; 64.5%)
of a yellow oil. Rf ) 0.43 (hex./AcOEt 97:3). MS (IS): [M + Na]+
) 483.1; 485.1; 487.1 (1:2:1). 1H NMR (200 MHz, CDCl3, δ):
0.18 (s, 6H, 2 × CH3); 1.02 (s, 9H, tBu); 3.86 (s, 3H, OCH3); 5.59
(dd, J ) 43.6 Hz, J ) 11.0 Hz, 1H, CH,); 6.89 (d, J ) 8.4 Hz, 1H,
CHar); 7.06 (s, 1H, CH); 7.10 (d, J ) 8.4 Hz, 1H, CH). Anal.
(C15H22Br2F2O2Si) C, H.
(E)-1-Bromo-1,2-difluoro-2-(4-methoxy-3-tert-butyl-dimethyl-
silyloxyphenyl)ethene (25a). Step 1. Preparation of the tetrameth-
ylpiperidide solution. Dissolved in 4 mL of anhydrous THF were
1.9 mL (11.7 mmol; 3 equiv) of 2,2,6,6-tetramethylpiperidine; the
solution was cooled to -80 °C and then 3.9 mL (9.8 mmol; 2.5
equiv) of a 2.5 M solution BuLi in hexane were added. The mixture
was stirred for 2 h at 0 °C. Step 2. Dehydrobromination. A solution
of 1.8 g (3.9 mmol) of compound 24a in 5 mL of anhydrous THF
was added to the tetramethyl piperidide solution previously cooled
down to -100 °C. After 1 h, the reaction was washed with 10 mL
of HCl (0.1 N), and the aqueous phase was back-extracted with 2
× 10 mL Et2O. The organic extracts were collected and dried over
anhydrous sodium sulfate and then purified on preparative silica
plates with n-hexane/ethyl acetate 97:3 to give 857 mg (2.3 mmol;
59%) of product as a pale yellow oil. According to the final product
Docking was carried out with Glide software.42 Docking
procedure was first constituted by receptor grid generation: two
cubic boxes centered on the cocrystallized DAMA-colchicine were
created with a range of 24 and 10 Å, respectively, on each side.
Ligands were docked flexibly using default scaling for van der
Waals radii and setting multiple solutions for each molecule.43
Postdocking minimization was performed using MM-GBSA ap-
proach,44 implemented in Prime software,42 with the rigid receptor
approximation. OPLS-AA force field was used in both calculations.
Chemistry. General Remarks: 1H-, 13C-, and 19F-NMR spectra
were recorded, unless otherwise indicated, in CDCl3 solution at
200- or 300-MHz on a Varian Gemini-200 or a Varian Gemini-
300 instrument. Chemical shift values are given in ppm and
coupling constants in Hz. Melting points were determined on a
Mettler DSC-30 and are uncorrected. Electro-Spray mass spectra
were recorded on a Waters Micromass ZQ-2000 instrument. Thin
layer chromatography (TLC) was carried out using Merck precoated

2716 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 9
Alloatti et al.
7 configuration (assigned by 19F-NMR), the stereochemistry of the
two fluorines on the double bond was cis. Rf ) 0.8 in hex/acetone
8:2. MS (IS): [M + Na]+ ) 401.4; 403.4 (1:1). 1H NMR (200
MHz, CDCl3, δ): 0.19 (s, 6H, 2 × CH3); 1.02 (s, 9H, tBu); 3.86 (s,
3H, OCH3); 6.88 (d, J ) 8.4 Hz, 1H, CH); 7.16 (d, J ) 2.2 Hz,
1H, CH); 7.23 (dd, J ) 8.4 Hz, J ) 2.2 Hz, 1H, CH). Anal.
(C15H21BrF2O2Si) C, H.
hexane/ethyl acetate, gave 57 mg (0.15 mmol; 48%) of Z-isomer
as a yellow oil (11a). Rf ) 0.17 in hex/acetone 8:2. MS (IS): [M
+ H]+ ) 382.4. 1H NMR (200 MHz, CDCl3, δ): 3.76 (s, 6H, 2 ×
OCH3), 3.90 (s, 3H, OCH3), 3.98 (s, 3H, OCH3), 6.60 (s, 2H, 2 ×
CHar), 7.02 (d, J ) 8.8 Hz, 1H, CHar), 7.39 (dd, J ) 8.8 Hz, J )
2.2 Hz, 1H, CHar), 7.90 (s, J ) 2.2 Hz, 1H, CHar). 19F NMR (282
MHz, CDCl3, δ): -123.7 (d, JFF ) 12.2 Hz); -131.7 (d, JFF )
12.2 Hz). A small amount of the E-isomer (11b) was isolated: 1H
NMR (200 MHz, CDCl3, δ): 3.93 (s, 9H, 3 × OCH3); 4.04 (s, 3H,
OCH3); 7.01 (s, 2H, 2 × CH); 7.20 (d, J ) 8.8 Hz, 1H, CH); 7.93
(dd, J ) 8.8 Hz, J ) 2.2 Hz, 1H, CH); 8.26 (d, J ) 2.2 Hz, 1H,
CH). 19F NMR (282 MHz, CDCl3, δ): -150.1 (d, JFF ) 122.1
Hz); -153.2 (d, JFF ) 122.1 Hz). Anal. (C18H17F2NO6) C, H, N.
(Z)-1,2-Difluoro-1-(3,4,5-trimethoxyphenyl)-2-(3-amino-4-
methoxyphenyl)ethene (14a). To a solution of 40 mg (0.105 mmol)
of nitro-stilbene 11a in AcOH (5 mL), 75 mg (1.15 mmol; 11 equiv)
of zinc powder was added; the mixture was stirred at room
temperature for 1.5 h. The reaction mixture was filtered over celite
and the filtrate evaporated to dryness. The crude product was
purified by chromatography on silica gel with CH2Cl2, then on
preparative HPLC to give 32 mg (0.091 mmol; 87%) of a waxy
solid (14a). Rf ) 0.31 in CH2Cl2. MS (IS): [M + H]+ ) 352.3. 1H
NMR (300 MHz, CDCl3, δ): 3.71 (s, 6H, 2 × OCH3), 3.86 (s, 3H,
OCH3), 3.87 (s, 3H, OCH3), 6.58 (s, 2H, 2 × CHar), 6.73 (d, 1H,
CHar), 6.89 (m, 2H, 2 × CHar). 19F NMR (282 MHz, CDCl3, δ):
-125.8 (d, JFF ) 15.3 Hz), -131.4 (d, JFF ) 15.3 Hz). Anal.
(C18H19F2NO4) C, H, N.
(Z)-1,2-Difluoro-1-(3,4,5-trimethoxyphenyl)-2-(4-methoxy-3-
tert-butyl-dimethyl-silyloxyphenyl)ethene (26). A mixture of 750
mg (1.98 mmol; 1 equiv) of stirene 25a, 1.260 g (5.94 mmol; 3
equiv) of 3,4,5-trimethoxyphenylboronic acid, 4 mL of Na2CO3 2
M aqueous solution and 104 mg (0.09 mmol; 0.05 equiv) of
Pd(Ph3P)4 in 20 mL of toluene was refluxed overnight. The solution
was then cooled down to room temperature, dried over anhydrous
sodium sulfate, and the crude mixture was passed through a short
silica gel column to remove catalyst. The crude product was purified
by chromatography on silica gel plates with hexane/acetone 8:2 to
give 740 mg (1.6 mmol; 81%) of a colorless oil. Rf ) 0.36 in hex/
acetone 8:2. MS (IS): [M + NH4]+ ) 484.1; [2M + NH4]+
)
1
950.1. H NMR (200 MHz, CDCl3, δ): 0.08 (s, 6H, 2 × CH3);
0.94 (s, 9H, tBu); 3.71 (s, 6H, 2 × OCH3); 3.82 (s, 3H, OCH3);
3.86 (s, 3H, OCH3); 6.57 (s, 2H, 2 × CH); 6.81 (d, J ) 8.4 Hz,
1H, CH); 6.84 (q, J ) 2.2 Hz, J ) 0.8 Hz, 1H, CH); 7.00 (dq, J )
8.4 Hz, J ) 2.2 Hz, J ) 0.8 Hz, 1H, CH). Anal. (C24H32F2O5Si)
C, H.
(Z)-1,2-Difluoro-1-(3,4,5) trimethoxyphenyl)-2-(3-hydroxy-4-
methoxyphenyl)ethene (7). A 1 M solution of tetrabutylammonium
fluoride in THF (9.4 mmol; 2 equiv) was dripped, at 0 °C and under
inert atmosphere, into a solution of 2.2 g (4.7 mmol) of stilbene
26 in 10 mL of anhydrous THF (stored on molecular sieves). The
reaction mixture was allowed to warm up to room temperature and
the reaction completed after four h. The mixture was poured onto
ice, the aqueous phase extracted with Et2O (3 × 20 mL); the organic
extracts were collected and dried over anhydrous Na2SO4. The crude
mixture was purified by chromatography on silica gel with
n-hexane/acetone 8:2 to give 1.361 g (3.9 mmol; 83%) of a white
(E)-1,2-Difluoro-1-(3,4,5-trimethoxyphenyl)-2-(3-amino-4-
methoxyphenyl)ethene (14b). The same protocol applied to the
E-isomer 11b gave, with comparable yield, the E-isomer of the
final aniline (14b): [M + H]+ ) 352.3. 1H NMR (200 MHz, CDCl3,
δ): 3.91 (s, 6H, 2 × OCH3); 3.92 (s, 3H, OCH3); 3.94 (s, 3H,
OCH3); 6.90 (d, J ) 8.1 Hz, 1H, CH); 6.99 (s, 2H, 2 × CH); 7.25
(dd, J ) 8.1 Hz, J ) 2.2 Hz, 1H, CH); 7.28 (d, J ) 2.2 Hz, 1H,
CH). 19F NMR (282 MHz, CDCl3, δ): -150.6 (d, JFF ) 122.1
Hz); -152.6 (d, JFF ) 122.1 Hz). Anal. (C18H19F2NO4) C, H, N.
1-Bromo-1-(4-methoxyphenyl)-2-fluoro-2-phenyl-ethane (28).
To a solution of 231 mg (1.3 mmol; 1.3 equiv) of N-bromo-
succinimide in 4 mL of dry CH2Cl2 and 2 mL of dry Et2O,
previously cooled down to –20 °C, were added 1.25 mL of HF ·Pyr.
After 10 min at this temperature, 210 mg of (Z)-4-methoxy-stilbene
(27), solubilized in 1 mL of dry Et2O, were added to the mixture.
The reaction was run to completion over one h and the mixture
diluted with 20 mL Et2O and poured into water. The organic phase
was neutralized with 5% NaHCO3 (10 mL) and dried over
anhydrous sodium sulfate. The crude product was purified by silica
gel chromatography using hexane/EtOAc 95:5 to give 196 mg (0.63
mmol; 63%) of a yellow solid. Rf ) 0.55 (hexane/EtOAc 85:15).
1
solid; mp ) 134–137 °C. MS (IS): [M + H]+ ) 353.0. H NMR
(200 MHz, CDCl3, δ): 3.71 (s, 6H, 2 × OCH3); 3.87 (s, 3H, OCH3);
3.91 (s, 3H, OCH3); 5.64 (s, 1H, OH); 6.58 (s, 2H, 2 × CH); 6.80
(d, J ) 8.4 Hz, 1H, CH); 6.91 (dq, J ) 8.4 Hz, J ) 2.2 Hz, 1.1
Hz, 1H, CH); 7.00 (d, J ) 2.2 Hz, 1H, CH). 19F NMR (282 MHz,
CDCl3, δ): -126.2 (d, JFF ) 14.8 Hz), -130.3 (d, JFF ) 14.8
Hz). Anal. (C18H18F2O5) C, H.
2,2-Dibromo-2-fluoro-1-(3-nitro-4-methoxy-phenyl)ethanol
(23b). The same procedure used for compound 23a, after purifica-
tion on a silica gel column using hexane/ethyl acetate 95:5, gave
1.146 g (3.1 mmol, 57.4%) of a yellow oil. Rf ) 0.53 (hex/AcOEt
1
6:4). MS (IS): [M - 1]- ) 371.8. H NMR (200 MHz, CDCl3,
1
[M + H]+ ) 309.2/311.2. H NMR (200 MHz, CDCl3, δ): 3.85
δ): 3.2 (bs, 1H, OH); 4.0 (s, 3H, OCH3); 5.15 (dd, J ) 2.9 Hz, J
) 8.0 Hz, 1H, CH); 7.13 (d, J ) 8.8 Hz, 1H, CHar); 7.76 (dd, J )
8.8 Hz, J ) 2.2 Hz, 1H, CHar); 8.10 (d, J ) 2.2 Hz, 1H, CHar).
Anal. (C9H8Br2FNO4) C, H, N.
(s, 3H, OCH3), 5.15 (dd, J ) 7.5 Hz, J ) 13.9 Hz, 1H, CH), 5.85
(dd, J ) 7.5 Hz, J ) 45.8 Hz, 1H, CH), 6.90 (d, J ) 8.7 Hz, 2H,
2 × CHar), 7.25 (d, J ) 8.7 Hz, 2H, 2 × CHar), 7.35–7.50 (m,
5H, 5 × CHar).
1,1-Dibromo-1,2-difluoro-2-(3-nitro-4-methoxy-phenyl) Ethane
(24b). The same procedure used for compound 24a, after purifica-
tion on a silica gel column using hexane/ethyl acetate 7:3, gave
960 mg (2.6 mmol; 84%) of a yellow oil. Rf ) 0.493 (hex/AcOEt
(E)-1-(4-Methoxyphenyl)-2-fluoro-2-phenyl-ethene (30). A
solution of 25 mg (0.08 mmol) of compound 28 and 14 mL (0.09
mmol; 1.1 equiv) of DBU in 2 mL DMSO was stirred at 80 °C for
3 h. The reaction mixture was poured into water (10 mL), extracted
with EtOAc (3 × 15 mL), and dried over anhydrous sodium sulfate.
The crude mixture was purified by silica gel chromatography using
hexane/EtOAc 97:3 to give 16 mg (0.07 mmol; 88%) of a pale
yellow oil. Rf ) 0.67 (hexane/ EtOAc 9:1). [M + H]+ ) 229.2.
1H NMR (200 MHz, CDCl3, δ): 3.70 (s, 3H, OCH3), 6.25 (d, J )
21.6 Hz, 1H, CH), 6.70 (d, J ) 9.2 Hz, 2H, 2 × CHar), 7.00–7.15
(m, 5H, 5 × CHar), 7.25 (d, J ) 9.2 Hz, 2H, 2 × CHar).
5-[(E/Z)-2-Bromo-2-fluoro-vinyl]-2-methoxy-phenol (31a). A
solution of CFBr3 (241 µL; 2.4 mmol; 1.2 equiv) and Ph3P (1.153
g; 4.4 mmol; 2.2 equiv) in 25 mL of CH2Cl2 was stirred for 0.5 h
in an ice bath; isovanillin (304.4 mg; 2 mmol) was added and the
reaction mixture was then kept in an ice bath for 30 min and at
room temperature for 2 h until completion. After dilution with
CH2Cl2 (30 mL) and washing with water (2 × 20 mL), the organic
1
7:3). H NMR (200 MHz, CDCl3, δ): 4.03 (s, 3H, OCH3); 5.71
(dd, J ) 43.2 Hz, J ) 9.5 Hz, 1H, CH,); 7.18 (d, J ) 8.8 Hz, 1H,
CHar); 7.76 (dd, J ) 8.8 Hz, J ) 1.8 Hz, 1H, CHar); 8.08 (d, J )
1.8 Hz, 1H, CHar). Anal. (C9H7Br2F2NO3) C, H, N.
(Z)-1-Bromo-1,2-difluoro-2-(3-nitro-4-methoxy-phenyl)-
ethene (25b). The same procedure used for compound 25a, after
purification on a silica gel column using hexane/ethyl acetate 8:2,
gave 100 mg (0.34 mmol; 13%) of product as a yellow oil. Rf )
1
0.36 in hex/acetone 8:2. MS (IS): [M + Na]+ ) 315.8/317.8. H
NMR (200 MHz, CDCl3, δ): 4.03 (s, 3H, OCH3); 7.18 (d, J ) 9.2
Hz, 1H, CHar); 7.78 (dd, J ) 9.2 Hz, J ) 2.6 Hz, 1H, CHar); 8.09
(d, J ) 2.6 Hz, 1H, CHar). Anal. (C9H6BrF2NO3) C, H, N.
(Z/E)-1,2-Difluoro-1-(3,4,5-trimethoxyphenyl)-2-(3-nitro-4-
methoxy-phenyl)ethene (11a,b). The same procedure used for
compound 26a, after purification on a silica gel column using

Fluorinated Combretastatin Analogues
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 9 2717
layer was dried over anhydrous sodium sulfate. Chromatographic
separation of the crude product on silica gel with hexane/EtOAc
90:10 gave 150 mg (0.47 mmol; 24%) of an 8:2 E/Z mixture. Rf )
0.66 in hexane/EtOAc 6:4 for the stereoisomer mixture. A small
amount of each of the two geometric isomers (Z-isomer as an oil
and E-isomer as a white solid) was isolated via preparative HPLC
to confirm the structure. MS (IS): [M - H]- ) 245.1/247.1.
E-isomer, and Rf ) 0.26 for Z-isomer. MS (IS): [M + Na]+
)
1
386.3. E-Isomer (12a): H NMR (200 MHz, CDCl3, δ): 3.74 (s,
6H, 2 × OCH3); 3.89 (s, 3H, OCH3); 3.95 (s, 3H, OCH3); 6.34 (d,
J ) 20.5 Hz, 1H, CH); 6.66 (s, 2H, 2 × CH); 6.97 (d, J ) 8.8 Hz,
1H, CH); 7.35 (dd, J ) 8.8 Hz, J ) 2.2 Hz, 1H, CH); 7.72 (d, J )
1
2.2 Hz, 1H, CH); mp ) 124–127 °C. Z-Isomer (12b): H NMR
(200 MHz, CDCl3, δ): 3.91 (s, 3H, OCH3); 3.94 (s, 6H, 2 × OCH3);
4.00 (s, 3H, OCH3); 6.18 (d, J ) 38.5 Hz, 1H, CH); 6.85 (s, 2H,
2 × CH); 7.11 (d, J ) 8.8 Hz, 1H, CH); 7.81 (dd, J ) 8.8 Hz, J
) 2.2 Hz, 1H, CH); 8.13 (d, J ) 2.2 Hz, 1H, CH). Tdec ) 220 °C.
Anal. (C18H18FNO6) C, H, N.
1
E-Isomer: H NMR (200 MHz, CDCl3, δ): 3.93 (s, 3H, OCH3);
5.62 (bs, 1H, OH); 6.59 (d, J ) 15.4 Hz, 1H, CH); 6.85 (d, J )
8.1 Hz, 1H, CH); 6.99 (dd, J ) 8.1 Hz, J ) 2.2 Hz, 1H, CH); 7.16
1
(d, J ) 2.2 Hz, 1H, CH). Z-Isomer: H NMR (200 MHz, CDCl3,
δ): 3.92 (s, 3H, OCH3); 5.61 (s, 1H, OH); 5.88 (d, J ) 32.8 Hz,
1H, CH); 6.82 (d, J ) 8.4 Hz, 1H, CH); 6.90 (dd, J ) 8.4 Hz, J )
2.2 Hz, 1H, CH); 7.06 (d, J ) 2.2 Hz, 1H, CH); mp ) 69–71 °C.
Anal. (C9H8BrFO2) C, H.
5-[(E)-2-Fluoro-2-(3,4,5-trimethoxy-phenyl)-vinyl]-2-methoxy-
phenylamine (15a). Zinc powder (40 mg; 0.6 mmol; 20 equiv)
was added at room temperature to a solution of nitro stilbene 12a
(12 mg; 0.03 mmol) in 2 mL of acetic acid; the reaction mixture
was stirred overnight at RT. After dilution with EtOAc (10 mL),
the organic layer was washed with 2 N aq NaOH (2 × 5 mL), and
the crude product was purified by chromatography on silica gel
with hexane/EtOAc 7:3 to give 9 mg (0.027 mmol; 89%) of a pale
5-[(E/Z)-2-Fluoro-2-(3,4,5-trimethoxy-phenyl)-vinyl]-2-meth-
oxy-phenol (8a,b). A solution of alcohol 31a (80 mg; 0.32 mmol),
3,4,5-trimethoxybenzenboronic acid (204 mg; 0.96 mmol; 3 equiv),
Pd(Ph3P)4 (20 mg; 0.017 mmol; 0.05 equiv), and 1 mL of 2 M aq
Na2CO3 in 10 mL toluene was refluxed overnight until completion.
After the addition of anhydrous sodium sulfate, the reaction mixture
was passed through a short silica pad to remove catalyst. Chro-
matography of the crude product on silica gel with hexane/EtOAc
85:15 to 8:2 gave 70 mg (0.21 mmol; 66%) of a 2.6:1 Z/E mixture
yellow oil. Rf ) 0.37 in hexane/EtOAc 7:3. MS (IS): [M + H]+
)
334.2, [M + Na]+ ) 356.2. 1H NMR (200 MHz, CDCl3, δ): 3.73
(s, 6H, 2 × OCH3); 3.85 (s, 3H, OCH3); 3.88 (s, 3H, OCH3); 6.35
(d, J ) 21.9 Hz, 1H, CH); 6.67–6.73 (m, 5H, 5 × CH). Anal.
(C18H20FNO4) C, H, N.
1
(by H NMR), which could not be separated even by preparative
5-[(Z)-2-Fluoro-2-(3,4,5-trimethoxy-phenyl)-vinyl]-2-methoxy-
phenylamine (15b). The same protocol as 15a gave 15b as a pale
HPLC but only after derivatization as silyl ether. Deprotection of
the silyl derivative (33) with TBAF (see compound 7) gave the
final product as a colorless oil. Rf ) 0.13 in hexane/EtOAc 8:2.
MS (IS): [M - H]- ) 333.2. E-Isomer (8a): 1H NMR (200 MHz,
CDCl3, δ): 3.73 (s, 6H, 2 × OCH3); 3.89 (s, 6H, 2 × OCH3); 5.55
(s, 1H, OH); 6.36 (d, J ) 22 Hz, 1H, CH); 6.70–6.75 (m, 4H, 4 ×
CH); 6.85 (d, J ) 1.5 Hz, 1H, CH). Z-Isomer (8b): 1H NMR (200
MHz, CDCl3, δ): 3.90 (s, 6H, 2 × OCH3); 3.94 (s, 6H, 3 × OCH3);
5.64 (s, 1H, OH); 6.14 (d, J ) 39.6 Hz, 1H, CH); 6.85 (s, 2H, 2 ×
CH); 6.87 (d, J ) 8.4 Hz, 1H, CH); 7.14 (dd, J ) 8.4 Hz, J ) 1.8
Hz, 1H, CH); 7.31 (d, J ) 1.8 Hz, 1H, CH). Anal. (C18H19FO5)
C, H.
4-[(E/Z)-2-Bromo-2-fluoro-vinyl]-1-methoxy-2-nitro-ben-
zene (31b). A solution of CFBr3 (241 µL; 2.4 mmol; 1.2 equiv)
and Ph3P (1.153 g; 4.4 mmol; 2.2 equiv) in 25 mL of CH2Cl2 was
stirred for 0.5 h in an ice bath; 3-nitro-4-methoxybenzaldehyde
(362.4 mg; 2 mmol) was added and the reaction mixture was kept
in an ice bath for 30 min and at room temperature for 2 h until
completion. After dilution with CH2Cl2 (30 mL) and washing with
water (2 × 20 mL), the organic layer was dried over anhydrous
sodium sulfate. Chromatographic separation of the crude product
on silica gel with hexane/EtOAc 90:10 gave 130 mg (0.47 mmol;
24%) of a 65.8:34.2 E/Z mixture. Irradiation of the mixture (at a
concentration of 1 mg/mL in CH3CN) with a 160 W lamp at 366
nm led to a 45.4:54.6 E/Z ratio. Rf ) 0.61 in hexane/EtOAc 7:3. A
small amount of each of the two geometric isomers was isolated
as yellow oils via preparative HPLC to confirm the structure. MS
(IS): [MH]+ ) 276.1/278.1 [M + Na]+ ) 298.1/300.1. Z-Isomer:
1H NMR (200 MHz, CDCl3, δ): 4.00 (s, 3H, OCH3); 6.63 (d, J )
14.3 Hz, 1H, CH); 7.11 (d, J ) 8.8 Hz, 1H, CH); 7.67 (dd, J ) 8.8
Hz, J ) 2.2 Hz, 1H, CH); 8.04 (d, J ) 2.2 Hz, 1H, CH). E-Isomer:
1H NMR (200 MHz, CDCl3, δ): 4.00 (s, 3H, OCH3); 5.96 (d, J )
31.9 Hz, 1H, CH); 7.08 (d, J ) 8.8 Hz, 1H, CH); 7.60 (dd, J ) 8.8
Hz, J ) 2.2 Hz, 1H, CH); 7.90 (d, J ) 2.2 Hz, 1H, CH). Anal.
(C9H7BrFNO3) C, H, N.
yellow oil. Rf ) 0.29 in hexane/EtOAc 7:3. MS (IS): [M + H]+
)
334.2, [M + Na]+ ) 356.2. 1H NMR (200 MHz, CDCl3, δ): 3.90
(s, 3H, OCH3); 3.91 (s, 3H, OCH3); 3.93 (s, 6H, 2 × OCH3); 6.13
(d, J ) 39.6 Hz, 1H, CH); 6.83 (s, 2H, 2 × CH); 6.84 (d, J ) 8.4
Hz, 1H, CH); 7.11 (dd, J ) 8.4 Hz, J ) 2.2 Hz, 1H, CH); 7.25 (d,
J ) 2.2 Hz, 1H, CH). Anal. (C18H20FNO4) C, H, N.
2,2,2-Tribromo-1-[3-(tert-butyl-dimethyl-silanyloxy)-4-meth-
oxyphenyl]-ethanol (34). To a solution of 22a (2.66 g; 10 mmol)
and CBr4 (5.31 g; 16 mmol; 1.6 equiv) in 50 mL of Et2O/THF 1:1,
previously cooled down to –129 °C, a 1.6 M BuLi solution (9 mL;
14 mmol; 1.4 equiv) was added dropwise in an inert atmosphere.
The reaction mixture was allowed to warm up to –70 °C until
completion; the solution was then diluted with EtOAc (40 mL) and
washed with NH4Cl saturated solution (40 mL). The crude product
was purified by chromatography on silica gel with hexane/EtOAc
9:1 to give 3.3 g (6.4 mmol; 40%) of product as an orange oil. Rf
) 0.42 in hexane/EtOAc 9:1. MS (IS): [M + Na]+ ) 538.9/540.9/
542.9/544.9. 1H NMR (200 MHz, CDCl3, δ): 0.18 (s, 6H, 2 ×
CH3); 1.02 (s, 9H, 3 × CH3); 3.50 (d, J ) 3.7 Hz, 1H, OH); 3.84
(s, 3H, OCH3); 5.09 (d, J ) 3.7 Hz, 1H, CH); 6.85 (d, J ) 8.4 Hz,
1H, CH); 7.22 (d, J ) 2.2 Hz, 1H, CH); 7.27 (dd, J ) 8.4 Hz, J )
2.2 Hz, 1H, CH). Anal. (C15H23Br3O3Si) C, H.
tert-Butyl-[2-methoxy-5-(2,2,2-tribromo-1-fluoro-ethyl)-phe-
noxy]-dimethyl-silane (35). (Diethylamino)sulfur trifluoride 0.65
mL (5 mmol; 1.6 equiv) in 10 mL of CH2Cl2 was added to a
solution of 1.6 g (3.1 mmol) alcohol 34 in 10 mL CH2Cl2 at -78
°C. The reaction mixture was allowed to warm up to 0 °C over a
period of 2 h, quenched with 25 mL of NaHCO3 saturated solution,
and diluted with 20 mL of CH2Cl2. The organic phase was dried
over anhydrous sodium sulfate and purified by chromatography on
silica gel using hexane/ethyl acetate 95:5 to give 1.213 g (2.32
mmol; 74.8%) of a yellow oil. Rf ) 0.60 (hex/AcOEt 9:1). MS
(IS): [M + Na]+ ) 540.8/542.8/544.9/546.9. 1H NMR (200 MHz,
CDCl3, δ): 0.20 (s, 6H, 2 × CH3); 1.03 (s, 9H, 3 × CH3); 3.87 (s,
3H, OCH3); 5.72 (d, J ) 42.9 Hz, 1H, CH); 6.90 (d, J ) 8.4 Hz,
1H, CH); 7.23 (d, J ) 2.2 Hz, 1H, CH); 7.28 (dd, J ) 8.4 Hz, J )
2.2 Hz, 1H, CH). Anal. (C15H22Br3FO2Si) C, H.
5-[(E/Z)-1-Fluoro-2-(4-methoxy-3-nitro-phenyl)-vinyl]-1,2,3-
trimethoxy-benzene (12a,b). A solution of alcohol 31b (104 mg;
0.38 mmol), 3,4,5-trimethoxybenzenboronic acid (242 mg; 1.14
mmol; 3 equiv), Pd(Ph3P)4 (23 mg; 0.02 mmol; 0.05 equiv), and 1
mL of 2 M aq Na2CO3 in 10 mL toluene was refluxed overnight
until completion. After addition of anhydrous sodium sulfate, the
reaction mixture was passed through a short silica pad to remove
catalyst. Chromatography of the crude product on silica gel with
hexane/EtOAc 90:10 to 85:15 gave 25 mg (0.07 mmol; 18%) of
E-isomer and 23 mg (0.06 mmol; 16%) of Z-isomer, both isomers
obtained as yellow solids. Rf ) 0.39 in hexane/EtOAc 7:3 for
tert-Butyl-[5-(2,2-dibromo-1-fluoro-vinyl)-2-methoxy-phenoxy]-
dimethyl-silane (36). DBU (0.411 mL; 2.7 mmol; 1.5 equiv) was
added to a solution of 940 mg (1.8 mmol) of 35 in 10 mL of CH2Cl2
at room temperature, and the reaction mixture stirred overnight at
RT. The solution was diluted with CH2Cl2 (30 mL) and washed
with water (2 × 20 mL). The crude product was purified by silica
gel chromatography with hexane/EtOAc 98:2 to give 658 mg (1.49
mmol; 83%) of a pale yellow oil. Rf ) 0.4 in hexane/EtOAc 96:4.

2718 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 9
Alloatti et al.
1
MS (IS): [M + Na]+ ) 461.0/463.0/464.0. H NMR (200 MHz,
2 × CH); 7.12 (d, J ) 2.2 Hz, 1H, CH); 7.23 (dd, J ) 8.4 Hz, J
) 2.2 Hz, 1H, CH). Anal. (C24H33FO5Si) C, H.
CDCl3, δ): 0.20 (s, 6H, 2 × CH3); 1.04 (s, 9H, 3 × CH3); 3.86 (s,
3H, OCH3); 6.89 (d, J ) 8.4 Hz, 1H, CH); 7.24 (d, J ) 2.2 Hz,
1H, CH); 7.31 (dd, J ) 8.4 Hz, J ) 2.2 Hz, 1H, CH). Anal.
(C15H21Br2FO2Si) C, H.
5-[(E)-1-Fluoro-2-(3,4,5-trimethoxy-phenyl)-vinyl]-2-methoxy-
phenol (9a). A solution of 38 (23 mg; 0.051 mmol) was dissolved
in 2 mL of THF, and 100 µL (0.100 mmol; 2 equiv) of 1 M TBAF
in THF were added. The reaction was run to completion in 1 h.
After dilution with EtOAc (5 mL) and washing with H2O (5 mL),
the crude product was chromatographed on silica gel with hexane/
EtOAc 7:3 to give 12 mg (0.036; 71%) of a waxy white solid. Rf
{5-[(Z/E)-2-Bromo-1-fluoro-2-(3,4,5-trimethoxy-phenyl)-vinyl]-
2-methoxy-phenoxy}-tert-butyl-dimethyl-silane (37). A solution
of 36 (232 mg; 0.53 mmol), 3,4,5-trimethoxybenzenboronic acid
(330 mg; 1.56 mmol; 3 equiv), Pd(Ph3P)4 (30 mg; 0.026 mmol;
0.05 equiv), and 2 mL of 2 M aq Na2CO3 in 10 mL of toluene was
refluxed overnight until completion. After addition of anhydrous
sodium sulfate, the reaction mixture was passed through a short
silica pad to remove catalyst. Chromatography of the crude product
on silica gel with hexane/EtOAc 95: 5 to 8:2 gave 87 mg (0.17
mmol; 32%) of the Z-isomer (by 1H NMR NOE) and 12 mg (0.023
mmol; 4%) of the E-isomer; both isomers obtained as pale yellow
oils. Rf ) 0.3 in hexane/EtOAc 9:1 for Z-isomer. Rf ) 0.35 in
hexane/EtOAc 9:1 for E-isomer. MS (IS): [M + Na]+ ) 549.3/
551.3. Z-Isomer: 1H NMR (200 MHz, CDCl3, δ): 0.02 (s, 6H 2 ×
CH3); 0.92 (s, 9H, 3 × CH3); 3.75 (s, 6H, 2 × OCH3); 3.79 (s, 3H,
OCH3); 3.86 (s, 3H, OCH3); 6.57 (s, 2H, 2 × CH); 6.69 (d, J )
2.2 Hz, 1H, CH); 6.72 (d, J ) 9.5 Hz, 1H, CH); 6.92 (dd, J ) 8.5
Hz, J ) 2.2 Hz, 1H, CH). E-Isomer: 1H NMR (200 MHz, CDCl3,
δ): 0.17 (s, 6H, 2 × CH3); 1.00(s, 9H, 3 × CH3); 3.85 (s, 3H,
OCH3); 3.87 (s, 3H, OCH3); 3.88 (s, 6H, 2 × OCH3); 6.82 (s, 2H,
2 × CH); 6.87 (d, J ) 8.4 Hz, 1H, CH); 7.32 (d, J ) 2.2 Hz, 1H,
CH); 7.37 (dd, J ) 8.4 Hz, J ) 2.2 Hz, 1H, CH). Anal.
(C24H32BrFO5Si) C, H.
1
) 0.20 in hexane/EtOAc 8:2. MS (IS): [M - H]- ) 333.1. H
NMR (300 ΜHz, CDCl3 δ): 3.71 (s, 6H, 2 × OCH3); 3.85 (s, 3H,
OCH3); 3.92 (s, 3H, OCH3); 5.61 (s, 1H, OH); 6.43 (s, 2H, 2 ×
CH); 6.80 (dd, J ) 8.4 Hz, J ) 0.7 Hz, 1H, CH); 7.01 (dq, J ) 8.4
Hz, J ) 1.8 Hz, J ) 0.7 Hz, 1H, CH); 7.10 (d, J ) 1.8 Hz, 1H,
CH). Anal. (C18H19FO5) C, H.
5-[(Z)-1-Fluoro-2-(3,4,5-trimethoxy-phenyl)-vinyl]-2-methoxy-
phenol (9b). Analogous deprotection of the Z-isomer of 38 gave
9b: 1H NMR (300 ΜHz, CDCl3 δ): 3.92 (s, 6H, 2 × OCH3); 5.67
(bs, 1H, OH); 6.12 (d, J ) 39.2 Hz, 1H, CH); 6.88 (s, 2H, 2 ×
CH); 6.89 (d, J ) 8.1 Hz, 1H, CH); 7.18 (dd, J ) 8.06 Hz, J ) 2.2
Hz, 1H, CH); 7.21 (d, J ) 2.2 Hz, CH). Anal. (C18H19FO5) C, H.
5-[(Z)-2-Bromo-1-fluoro-2-(3,4,5-trimethoxy-phenyl)-vinyl]-
2-methoxy-phenol (10). A solution of 37 (Z-isomer) (20 mg; 0.038
mmol) was dissolved in 2 mL THF, 76 µL (0.076 mmol; 2 equiv)
of 1 M TBAF in THF were added. The reaction was run to
completion in 1 h. After dilution with EtOAc (5 mL) and washing
with H2O (5 mL), the crude product was chromatographed on silica
gel with hexane/EtOAc 7:3 to give 14 mg (0.034 mmol; 89%) of
a white solid. Rf ) 0.11 in hexane/EtOAc 8:2. MS (IS): [M - H]–
1
) 411.1/413.1, [M + Na]+ ) 435.1/437.1. H NMR (300 MHz,
tert-Butyl-{5-[(E/Z)-1-fluoro-2-(3,4,5-trimethoxy-phenyl)-vinyl]-
2-methoxy-phenoxy}-dimethyl-silane (38a,b). Both geometric
isomers of 38 were obtained by two different approaches. Via 37:
Tributyltin hydride (82 µL; 0.28 mmol; 4.4 equiv) and AIBN (5
mg; 0.03; 0.47 equiv) were added to a solution of 37 (34 mg; 0.064
mmol) in 3 mL of 1,2-dichloroethane and refluxed. After 16 h,
another 82 µL of Bu3SnH and 5 mg of AIBN were necessary to
drive the reaction to completion. After cooling to room temperature,
the reaction mixture was diluted with CH2Cl2 (20 mL), and washed
with water (2 × 10 mL). Chromatography of the crude product on
silica gel with hexane/EtOAc 85:15 gave 23 mg (0.051 mmol; 80%)
of a pale yellow oil.
Via 39: (a) Debromination. Tributyltin hydride (358 µL; 1.23
mmol; 4.4 equiv) and AIBN (23 mg; 0.14; 0.5 equiv) were added
to a solution of 36 (123 mg; 0.28 mmol) in 3 mL of 1,2-
dichloroethane and refluxed. After 16 h, another 358 µL of Bu3SnH
and 23 mg of AIBN were necessary to drive the reaction to
completion. After cooling to room temperature, the reaction mixture
was diluted with CH2Cl2 (20 mL) and washed with water (2 × 10
mL). Chromatography of the crude product on silica gel with
hexane/EtOAc 98:2 gave 62 mg (0.25 mmol; 89%) of product as
a mixture of the two isomers Z/E (1:2.2), whose ratio was
CDCl3, δ): 3.76 (s, 6H, 2 × OCH3); 3.88 (s, 6H, 2 × OCH3); 5.56
(s, 1H, OH); 6.57 (s, 2H, 2 × CH); 6.67 (d, J ) 8.8 Hz, 1H, CH);
6.73 (dd, J ) 8.8 Hz, J ) 1.8 Hz, 1H, CH); mp ) 104–107 °C.
Anal. (C18H18BrFO5) C, H.
5-[(Z/E)-2-Bromo-2-fluoro-vinyl]-1,2,3-trimethoxy-benzene
(41a,b). A solution of CFBr3 (600 µL; 6 mmol; 1.2 equiv) and
Ph3P (5.8 g; 22 mmol; 4.4 equiv) in 40 mL of CH2Cl2 was stirred
for 0.5 h in an ice bath; 3,4,5-trimethoxybenzaldheyde (40; 980
mg; 5 mmol) was added and the reaction mixture then kept in an
ice bath for 30 min and at room temperature for 2 h until
completion. After dilution with CH2Cl2 (30 mL) and washing with
water (2 × 20 mL), the organic layer was dried over anhydrous
sodium sulfate. Chromatographic separation of the crude product
on silica gel with hexane/EtOAc 90:10 gave 600 mg (2.1 mmol;
42%) of a 1:1 E/Z mixture. Rf ) 0.23 in hexane/EtOAc 9:1. A
small amount of each of the two geometric isomers was isolated
via preparative HPLC, as pale yellow oils, to confirm the structure.
1
MS (IS): [MH]+ ) 291.1/294.0. Z-Isomer (41a): H NMR (200
MHz, CDCl3, δ): 3.88 (s, 9H, 3 × OCH3); 6.62 (d, J ) 15.4 Hz,
1
1H, CH); 6.75 (s, 2H, 2 × CH). E-Isomer (41b): H NMR (200
MHz, CDCl3, δ): 3.87 (s, 9H, 3 × OCH3); 5.93 (d, J ) 32.2 Hz,
1H, CH); 6.65 (s, 2H, 2 × CH). Anal. (C11H12BrFO3) C, H.
5-[(E/Z)-2-Fluoro-2-(4-methoxy-3-nitrophenyl)-vinyl]-1,2,3-
trimethoxy-benzene (13a,b). A solution of stirene 41 (102 mg;
0.35 mmol), of 3-nitro-4-methoxyphenylboronic acid (234 mg; 1.19
mmol; 3.4 equiv), Pd(Ph3P)4 (21 mg; 0.018 mmol; 0.05 equiv),
and 1 mL of 2 M aq Na2CO3 in 10 mL toluene was refluxed
overnight until completion. After addition of anhydrous sodium
sulfate, the reaction mixture was passed through a short silica pad
to remove catalyst. Chromatography of the crude product on silica
gel with hexane/EtOAc 90:10 to 85:15 gave 35 mg (0.097 mmol;
27%) of E-isomer as a yellow oil and 373 mg (0.10 mmol; 29%)
of Z-isomer as a yellow solid. Rf ) 0.36 in hexane/EtOAc 7:3 for
E-isomer and Rf ) 0.23 for Z-isomer. MS (IS): [M + Na]+ ) 386.3.
E-Isomer (13a): 1H NMR (200 MHz, CDCl3, δ): 3.74 (s, 6H, 2 ×
OCH3); 3.86 (s, 3H, OCH3); 3.98 (s, 3H, OCH3); 6.42 (s, 2H, 2 ×
CH); 6.47 (d, J ) 20.2 Hz, 1H, CH); 7.01 (d, J ) 9.2 Hz, 1H,
CH); 7.59 (dd, J ) 9.2 Hz, J ) 2.2 Hz, 1H, CH); 8.00 (d, J ) 2.2
Hz, 1H, CH). Z-Isomer (13b): 1H NMR (200 MHz, CDCl3, δ):
3.90 (s, 3H, OCH3); 3.92 (s, 6H, 2 × OCH3); 4.03 (s, 3H, OCH3);
6.22 (d, J ) 38.8 Hz, 1H, CH); 6.90 (s, 2H, 2 × CH); 7.15 (d, J
1
determined both by H NMR and by HPLC. (b) Suzuki reaction.
A solution of debrominated product obtained in the previous step
(62 mg; 0.25 mmol), 3,4,5-trimethoxybenzenboronic acid (159 mg;
0.75 mmol; 3 equiv), Pd(Ph3P)4 (14 mg; 0.012 mmol; 0.05 equiv),
and 1 mL of 2 M aq Na2CO3 in 10 mL toluene was refluxed
overnight until completion. After addition of anhydrous sodium
sulfate, the reaction mixture was passed through a short silica pad
to remove catalyst. Chromatography of the crude product on silica
gel with hexane/EtOAc 95:5 to 8:2 gave 13 mg (0.029 mmol; 12%)
of the Z-isomer and 11 mg (0.025 mmol; 10%) of the E-isomer.
MS (IS): [M + Na]+ ) 449.3/451.3. E-Isomer (38a): Rf ) 0.7 in
hexane/EtOAc 9:1. 1H NMR (200 MHz, CDCl3, δ): 0.09 (s, 6H 2
× CH3); 0.95 (s, 9H, 3 × CH3); 3.71 (s, 6H, 2 × OCH3); 3.83 (s,
3H, OCH3); 3.84 (s, 3H, OCH3); 6.31 (d, J ) 20.9 Hz, 1H, CH);
6.42 (s, 2H, 2 × CH); 6.80 (d, J ) 8.4 Hz, 1H, CH); 6.94 (d, J )
2.2 Hz, 1H, CH); 7.081 (dd, J ) 8.4 Hz, J ) 2.2 Hz, 1H, CH).
Z-Isomer (38b): Rf ) 0.6 in hexane/EtOAc 9:1. 1H NMR (200 MHz,
CDCl3, δ): 0.21 (s, 6H, 2 × CH3); 1.04 (s, 9H, 3 × CH3); 3.87 (s,
3H, OCH3); 3.89 (s, 3H, OCH3); 3.92 (s, 6H, 2 × OCH3); 6.08 (d,
J ) 39.2 Hz, 1H, CH); 6.88 (d, J ) 8.4 Hz, 1H, CH); 6.89 (s, 2H,

Fluorinated Combretastatin Analogues
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 9 2719
) 9.2 Hz, 1H, CH); 7.80 (dd, J ) 9.2 Hz, J ) 2.2 Hz, 1H, CH);
8.12 (d, J ) 2.2 Hz, 1H, CH); mp ) 138–141 °C. Anal.
(C18H18FNO6) C, H, N.
(Z)-1,2-Difluoro-1-(3,4,5-trimethoxyphenyl)-2-(3-hydroxy-4-
methoxyphenyl)ethene O-Disodium Phosphate (43). According
to the cited three-step protocol,5,30 we obtained the disodium
phosphate prodrug of 7 as a white solid: [M - 2Na + H]- ) 431.2
1H NMR (300 MHz, D2O, δ): 3.55 (s, 6H, 2 × OCH3); 3.63 (s,
3H, OCH3); 3.70 (s, 3H, OCH3); 6.59 (s, 2H, 2 × CH); 6.72 (d, J
) 8.1 Hz, 1H, CH); 6.78 (d, J ) 8.1 Hz, 1H, CH); 7.48 (s, 1H,
CH). Anal. (C18H17F2O8PNa2) C, H.
5-[(E)-1-Fluoro-2-(3,4,5-trimethoxy-phenyl)-vinyl]-2-methoxy-
phenylamine (16a). Zinc in powder (97 mg; 1.48 mmol; 20 eq.)
was added, at room temperature, to a solution of stilbene 13a (27
mg; 0.074 mmol) in 2 mL of acetic acid; the reaction mixture was
stirred overnight at RT. After dilution with EtOAc (10 mL), the
organic layer was washed with 2 N aq NaOH (2 × 5 mL), and the
crude product was purified by chromatography on silica gel with
hexane/EtOAc 7:3 to give 20 mg (0.060 mmol; 81%) of a pale
yellow oil (16a). Rf ) 0.49 in hexane/EtOAc 6:4. MS (IS): [M +
Biology
Cell Culture. Primary cultures of bovine microvascular endot-
helial cells (BMEC) were obtained from bovine adrenal glands as
described by Folkman.45 BMEC were maintained in DMEM,
supplemented with 20% fetal calf serum (FCS), 50 units/mL heparin
(Sigma, St. Louis, MO), 50 µg/mL bovine brain extract, and 100
units/mL gentamycin.
NCI-H460 human nonsmall cell lung carcinoma and HT29 colon
carcinoma were purchased from ATCC and cultured according to
manufacturer’s instructions.
1
Na]+ ) 356.2. H NMR (200 MHz, CDCl3, δ): 3.71 (s, 6H, 2 ×
OCH3); 3.85 (s, 3H, OCH3); 3.88 (s, 3H, OCH3); 6.28 (d, J ) 21.6
Hz, 1H, CH); 6.44 (s, 2H, 2 × CH); 6.74 (d, J ) 9.2 Hz, 1H, CH);
6.91–6.94 (m, 2H, 2 × CH). Anal. (C18H20FNO4) C, H, N.
5-[(Z)-1-Fluoro-2-(3,4,5-trimethoxy-phenyl)-vinyl]-2-methoxy-
phenylamine (16b). The above-described synthetic procedure gave
the Z-isomer 16b with 77% yield as a pale yellow oil. Rf ) 0.50 in
1
hexane/EtOAc 6:4. MS (IS): [M + Na]+) 356.2. H NMR (200
Tubulin Polymerization Assay. The tubulin polymerization test
was performed by CytoDINAMIX ScreenTM (Cytoskeleton Inc.,
Denver, CO). A total of 100 µL of 3 mg/mL HTS tubulin in G-PEM
buffer plus 5% glycerol at 4 °C were pipetted into wells of the
96-well plate and incubated at 37 °C in presence or absence of
single compounds. Tubulin polymerization was detected by measur-
ing the absorbance of the solution (340 nm) for 60 min. Because
the amount of tubulin polymerized is directly proportional to the
AUC (area under the curve), we used AUC to determine the
concentration that inhibited tubulin polymerization by 50% (IC50).
The AUC of the untreated control was set to 100% polymerization
(maximal attainable), and the IC50 was calculated by nonlinear
regression.
Cytotoxicity Assay. To test the effect of the drugs on cell growth,
the above cell lines were resuspended at different concentrations
in 200 µL of appropriate growth medium and seeded in 96-well
plates. A total of 24 h after plating, scalar concentrations of each
drug were added to the cells. The drugs were removed after 24 h
and, after an additional 48 h of recovery, the number of surviving
cells was determined by staining with sulforhodamine B test as
previously described.46 Optical density was measured at 564 nm.
Results, expressed as concentrations that inhibit 50% of cell growth
(IC50), were calculated by the ALLFIT program.
Flow Cytometry. NCI-H460 cells were seeded in 100 mm plastic
dishes in RPMI-1640 complete medium, allowed to adhere, and
exposed for 24 h to the various molecules (IC80 dose). The cells
were then harvested with trypsin/EDTA, pooled with the respective
supernatant, and fixed in ice-cold 70% ethanol at 4 °C. On the day
of analysis, ethanol was removed by centrifugation, and cells were
rinsed twice with PBS and treated with RNase (75 KU/mL) for 30
min at 37 °C. Propidium iodide (PI) was finally added (50 µg/mL)
to stain cellular DNA, and samples (2 × 104 cells) were processed
on a FACScan flow cytometer (Becton Dickinson) to assess cell
cycle distribution and apoptosis. Apoptosis was evaluated by
measuring the percentage of cells with hypodiploid DNA content
(sub-G1/0 population) with CELLQuest software, while cell cycle
analysis was performed with ModFit software.
Immunofluorescence. For tubulin immunostaining, the cells,
cultured on gelatin-coated coverslips, were treated with the
compounds at concentrations near their respective IC80values. After
24 h, the cells were washed with PBS and then fixed in cold
methanol for 5 min and acetone for 1 min. Next, the fixed cells
were incubated in a PBS solution containing 0.1% Triton X-100
and 0.5% acetic acid for 10 min and, after three washings in PBS,
in a solution containing 2 µg/mL of the monoclonal anti-R-tubulin
antibody (Sigma Chemical Co., St. Louis, MO) for 1 h at RT in a
humidified chamber. Following the primary antibody incubation,
the cells were washed three times in PBS and incubated with an
AlexaFluor-conjugated goat antimouse secondary antibody (Mo-
lecular Probes, Leiden, The Netherlands) at a concentration of 20
µg/mL for 1 h at RT in the dark. For nuclear staining, the cells
were incubated with Hoechst 33342 (Sigma Chemicals) at a
MHz, CDCl3, δ): 3.89 (s, 3H, OCH3); 3.91 (s, 6H, 2 × OCH3);
3.92 (s, 3H, OCH3); 6.10 (d, J ) 39.2 Hz, 1H, CH); 6.85 (d, J )
10.6 Hz, 1H, CH); 6.88 (s, 2H, 2 × CH); 7.107–7.142 (m, 2H, 2
× CH). Anal. (C18H20FNO4) C, H, N.
6-[(E)-2-Fluoro-2-(3,4,5-trimethoxy-phenyl)-vinyl]-benzo-
[b]thiophen-4-ol (18a). The procedure used for compound 31a gave
intermediate 42a, which without further purification underwent
Suzuki coupling to give, after preparative HPLC purification with
CH3CN/H2O 60:40, 9 mg (0.025 mmol; 18% in two steps) of the
final compound 44a as a yellow solid. Rf ) 0.50 in hexane/EtOAc
7:3. [M - H]- ) 359.3. 1H NMR (200 MHz, CDCl3, δ): 3.66 (s,
6H, 2 × OCH3); 3.,88 (s, 3H, OCH3); 6.49 (d, J ) 21.6 Hz, 1H,
CH); 6.61 (s, 1H,CH); 6.71 (s, 2H, 2 × CH); 7.34 (d, J ) 5.4 Hz,
1H, CH); 7.36 (s, 1H, CH); 7.42 (d, J ) 5.4 MHz, 1H, CH); mp
) 158–161 °C. Anal. (C19H17FO4S) C, H, S.
6-[(Z)-2-Fluoro-2-(3,4,5-trimethoxy-phenyl)-vinyl]-benzo-
[b]thiophen-4-ol (18b). Analogous procedure gave the Z-isomer
1
as yellow solid: Rf ) 0.33 in Hexane/EtOAc 7:3. H NMR (200
MHz, CDCl3, δ): 3.92 (s, 3H, OCH3); 3.95 (s, 6H, 2 × OCH3);
6.28 (d, J ) 39.2 Hz, 1H, CH); 6.88 (s, 2H, 2 × CH); 7.12 (s, 1H,
CH); 7.39 (d, J ) 5.5 Hz, 1H, CH); 7.47 (d, J ) 5.5 Hz, 1H, CH);
7.72 (s, 1H, CH); mp ) 134–137 °C. Anal. (C19H17FO4S) C, H, S.
6-[(E)-2-Fluoro-2-(3,4,5-trimethoxy-phenyl)-vinyl]-benzofu-
ran-4-ol (19a). The procedure applied for 18 gave compound 19a
in a 22% yield (two steps) as a pale yellow oil. Rf ) 0.37 in hexane/
1
EtOAc 7:3. [M - H]- ) 343.2. H NMR (200 MHz, CDCl3, δ):
3.68 (s, 6H, 2 × OCH3); 3.,89 (s, 3H, OCH3); 6.49 (d, J ) 21.5
Hz, 1H, CH); 6.53 (bs, 1H, CH); 6.70 (s, 2H, 2 × CH); 6.81 (dd,
J ) 2.2 Hz, J ) 0.7 Hz, 1H, CH); 7.03 (bs, 1H, CH); 7.54 (d, J )
2.2 Hz, 1H, CH). Anal. (C19H17FO5) C, H.
6-[(Z)-2-Fluoro-2-(3,4,5-trimethoxy-phenyl)-vinyl]-benzofu-
ran-4-ol (19b). Analogous procedure of 19a gave the Z-isomer
1
(19b) as a yellow oil: Rf ) 0.23 in hexane/EtOAc 7:3. H NMR
(200 MHz, CDCl3, δ): 3.91 (s, 3H, OCH3); 3.95 (s, 6H, 2 × OCH3);
6.26 (d, J ) 38.8 Hz, 1H, CH); 6.86 (d, J ) 2.2 Hz, 1H, CH); 6.88
(s, 2H, 2 × CH); 6.99 (s, 1H, CH); 7.44 (s, 1H, CH); 7.60 (d, J )
2.2 Hz, 1H, CH). Anal. (C19H17FO5) C, H.
5-[(E/Z)-2-Fluoro-2-(3-fluoro-4-methoxy-phenyl)-vinyl]-1,2,3-
trimethoxy-benzene (17a,b). The same procedure used for com-
pound 13 gave, after purification by chromatography on silica gel
with hexane/EtOAc 9:1, the final product in a 40% yield (Rf )
0.38 in hexane/EtOAc 8:2) together with its Z-isomer (22%; Rf )
0.30 in hexane/EtOAc 8:2). E-Isomer 17a: [M + H]+ ) 337.2,
[M + Na]+ ) 359.2. 1H NMR (200 MHz, CDCl3, δ): 3.71 (s, 6H,
2 × OCH3); 3.85 (s, 3H, OCH3); 3.90 (s, 3H, OCH3); 6.36 (d, J )
22.3 Hz); 6.41 (s, 2H, 2 × CH); 6.89 (t, J ) 8.5 Hz, 1H, CH);
7.19–7.28 (m, 2H, 2 × CH); white solid, mp ) 101–104 °C.
Z-Isomer 17b: 1H NMR (200 MHz, CDCl3, δ): 3.89 (s, 3H, OCH3);
3.91 (s, 6H, 2 × OCH3); 3.94 (s, 3H, OCH3); 6.13 (d, J ) 39.2
Hz, 1H, CH); 6.88 (s, 2H, 2 × CH); 6.99 (t, J ) 8.4 Hz, 1H, CH);
7.32–7.40 (m, 2H, 2CH). Anal. (C18H18F2O4) C, H.

2720 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 9
Alloatti et al.
(13) (a) Cushman, M.; Nagarathnam, D.; Gopal, D.; He, H.-M.; Lin, C. M.;
Hamel, E. Synthesis and evaluation of analogues of (Z)-1-(4-
methoxyphenyl)-2-(3,4,5-trimethoxyphenyl)ethene as potential cyto-
toxic and antimitotic agents. J. Med. Chem. 1992, 35, 2293–2306. (b)
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.
concentration of 10 µg/mL for 10 min at RT in the dark. Coverslips
were then mounted with 50% glycerol in PBS. Cells were analyzed
with an epifluorescent microscope (Polyvar, Zeiss AG, Germany)
and imaged with a cooled CCD camera (Photometrics, Roper
Scientific Inc., CA). Images were then processed with Metamorph
software (Universal Imaging, Downingtown, PA).
(14) Welch, J. T., Ed. SelectiVe Fluorination in Organic and Bioorganic
Chemistry; ACS Symposium Series 456: Washington D.C., 1991.
(15) Seebach, D. Organic synthesis-Where now. Angew. Chem., Int. Ed.
Engl. 1990, 29, 1320.
(16) Hertel, L. W.; Ternansky, R. J. Stud. Org. Chem. (Amsterdam) 1993,
48, 23.
(17) Biffinger, J. C.; Kim, H. W.; DiMagno, S. G. The polar hydrophobicity
of fluorinated compounds. ChemBioChem 2004, 5, 622–627.
(18) Giannini, G. Fluorinated anthracyclines: Synthesis and biological
activity. Med. Chem. ReV.sOnline 2004, 1, 47–71.
(19) Ringel, I.; Jaffe, D.; Alerhand, S.; Boye, O.; Muzaffar, A.; Brossi, A.
Fluorinated colchicinoids: Antitubulin and cytotoxic properties. J. Med.
Chem. 1991, 34, 3334–3338.
Acknowledgment. The authors would like to thank Dr.
Michel Simard (Université de Montréal) for X-ray analysis and
Prof. Stephen Hanessian for critical revision of the manuscript
and helpful suggestions.
1
Supporting Information Available: H NMR, 19F NMR, and
MS of new derivatives, as well as crystal data for compound 7.
This material is available free of charge via the Internet at http://
pubs.acs.org.
(20) Hadfield, J. A.; McGown, A. T.; Mayalarp, S. P.; Land, E. J.; Hamblett,
I.; Gaukroger, K.; Lawrence, J. N.; Hepworth, L. A.; Butler, J.
Preparation of substituted stilbenes as antitumor agents, Patent
WO02050007, 2005.
(21) Gaukroger, K.; Hadfield, J. A.; Lawrence, N. J.; Nolan, S.; McGown,
A. T. Structural requirements for the interaction of combretastatins
with tubulin: how important is the trimethoxy unit. Org. Biomol. Chem.
2003, 1, 3033–3037.
References
(1) (a) Pettit, G. R.; Singh, S. B.; Hamel, E.; Lin, C. M.; Alberts, D. S.;
Garcia-Kendall, D. Isolation and structure of the strong cell growth
and tubulin inhibitor combretastatin A-4. Experientia 1989, 45, 209–
211. (b) Pettit, G. R.; Cragg, G. M.; Herald, D. L.; Schmidt, J. M.;
Lobavanijaya, P. Isolation and structure of combretastatin. Can.
J. Chem. 1982, 60, 1347–1376.
(2) Nam, N. H. Combretastatin A-4 analogues as antimitotic anti-tumor
agents. Curr. Med. Chem 2003, 10, 1697–1722, and references cited
therein.
(3) (a) Hori, K.; Saito, S. Microvascular mechanisms by which the
combretastatin A-4 derivative AC7700 (AVE8062) induces tumour
blood flow stasis. Br. J. Cancer 2003, 89, 1334–1344, and references
cited therein. (b) Hori, K.; Saito, S. Induction of tumour blood flow
stasis and necrosis: a new function for epinephrine similar to that of
combretastatin A-4 derivative AVE8062 (AC7700). Br. J. Cancer
2004, 90, 549–553. (c) Tozer, G. M.; Prise, V. E.; Wilson, J.; Locke,
R. J.; Vojnovic, B.; Stratford, M. R.; Dennis, M. F.; Chaplin, D. J.
Combretastatin A-4 phosphate as a tumor vascular-targeting agent:
Early effects in tumors and normal tissues. Cancer Res. 1999, 59,
1626–1634.
(4) Chaplin, D. J.; Hill, S. A. The development of combretastatin A4
phosphate as a vascular targeting agent. Int. J. Radiat. Oncol., Biol.,
Phys. 2002, 54, 1491–1496.
(5) Pettit, G. R.; Temple, C, Jr.; Narayanan, V. L.; Varma, R.; Boyd,
M. R.; Rener, G. A.; Bansal, N. Antineoplastic agents 322. Synthesis
of combretastatin A-4 prodrugs. Anti-Cancer Drug Des. 1995, 10, 299–
309.
(6) Pettit, G. R.; Lippert, J. W. Synthesis of the combretastatin A-1 and
combretastatin B-1 prodrugs. Anti-Cancer Drug Des. 2000, 15, 203–
216.
(7) Kirwan, I. G.; Loadman, P. M.; Swaine, D. J.; Anthoney, D. A.; Pettit,
G. R.; et al. Comparative preclinical pharmacokinetic and metabolic
studies of the combretastatin prodrugs combretastatin A4 phosphate
and A1 phosphate. Clin. Cancer Res. 2004, 10, 1446–1453.
(8) Hatanaka, T.; Fujita, K.; Ohsumi, K.; Nakagawa, R.; Fukuda, Y.; Nihei,
Y.; Suga, Y.; Akiyama, Y.; Tsuji, T. Novel B-ring modified combre-
tastatin analogues: syntheses and antineoplastic activity. Bioorg. Med.
Chem. Lett. 1998, 8, 3371–3374.
(9) Ohsumi, K.; Nakagawa, R.; Fukuda, Y.; Hatanaka, T.; Morinaga, Y.;
Nihei, Y.; Ohishi, K.; Suga, Y.; Akiyama, Y.; Tsuji, T. Novel
combretastatin analogues effective against murine solid tumors: Design
and structure-activity relationships. J. Med. Chem. 1998, 41, 3022–
3032.
(22) David, P. D. Preparation and use of cis-stilbenes with vascular
damaging activity, Patent WO0112579, 2001.
(23) Lawrence, J. N.; Hepworth, L. A.; Rennison, D.; McGown, A. T.;
and Hadfield, J. A. Synthesis and anticancer activity of fluorinated
analogues of combretastatin A-4. J. Fluorine Chem. 2003, 123, 101–
108.
(24) Ducki, S.; Mackenzie, G.; Lawrence, N. J.; Snyder, J. P. Quantitative
structure-activity relationship (5D-QSAR) study of combretastatin-
like analogues as inhibitors of tubuline assembly. J. Med. Chem. 2005,
48, 457–465.
(25) (a) Bellucci, G.; Berti, G.; Chiappe, C.; Lippi, A.; Marioni, F. The
metabolism of carbamazepine in humans: Steric course of the
enzymatic hydrolysis of the 10,11-epoxide. J. Med. Chem. 1987, 30,
768–773. (b) Kevin Park, B.; Kitteringham, N. R.; O’Neill, P. M.
Metabolism of fluorine-containing drugs. Annu. ReV. Pharmacol.
Toxicol. 2001, 41, 443–470. (c) Aprile, S.; Del Grosso, E.; Tron, G. C.;
Grosa, G. In vitro metabolism study of combretastatin A-4 in rat and
human liver microsomes. Drug Metab. Dispos. 2007, 35, 2252–2261.
(26) (a) Giannini, G., Fluorocombretastatin and derivatives thereof, Patent
WO05007603, 2005. (b) Drug Data Report, 2005; 27, 361, and
reference therein reported. (c) Tron, G. C.; Pirali, T.; Sorba, G.; Pagliai,
F.; Busacca, S.; Genazzani, A. A. J. Med. Chem. 2006, 49, 3033–
3044. (d) Mahindroo, N.; et al. Antitubulin agents for the treatment
of cancer - a medicinal chemistry update. Expert Opin. Ther. Pat.
2006, 16, 647–691.
(27) Simoni, D.; Romagnoli, R.; Baruchello, R.; Rondanin, R.; Rizzi’, M.;
Pavani, M. G.; Alloatti, D.; Giannini, G.; Marcellini, M.; Riccioni,
T.; Castorina, M.; Guglielmi, M. B.; Bucci, F.; Carminati, P.; Pisano,
C. Novel combretastatin analogues endowed with antitumor activity.
J. Med. Chem. 2006, 49, 3143–3152.
(28) Ravelli, R. B.; Gigant, B.; Curmi, P. A.; Jourdain, I.; Lachkar, S.;
Sobel, A.; Knossow, M. Insight into tubulin regulation from a complex
with colchicine and a stathmin-like domain. Nature 2004, 428, 198–
202.
(29) Shimizu, M.; Yamada, N.; Takebe, Y.; Hata, T.; Kuroboshi, M.;
Hiyama, T. Generation and carbonyl addition reactions of dibromof-
luoromethyllithium derived from tribromofluoromethane as applied
to the stereoselective synthesis of fluoro olefins and 2-bromo-2-fluoro-
1,3-alkanediols. Bull. Chem. Soc. Jpn. 1998, 71, 2903–2921.
(30) Burton, D. J.; Yang, Z. Y.; Qiu, W. Fluorinated ylides and related
compounds. Chem. ReV. 1996, 96, 1641–1715.
(10) Lawrence, N. J.; Ghani, F. A.; Hepworth, L. A.; Hadfield, J. A.;
McGown, A. T.; Pritchard, R. G. The synthesis of E- and Z-
combretastatin A-4 and a phenantrene from Combretum caffrum.
Synthesis 1999, 9, 1656–1660.
(31) Pettit, G. R.; Moser, B. R.; Boyd, M. R.; Schmidt, J. M.; Pettit, R. K.;
Chapuis, J. C. Antineoplastic agents 460. Synthesis of combretastatin
A-2 prodrugs. Anti-Cancer Drug Des. 2001, 16, 185–193.
(32) Pettit, G. R.; Rhodes, M. R.; Herald, D. L.; Hamel, E.; Schmidt, J. M.;
Pettit, R. K. Antineoplastic agents. 445. Synthesis and evaluation of
structural modifications of (Z)- and (E)-combretastatin A-4. J. Med.
Chem. 2005, 48, 4087–4099, and references cited therein.
(33) Ismail, F. Important fluorinated drugs in experimental and clinical use.
J. Fluorine Chem. 2002, 118, 27–33.
(11) Simoni, D.; Grisolia, G.; Giannini, G.; Roberti, M.; Rondanin, R.;
Piccagli, L.; Baruchello, R.; Rossi, R.; Romagnoli, R.; Invidiata, F. P.;
Grimaudo, S.; Jung, M. K.; Hamel, H.; Gebbia, N.; Crosta, L.;
Abbadessa, V.; Di Cristina, A.; Dusonschet, L.; Meli, M.; Tolomeo,
M. Heterocyclic and phenyl double-bond-locked combretastatin
analogues possessing potent apoptosis-inducing activity in HL60 and
in MDR cell lines. J. Med. Chem. 2005, 48, 723–736.
(12) Wang, L.; Woods, K. W.; Li, Q.; Barr, K. J.; McCroskey, R. W.;
Hannic, S. M.; Gherke, L.; Credo, R. B.; Hui, Y. H.; Marsh, K.;
Warner, R. L.; Jang, Y.; Zielinski-Mozng, N.; Frost, D.; Rosenberg,
S. H.; Sham, H. L. Potent, orally active heterocycle-based Combre-
tastatin A-4 analogues: Synthesis, structure-activity relationship,
pharmacokinetics, and in vivo antitumor activity evaluation. J. Med.
Chem., 2002, 45, 1697–1711.
(34) Maestro, Version 8.0, Impact, Version 4.5, Schrödinger, LLC: New
York, NY, 2007.
(35) Goodford, P. A computational procedure for determining energetically
favorable binding sites on biologically important macromolecules.
J. Med. Chem. 1985, 28, 849–857.

Fluorinated Combretastatin Analogues
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 9 2721
(36) Jorgensen, W. L.; Tirado-Rives, J. The OPLS force field for proteins.
(43) Friesner, R. A.; Banks, J. L.; Murphy, R. B.; Halgren, T. A.; Klicic,
J. J.; Mainz, D. T.; Repasky, M. P.; Knoll, E. H.; Shelley, M.; Perry,
J. K.; Shaw, D. E.; Francis, P.; Shenkin, P. S. Glide: A New approach
for rapid, accurate docking and scoring. 1. method and assessment of
docking accuracy. J. Med. Chem. 2004, 47, 1739–1749.
(44) Lyne, P. D.; Lamb, M. L.; Saeh, J. C. Accurate prediction of the
relative potencies of members of a series of kinase inhibitors using
molecular docking and MM-GBSA Scoring. J. Med. Chem. 2006, 49,
4805–4808.
J. Am. Chem. Soc. 1988, 110, 1657–1666.
(37) Weber, A. E.; Halgren, T. A.; Doyle, J. J.; Lynch, R. J.; Siegl, P. K. S.;
Parsons, W. H.; Greenlee, W. J.; Patchett, A. A. Design and synthesis
of P2-P1′-linked macrocyclic human renin inhibitors. J. Med. Chem.
1991, 34, 2692–2701.
(38) Halgren, T. New method for fast and accurate binding-site identification
and analysis. Chem. Biol. Drug Des. 2007, 69, 146–148.
(45) Folkman, J.; Haudenschild, C. C.; Zetter, B. R. Long-term culture of
capillary endothelial cells. Proc. Natl. Acad. Sci. U.S.A. 1979, 76,
5217–5221.
(46) Skehan, P.; Storeng, R.; Scudiero, D.; Monks, A.; McMahon, J.;
Vistica, D.; Warren, J. T.; Bokesch, H.; Kenney, S.; Boyd, M. R. New
colorimetric cytotoxicity assay for anticancer-drug screening. J. Natl.
Cancer Inst. 1990, 82, 1107–1112.
(39) MacroModel, Version 9.5, Schrödinger, LLC: New York, NY, 2007.
(40) Spartan’06, Wavefunction, Inc.: Irvine, CA.
(41) Brinck T. The use of the electrostatic potential for analysis and
prediction of intermolecular interactions. In Theoretical Organic
Chemistry, ElseVier Science & Technology, Eds.; Elsevier: Amsterdam,
1998; vol 5, pp 51–93.
(42) Glide, Version 4.5, Prime, Version 1.6, Schrödinger, LLC: New York,
NY, 2007.
JM701362M