Synthesis and structure-activity relationships of constrained heterocyclic analogues of combretastatin A4

Article abstract of DOI:10.1002/cmdc.201100154

A series of combretastatin A4 (CA4) analogues with a lactam or lactone ring fused to the trimethoxyphenyl or the B-phenyl moiety were synthesized in an efficient and stereoselective manner by using a domino Heck-Suzuki-Miyaura coupling reaction. The vascular-disrupting potential of these conformationally restricted CA4 analogues was assessed by various invitro assays: inhibition of tubulin polymerization, modification of endothelial cell morphology, and disruption of endothelial cell cords. Compounds were also evaluated for their growth inhibitory effects against murine and human tumor cells. B-ring-constrained derivatives that contain an oxindole ring (in contrast to compounds with a benzofuranone ring) as well as analogues bearing a six-membered lactone core fused to the trimethoxyphenyl ring are endowed with significant biological activity. The most potent compound of this series (oxindole 9b) is of particular interest, as it combines chemical stability and a biological activity profile characteristic of a vascular-disrupting agent.

Full text of DOI:10.1002/cmdc.201100154

DOI: 10.1002/cmdc.201100154
Synthesis and Structure–Activity Relationships of
Constrained Heterocyclic Analogues of Combretastatin A4
Martin Arthuis,[a, b] Renꢀe Pontikis,*[a, b] Guy G. Chabot,*[c, d] Johanne Seguin,[c, d]
Lionel Quentin,[c, d] Stꢀphane Bourg,[e] Luc Morin-Allory,[f] and Jean-Claude Florent*[a, b]
A series of combretastatin A4 (CA4) analogues with a lactam
or lactone ring fused to the trimethoxyphenyl or the B-phenyl
moiety were synthesized in an efficient and stereoselective
manner by using a domino Heck–Suzuki–Miyaura coupling re-
action. The vascular-disrupting potential of these conforma-
tionally restricted CA4 analogues was assessed by various in vi-
tro assays: inhibition of tubulin polymerization, modification of
endothelial cell morphology, and disruption of endothelial cell
cords. Compounds were also evaluated for their growth inhibi-
tory effects against murine and human tumor cells. B-ring-con-
strained derivatives that contain an oxindole ring (in contrast
to compounds with a benzofuranone ring) as well as ana-
logues bearing a six-membered lactone core fused to the tri-
methoxyphenyl ring are endowed with significant biological
activity. The most potent compound of this series (oxindole
9b) is of particular interest, as it combines chemical stability
and a biological activity profile characteristic of a vascular-dis-
rupting agent.
Introduction
Tumor vasculature represents an attractive target for cancer
therapy, as solid tumors require blood vessels for growth and
metastasis.[1] The antivascular approach, which exploits differ-
ences between normal and immature tumor blood vessels,
aims to selectively damage the newly established tumor vascu-
lature.[2] To date, vascular-disrupting agents (VDAs) are essen-
tially small molecules that inhibit tubulin polymerization.[3] Mi-
crotubules are dynamic polymers of a and b tubulin that play
a crucial role in several cellular processes, including cell divi-
sion, cell signaling, motility, transport, and maintenance of cell
shape.[4] Among VDAs that perturb microtubule dynamics,
combretastatin A4 (CA4) is the most promising compound.[5]
Early works showed that the natural product CA4 strongly
inhibits the proliferation of a broad spectrum of human cancer
cells and acts as a tubulin polymerization inhibitor by binding
at the colchicine site. The water-soluble phosphate prodrug
CA4P, selected for further preclinical development, has drawn
significant attention due to its potent and selective effect on
tumor blood vessels, causing rapid vascular shutdown, leading
to necrosis of the central tumor, an area that is often resistant
to conventional anticancer treatments. The prodrug CA4P (for-
sbretabulin) is currently being evaluated, either alone or in
combination with paclitaxel or carboplatin, in clinical studies
for the treatment of patients with anaplastic thyroid carcino-
ma[6] or advanced cancer.[7,8]
to macromolecules, leading to blood flow shutdown in vivo.
The consequent decrease in nutrient supply induces necrosis
in the tumor center, with a viable rim remaining at the periph-
ery.[2,11]
The encouraging antivascular and antitumor profile of CA4
has contributed greatly to the current interest in the design
and synthesis of several CA4 analogues.[12] Through SAR stud-
ies, it has been established that the cis orientation of both
phenyl groups is an essential requirement for efficient tubulin
affinity, forcing the two aromatic rings to stay within an appro-
[a] Dr. M. Arthuis, Dr. R. Pontikis, Dr. J.-C. Florent
Institut Curie, Centre de Recherche
26 rue d’Ulm, Paris 75248 Cedex 05 (France)
Fax: (+33)156246631
E-mail: renee.pontikis@curie.fr
jean-claude.florent@curie.fr
[b] Dr. M. Arthuis, Dr. R. Pontikis, Dr. J.-C. Florent
CNRS, UMR 176, 26 rue d’Ulm, Paris 75248 Cedex 05 (France)
[c] Dr. G. G. Chabot, J. Seguin, L. Quentin
Universitꢀ Paris Descartes
Facultꢀ des Sciences Pharmaceutiques et Biologiques
4 avenue de l’Observatoire, 75006 Paris (France)
E-mail: guy.chabot@parisdescartes.fr
[d] Dr. G. G. Chabot, J. Seguin, L. Quentin
Laboratoire de Pharmacologie Chimique
Gꢀnꢀtique et Imagerie, INSERM U 1022, CNRS, UMR 8151
4 avenue de l’Observatoire, 75006 Paris (France)
By binding to tubulin in tumor endothelial cells, VDAs such
as CA4 cause rapid microtubule depolymerization. This
damage triggers disruption of the cell–cell junction involving
the protein vascular endothelial cadherin[9] and further cytoske-
letal rearrangements (assembly of actin stress fibers) through
activation of the Rho/Rho kinase pathway.[2a,10] The net result
of these effects is a rounding up and membrane blebbing of
endothelial cells, together with increased vessel permeability
[e] Dr. S. Bourg
Fꢀdꢀration de Recherche Physique et Chimie du Vivant
Universitꢀ d’Orlꢀans, CNRS, FR 2708
rue Charles Sadron, 45071 Orlꢀans Cedex 2 (France)
[f] Dr. L. Morin-Allory
ICOA–Universitꢀ d’Orlꢀans
CNRS, UMR 6005, BP 6759, 45067 Orlꢀans Cedex 2 (France)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201100206.
ChemMedChem 2011, 6, 1693 – 1705
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1693

R. Pontikis, G. G. Chabot, J.-C. Florent, et al.
MED
priate distance in the colchicine binding site. In fact, the cis
double bond of CA4 or analogues easily undergoes isomeriza-
tion, leading to trans isomers that display dramatically de-
creased inhibition of cancer cell growth and tubulin assem-
bly.[12]
Many linkers that constrain the two phenyl rings to a similar
cis-restricted conformation have been reported to increase the
activity and stability of the designed compounds relative to
CA4. This approach was mainly achieved by replacement of
the double bond with other rigid linkers[13] or different cyclic
moieties,[14] or by introducing another ring.[15] These considera-
tions led us to design a new series of conformationally restrict-
ed CA4 analogues by inserting an additional ring between the
cis-olefinic bond and one of the aromatic moieties of the CA4
structure.[16]
Figure 2. Retrosynthetic approach to the stereoselective preparation of
fused heterocyclic systems containing an exocyclic tri- or tetrasubstituted
double bond, by palladium-catalyzed domino reactions.
CꢀH activation, and CꢀC bond formation, from anilides and
aryl iodides as the starting materials (Path B).[19] We therefore
decided to investigate the preparation of heterocyclic CA4 ana-
logues containing a stereochemically defined substituted exo-
cyclic alkene, by using these palladium-catalyzed domino pro-
cesses.[20]
Herein we report the synthesis of a broad range of heterocy-
clic CA4 analogues with a five- or six-membered core of a dif-
ferent nature (lactam or lactone) fused to the trimethoxyphen-
yl ring or the B-phenyl moiety. We also extended the SAR of
this series by examining various substitutions on the B-ring.
Dienic compounds were also investigated as analogues of the
diarylbutadiene 1 that we previously reported as a potent tu-
bulin polymerization inhibitor (Figure 1).[17] The synthesized
compounds were then evaluated in vitro for their capacity to
act as potential VDAs. Moreover, the tubulin binding mode of
selected derivatives is proposed by molecular docking investi-
gations.
We first planned to synthesize CA4 analogues with a hetero-
cycle fused to the B-ring by using the tandem Heck carbocycli-
zation/Suzuki cross-coupling process (Scheme 1). As we previ-
ously reported, under the optimized reaction conditions
[Pd(OAc)2, PPh3, CsF] this tandem reaction occurred with the
N-protected propynamide 6c and the N-free butynamides 6a
and 6b as starting materials.[21] It should be pointed out that
the substitution of the newly formed double bond by a methyl
group (R2), in preventing isomerization of the double bond,
stabilizes the coupling products and furthermore allowed a
domino reaction with secondary amide substrates (X=NH).
Thus, the (E)-3-arylmethyleneoxindoles 9b, 9c and the (E,E)-3-
alkylideneoxindoles 10a, 10c were efficiently obtained, in a
stereoselective manner, from the 2-iodoanilides 6a–c and the
3,4,5-trimethoxyphenyl boronic acid 7 or the corresponding
styryl derivative 8, respectively. The alkyne esters 6d and 6e
were similarly treated to afford the corresponding lactones 9e
and 10d. These domino reactions required the use of the 2-io-
doaryl derivatives 6a–e, which were derived from 2-iodoani-
lines 2 and 3 or 2-iodophenols 4 and 5 by a carbodiimide-
mediated coupling with propynoic or butynoic acids.
Figure 1. Structures of CA4, the prodrug CA4P, and vinylogous analogue 1.
Results and Discussion
To access such heterocyclic derivatives, it seemed advanta-
geous to use a domino process involving an arene CꢀH activa-
Chemistry
The stereoselective preparation
of fused heterocyclic systems
bearing an exocyclic tri- or tetra-
substituted double bond has
been reported, by the use of pal-
ladium-catalyzed domino reac-
tions (Figure 2). One of the more
reliable synthetic pathways is
the
Heck
carbocyclization/
Suzuki–Miyaura coupling proto-
col (Path A) that uses alkyne-
tethered halogenated arenes
and boronic acids as coupling Scheme 1. Synthesis of B-ring-constrained CA4 analogues via tandem Heck–Suzuki–Miyaura reaction. Reagents
partners.[18] More recently, com-
pounds with an oxindole moiety
were successfully prepared by a
and conditions: a) 2, 3, 4, or 5, CH3ꢀCꢁCꢀCO2H, DCC, DMAP for 6d and 6e, CH2Cl2, 08C!RT, 3–18 h, 36–96%;
b) 1. 2, HꢀCꢁCꢀCO2H, DCC, CH2Cl2, 08C!RT, 3 h, 58%; 2. NaH, THF, 08C, 1 h, then BnBr, TBAB, RT, 2 h, 77%;
c) Coupling conditions A: boronic acid 7 or 8 (1.1 equiv), Pd(OAc)2 (5 mol%), PPh3 (10 mol%), CsF (3 equiv) in THF,
reflux, 1–3 h, 47–80%; for precursor 6e and boronic acid 8: RT, 2 h, 47%. [a] Presence of the Z isomer (EE/EZ:
sequence of carbopalladation, 97:3).
1694
www.chemmedchem.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1693 – 1705

Constrained Heterocyclic CA4 Analogues
tion that requires readily available non-halogenated aromatic
precursors. Zhu and co-workers reported the synthesis of 3-(di-
arylmethylene)oxindoles from N-methyl- or N-trimethylsilyle-
thoxymethyl (SEM)-3-arylpropynamides (Figure 2, path B, R=
phenyl) and aryl iodides as the other coupling partner.[19] Inter-
estingly, under the reported reaction conditions [Pd(OAc)2,
NaOAc, DMF], the methylacetylene derivative 12 proved to be
a suitable partner (Scheme 2). After heating at 808C for 2 h in
the presence of 3,4,5-trimethoxyphenyl iodide, the N-methylox-
indole 9 f was obtained stereoselectively, albeit in moderate
yield (46%).[22]
ported above (808C, 2 h), the expected oxindole 17 was ob-
tained, but in low yield (23%). Nevertheless, running the reac-
tion at reflux briefly (5 min) increased the yield to 46%.[26]
We next turned our attention to the synthesis of CA4 ana-
logues with a six-membered heterocycle fused to the trime-
thoxyphenyl ring: isoquinolinone and isochromanone deriva-
tives with a multi-substituted exo double bond. Unfortunately,
ester 18 failed to react following the direct arylation process
(Scheme 3). Instead, the Michael addition product of phenol
16 to the activated triple bond was recovered (ether 19). How-
ever, the desired isochroman-3-ones 21a–d and 22 were ob-
tained under the Heck–Suzuki coupling protocol from ester 20
(X=O) using several boronic acids as the trapping agents.[27]
Isoquinolin-3-ones 23a–f and 24 were prepared in the same
manner from the appropriate alkynamides 20 (X=NH, NR).
Access to the isomeric isochroman-1-one series was also at-
tempted (Scheme 3). Under the same procedure, reaction of
the benzoate 25, obtained from the known benzaldehyde
26,[28] with the 4-methoxyphenylboronic acid afforded the de-
sired lactone 27, but along with the corresponding direct cou-
pling product (ratio 6:4). Despite different reaction conditions,
the styryl and more electron-rich boronic acids failed to pro-
duce the expected products. The configuration of the newly
formed double bond of all the synthesized compounds was
unambiguously supported by NOESY experiments.[21,27]
Scheme 2. Synthesis of oxindoles 9 f and 17 via hydroarylation. Reagents
and conditions: a) CH3ꢀCꢁCꢀCO2H, DCC, CH2Cl2, 08C!RT, 3 h, 77% for 12,
88% for 14; b) Coupling conditions B: ArꢀI (1.15 equiv), Pd(OAc)2
(0.05 equiv), NaOAc (2 equiv), DMF; 808C, 2 h for 9 f (46%); reflux, 5 min for
17 (46%); c) NaH, THF, 08C, 30 min, then SEMCl, 08C!RT, 65 h, 75%.
Biological evaluation
The potential antivascular effects of the synthesized com-
pounds were assessed by their ability to inhibit tubulin poly-
merization (fluorimetric assay)[29] and to rapidly induce a
rounding up of endothelial cells.[30] This morphological test is
considered to be predictive of potential in vivo antivascular ac-
tivity, and it allows determination of the lowest active concen-
tration that causes rounding of immortalized HUVEC (EAhy 926
cells) within 2 h. Compounds were also evaluated for their
growth inhibitory effects against two murine cancer cell lines:
B16 melanoma cells and Lewis lung carcinoma (3LL) cells.
The biological activities of the B-ring-constrained CA4 ana-
logues, with CA4 and diarylbutadiene 1 as reference com-
Encouraged by these results, we envisioned extension of the
same methodology to the synthesis of CA4 analogues bearing
a hetero-ring annulated to the trimethoxyphenyl moiety. We
first studied the oxindole series. Because this reaction proceeds
well with a tertiary amide, we selected an acetylenic precursor
with a removable N-protecting group, the N-SEM-butynamide
15 (Scheme 2).[23,24] With 5-iodo-2-methoxyphenol 16[25] as the
other coupling partner, and under the same conditions as re-
Scheme 3. Synthesis of A-ring-constrained CA4 analogues.[a] Reagents and conditions: a) 18, coupling conditions B, 708C, 20 min; b) 20, ArꢀB(OH)2 or 8, cou-
pling conditions A, 1–10 h, 29–92%;[27] c) KMnO4, H2O, 758C, 5 h, 84%; d) butynol, DCC, DMAP, CH2Cl2, RT, 65 h, 67%; e) 4-methoxyphenylboronic acid, cou-
pling conditions A, 30 h, 36%. [a] See Table 2 for a complete description of X, R1, and R2 groups.
ChemMedChem 2011, 6, 1693 – 1705
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1695

R. Pontikis, G. G. Chabot, J.-C. Florent, et al.
MED
analogues of the dienic deriva-
Table 1. Inhibition of tubulin polymerization, morphological effects on endothelial cells, and cytotoxicity of B-
ring-constrained CA4 analogues.
tive 1.[34] The N-benzyloxindole
10c (a vinylogous analogue of
9c) presented the same biologi-
cal profile as 9c; that is, both
failed to interact with tubulin
and displayed a moderate effect
on the proliferation of 3LL cells
ITP
IC50 [mm][a]
Morphology
HUVEC [mm][b]
Cytotoxicity IC50 [mm][c]
(IC50 =5–6 mm).
Unfortunately,
Compd
X
R2
R1
3LL
B16
with regard to the potent com-
pound 9b, such a similarity was
not observed with the stabilized
N-free oxindole 10a, which
showed a significant loss of po-
tency. Surprisingly, its isomer,
9c
9g
9b
9 f
N-Bn
NH
NH
N-CH3
O
N-Bn
NH
H
H
H
OCH3
OCH3
20
>20
0.6
14
6.7
19
5.6
2.2
13
10
>10
0.05
0.3
5.3
19
0.09
0.79
>50
9.4
32
1.1
16
CH3
CH3
CH3
CH3
H
CH3
CH3
CH3
9e
OCH3
na[d]
na
>50
10c
10c
trans-10a
10d
CA4
1
–
–
–
–
5.7
8.4
3.1
22
>50
12
1.3
5
na
0.006
0.10
with
a
trans,trans-butadienic
NH
O
linker (trans-10a)[35] displayed a
slightly better inhibition of tubu-
lin polymerization. In this dienic
series a benzofuran core was
also unsuitable for activity (com-
pound 10d).
>50
>50
0.27
0.56
0.004
nd[e]
0.005
0.13
[a] Compound concentration required to inhibit tubulin polymerization (ITP) by 50%; in vitro microtubule as-
sembly was monitored using a fluorescent probe (DAPI).[29] [b] Morphological activity on immortalized HUVEC
(EAhy 926 cells) is expressed as the lowest concentration at which cell rounding was observed following incu-
bation with the tested drug for 2 h. [c] Compound concentration required to inhibit murine tumor cell prolifer-
ation by 50% after an incubation time of 48 h; 3LL: Lewis lung carcinoma; B16: melanoma. [d] Not active at
maximum solubility in cell culture medium. [e] Not determined.
Table 2 lists the biological ac-
tivity values of the synthesized
compounds bearing a heterocy-
clic moiety fused to the trime-
pounds, are presented in Table 1. The N-benzyloxindole 9c
does not markedly inhibit tubulin polymerization and was also
shown to gradually isomerize in solution to the (Z)-olefinic
form. This instability was previously observed, by our research
group and others, with similar oxindole derivatives bearing a
methine bridge.[16c,d,21] Nevertheless, as we previously reported,
substitution of the exocyclic double bond by a methyl group
leads to chemically stable compounds. Therefore, only deriva-
tives with a methyl-substituted alkene were further investigat-
ed.
The stabilized N-free oxindole 9g[31] was also devoid of bio-
logical activity. Interestingly, by analogy with CA4, substitution
at the C6 position of the oxindole core with a methoxy group
(compound 9b) considerably enhanced biological activity.
Compound 9b showed an inhibition of tubulin polymerization
(IC50 =0.6 mm) within the same order of magnitude as that of
CA4 (IC50 =0.27 mm) and clearly induced rapid changes in en-
dothelial cell morphology at low concentration (rounding up
at 50 nm).[32] Antiproliferative activity was also increased, but
did not reach the IC50 values of CA4 (IC50 =0.09 and 1.1 mm
versus 0.005 for CA4). However, methylation of the oxindole
nitrogen (compound 9 f) led to significant loss of anti-tubulin
activity,[33] although a morphological effect on endothelial cells
was observed at 0.3 mm. Replacement of the methoxyoxindole
ring by a methoxybenzofuran ring (compound 9e) had a detri-
mental effect on activity. Considering the above results, it ap-
pears that the key structural factors for potent constrained de-
rivatives bearing an olefinic linker are a methoxy group and a
free nitrogen atom on the heterocyclic ring.
thoxyphenyl ring. The N-SEM-oxindole derivative 17 is devoid
of any appreciable anti-tubulin and antiproliferative activities.
All the other products studied presented a larger six-mem-
bered lactam or lactone ring, therefore offering the possibility
to adopt a more suitable conformation to interact with tubu-
lin.
To estimate the effect of the position of the ester function
inside the ring on biological activity, the isomeric isochroma-
nones 21a and 27 were compared. Whereas compound 27
with the carbonyl group in the benzylic position was found in-
active, the 3-isochromanone 21a affected tubulin polymeri-
zation (3.9 mm), endothelial cell morphology, and proliferative
activity. Therefore, several 3-isochromanones variously substi-
tuted on the B-ring were prepared and evaluated. Replace-
ment of the 4-methoxy substituent with the lipophilic trifluoro-
methyl group (compound 21b) proved detrimental to activity.
However, the dimethylamino derivative 21c and compound
21d, which possesses the structural feature of CA4, displayed
a substantial ability to inhibit tubulin polymerization and pro-
duced a potent morphological effect at sub-micromolar con-
centrations. These three interesting isochromanones displayed
modest antiproliferative activity in the micromolar range. Sev-
eral isoquinolinones with a free or alkylated nitrogen atom and
a variously substituted B-ring were then evaluated. However,
these compounds (23a–f) were essentially devoid of activity. In
summary, for the A-ring-constrained CA4 analogues investigat-
ed, the dimethylaniline and methoxyphenol derivatives 21c
and 21d proved to be the best tubulin polymerization inhibi-
tors, with IC50 values (1.4 and 1.8 mm) slightly lower than that
of CA4.
Compounds 10, which did not bear the methoxy group es-
sential for activity in the previous series, were investigated as
1696
www.chemmedchem.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1693 – 1705

Constrained Heterocyclic CA4 Analogues
thelial cell morphology is a dis-
tinctive feature of the activity of
VDAs and is usually associated
with microtubule network disor-
ganization. Indeed, alteration of
endothelial cell morphology by
compounds 9b, 21c, 21d, and
22 is well correlated with their
ability to effectively inhibit tubu-
lin polymerization. However, cor-
relation between those assays
did not hold true for derivative
9 f, for which disruption of the
endothelial cell network could
result from alteration of the or-
ganization of other cytoskeletal
components.[9,30] It is noteworthy
that the desmethyl analogue of
Table 2. Inhibition of tubulin polymerization, morphological effects on endothelial cells, and cytotoxicity of A-
ring-constrained CA4 analogues.
ITP
IC50 [mm][a]
Morphology
HUVEC [mm][b]
Cytotoxicity IC50 [mm][c]
Compd
X
R1
R2
3LL
B16
17
27
–
–
O
O
O
O
NH
N-Bn
N-CH3
N-CH3
N-CH3
N-CH3
O
–
–
–
–
>20
nd[d]
3.9
>20
1.4
25
>30
1.3
19
39
1.9
12
1.8
1.0
20
15
23
28
23
25
0.9
55
1.1
13
0.85
1.2
nd
nd
nd
nd
nd
nd
3.1
nd
0.004
nd
21a
21b
21c
21d
23a
23b
23c
23d
23e
23 f
22
OCH3
CF3
H
H
H
OH
H
H
H
H
H
OH
–
12.5
0.65
0.65
12.5
>30
>30
>30
25
N(CH3)2
OCH3
OCH3
OCH3
OCH3
CF3
N(CH3)2
OCH3
–
1.8
>20
>20
>20
>20
>20
>20
1.0
12
0.27
0.56
9 f, which does not bear
a
methyl group on the olefinic
bond, was recently reported to
display not only a vascular-dis-
rupting effect, but also anti-an-
giogenic activities.[36]
25
0.12
12.5
0.006
0.10
24
CA4
1
N-CH3
–
–
>50
0.005
0.13
[a] Compound concentration required to inhibit tubulin polymerization (ITP) by 50%; in vitro microtubule as-
sembly was monitored using a fluorescent probe (DAPI).[29] [b] Morphological activity on immortalized HUVEC
(EAhy 926 cells) is expressed as the lowest concentration at which cell rounding was observed following incu-
bation with the tested drug for 2 h. [c] Compound concentration required to inhibit murine tumor cell prolifer-
ation by 50% after an incubation time of 48 h; 3LL: Lewis lung carcinoma; B16: melanoma. [d] Not deter-
mined.
The effect of these promising
compounds was next evaluated
on the proliferation of EAhy 926
endothelial cells. After incuba-
tion for 48 h, compounds 9b,
21c, 21d, and 22 exhibited sig-
nificant inhibition of endothelial
cell proliferation (Table 3). How-
With dienic derivatives, the lactam ring was also detrimental
to biological activity (compound 24). However, lactone 22 po-
tently inhibited tubulin polymerization (IC50 =1.0 mm) and
caused rounding up of endothelial cells (0.12 mm) at a concen-
tration lower than the cytotoxic IC50 values (micromolar range).
Notably, the constrained com-
ever, it should be noted that these compounds were not cyto-
toxic (IC50 >50 mm) following an incubation time of 2 h (experi-
mental conditions for the morphological test). These results in-
dicate that the morphological effects observed are not caused
by antiproliferative activity within the 2 h incubation period.
pound 22 displayed a similar
biological activity profile as that
of diene 1.
Altogether, these results show
that some constrained deriva-
tives exhibited interesting bio-
logical properties; these include
oxindoles 9b and 9 f, and the
isoquinolones 21c, 21d, and 22
(Table 3). These five compounds
are able to cause endothelial cell
rounding at low micromolar con-
centrations, with the B-ring-con-
strained derivative 9b still affect-
ing cellular shape at 50 nm. Rep-
resentative micrographs of the
Figure 3. Morphological effects of selected CA4 analogues on EAhy 926 endothelial cells. Cells in exponential
growth were exposed to the indicated compound for 2 h at the following concentration: control, 1% DMSO; CA4,
0.006 mm; 9b, 0.10 mm; 21c, 0.65 mm; 21d, 0.65 mm; 22, 0.16 mm. Representative micrographs were taken at
rounding up effects of some of
these compounds are presented
in Figure 3. Alteration of endo- 360ꢂ magnification.
ChemMedChem 2011, 6, 1693 – 1705
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1697

R. Pontikis, G. G. Chabot, J.-C. Florent, et al.
MED
Molecular docking
Table 3. Biological activities of representative constrained derivatives.
To determine the possible bind-
ing modes of this series of com-
pounds with tubulin, docking
studies were carried out with
the most potent compounds:
oxindole 9b and the A-ring-con-
strained derivatives 21c, 21d,
and 22, by using the reported
high-resolution crystal structure
of the tubulin–DAMA-colchicine
complex (PDB ID: 1SA0).[39] The
colchicine binding site is mostly
buried in the b subunit, bor-
dered by the S8 and S9 strands,
the T7 loop, and the H7 and H8
helices. The ligand also main-
tains few interactions with the
T5 loop of the neighboring a su-
bunit. Several important residues
for binding of colchicine-type li-
gands to tubulin have been
identified, including Cysb241,
ITP
Morphology
Cytotoxicity IC50 [mm]
Compd
IC50 [mm][a]
HUVEC [mm][b]
3LL[c]
B16[c]
1.1
16
1.8
1.0
0.9
0.004
0.13
HCT116[d] MCF7
Endothelial
cells[f]
IC1/2max [mm][d,e]
9b
9 f
21c
21d
22
0.6
14
0.05
0.3
0.09
0.79
0.85
1.2
3.1
0.005
nd
0.10
nd[g]
0.21
nd
0.039
0.008
nd
0.9
nd
0.51
nd
1.4
1.8
1.0
0.27
0.56
0.65
0.65
0.12
0.006
0.10
1.0
0.10
0.001
0.50[17]
0.079
0.001
nd
0.17
0.005
0.3
CA4
1
[a] Compound concentration required to inhibit tubulin polymerization (ITP) by 50%; in vitro microtubule as-
sembly was monitored using a fluorescent probe (DAPI).[29] [b] Morphological activity on immortalized HUVEC
(EAhy 926 cells) is expressed as the lowest concentration at which cell rounding was observed following incu-
bation with the tested drug for 2 h. [c] Compound concentration required to inhibit murine tumor cell prolifer-
ation by 50% after an incubation time of 48 h; 3LL: Lewis lung carcinoma; B16: melanoma. [d] Compound con-
centration required to inhibit human tumor cell proliferation by 50% after 72 h incubation. [e] IC50 values could
not be determined with the MCF7 cell line, as growth inhibition did not exceed 60%; IC1/2max is the half-maxi-
mal inhibition concentration. [f] Compound concentration required to inhibit immortalized endothelial cell
(EAhy 926) proliferation by 50% after 48 h incubation; following incubation for 2 h, IC50 >50 mm. [g] Not deter-
mined.
Metb259,
Val a181.[40]
Thra179,
and
Docking simulations were per-
formed with GOLD software ver-
sion 4.1 with default parameters
We next evaluated the ability of compounds 9b, 21c, 21d,
and 22 to disrupt a network of capillary-like tubular structures
formed by endothelial cells seeded on a Matrigel matrix.[37]
Their effects on the preformed cords are depicted in Figure 4
and are compared with CA4. Over the course of 2.5 h, these
constrained CA4 analogues can disrupt the tube-like structures
at concentrations of 1 mm (CA4, 9b, 9 f, and 22) and 10 mm (1,
21c, and 21d). These effects were observed at a concentration
previously shown non-cytotoxic for endothelial cells after a
short exposure time.
and without constraints on the ligand.[41] With such parame-
ters, GOLD generates poses until it finds three with RMSD<1.5
ꢃ. GoldScore was used as the scoring function to rank the
poses. To validate our docking process, DAMA-colchicine was
first extracted from the complex then re-docked. The three
top-scored conformers obtained superimposed well with the
original crystallized structure. CA4 was found to adopt a very
similar binding mode, the trimethoxyphenyl ring lies in a hy-
drophobic pocket close to Cysb241, and the 4’-methoxy group
occupies the same position as the corresponding group on
ring C of colchicine. As previously reported, CA4 interacts ex-
tensively with b-tubulin and with only two residues in a-tubu-
lin (Thr179 and Val181).[40,42]
To further assess the cytotoxic profile of oxindole 9b and
lactones 21c and 22, we investigated their effects against two
human cancer cell lines: HCT116 colon carcinoma cells and
MCF7 hormone-dependent breast carcinoma cells. Against
HCT116, these three compounds displayed antiproliferative ac-
tivities of 0.1–0.2 mm (Table 3). For the MCF7 cell line, IC50
values could not be determined, as growth inhibition did not
exceed 60%, even for CA4 (Supporting Information, table 1).
Whereas the reference CA4 inhibited the proliferation of the
four tumor cell lines with IC50 values in the nanomolar range,
our compounds exhibited good to moderate cytotoxicity de-
pending on the tumor cell type (IC50 values ranging from 0.1
to 1–3 mm). However, these compounds displayed significantly
lower cytotoxicity than CA4, while they showed a potent abili-
ty to inhibit the assembly of tubulin. In fact, this discrepancy
between cytotoxic activity and anti-tubulin activity represents
an interesting goal in the search for new antivascular
agents.[38]
Our docking process applied to the oxindole 9b identifies
two potential binding modes. In the best-scored pose, 9b
overlays correctly with CA4 (Figure 5A). The 3- and 4-methoxy
oxygen atoms of the TMP ring are involved in hydrogen bonds
with the thiol group of Cysb241, and the methoxy group of
the oxindole core establishes hydrophobic contacts with the
side chain of Metb259. Hydrophobic interactions are also ob-
served between the methyl group of the bridge and residues
Leu248 and Ala250 of the b-tubulin subunit. Furthermore, the
indole NH group forms a stable hydrogen bond (1.6 ꢃ) with
the amide oxygen atom of Asnb258.[43–45] This could explain
the observed good biological results for 9b and the loss of ac-
tivity for the N-methyl analogue 9 f, or for the derivative 9e,
which bears an oxygen atom in place of a hydrogen bond
donor atom. Furthermore, residues Leu248 and Ala250 of the
1698
www.chemmedchem.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1693 – 1705

Constrained Heterocyclic CA4 Analogues
Conclusions
We designed and prepared a set of conformationally restricted
CA4 analogues with a lactam or lactone ring fused to the tri-
methoxyphenyl or the B-phenyl moieties of CA4. Another
series included compounds with an (E,Z)-dienic linker. Access
to these compounds was efficiently achieved in a stereoselec-
tive manner by using a domino Heck–Suzuki–Miyaura coupling
reaction. Oxindole derivatives could be obtained via CꢀH aryla-
tion; however, the sequence could not be extended to com-
pounds with a lactam ring.
The vascular-disrupting potential of these molecules was as-
sessed with various in vitro assays. For the B-ring-constrained
derivatives with an olefinic linker, a methoxy group and a free
nitrogen atom on the heterocyclic moiety are required struc-
tural factors for potent activity. Although the oxindole 9b is
endowed with moderate antiproliferative activity, its ability to
disorganize the microtubule network is similar to that of the
reference CA4 and is well supported by the molecular model-
ing studies. In contrast, this structural modification is detrimen-
tal in the dienic series.
Results of the biological evaluation of compounds bearing a
six-membered heterocyclic core fused to the trimethoxyphenyl
ring indicate that a lactone moiety is well tolerated in both
series. The significant anti-tubulin activity displayed by the 3-
isochromanones 21c, 21d, and 22 indicates that structural
modification carried out on this A-part of the CA4 molecule, al-
though rarely considered so far, is a promising way to provide
access to novel antivascular compounds.
Figure 4. Effects of selected CA4 analogues on preformed endothelial cell
cords (HUVEC) on Matrigel. Micrographs shown were taken 2.5 h after addi-
tion of the compounds at the following concentration: control, 1% DMSO;
CA4, 1 mm; 9b, 1 mm; 9 f, 1 mm; 21c, 10 mm; 21d, 10 mm; 1, 10 mm; 22, 1 mm.
Original magnification of 40ꢂ.
The most active compound 9b was of particular interest, as
it combines chemical stability and the biological activity profile
characteristic of a vascular-disrupting agent. The promising an-
tivascular potentiality of this compound will be investigated
further with in vivo murine experiments.
b-tubulin subunit are involved in hydrophobic contacts with
the methyl group of the olefinic bridge.
In contrast, in the second binding mode (score value of 43.3,
regarding the score of 45.7 for the best pose), the TMP ring
lies on the edge of the colchicine binding site, pointing
toward the GTP bound to the a-subunit (Figure 5B). In this
case, hydrogen bonds could be established between two me-
thoxy oxygen atoms of the TMP ring, the 3-methoxy group
with the side chain amino group of Asna101 (3.0 ꢃ) and the 5-
methoxy group with the side chain hydroxy of Sera178 (3.0 ꢃ).
Hydrogen bond interactions with Asn101 and Ser178 of a-tu-
bulin are also found to be involved in the binding of potent
tubulin polymerization inhibitors which show an orientation
different from that of DAMA-colchicine in the tubulin binding
Experimental Section
Chemistry
Unless otherwise noted, all materials were obtained from commer-
cial sources and were used without purification. THF was dried
over sodium/benzophenone, CH2Cl2 over P2O5, DMF and MeOH
over molecular sieves. Some cycloadducts have already been de-
scribed by our research group and others: 6a, 6c,[49] 10c,[21] and
21–24.[27] Aniline 3,[50] phenol 5,[51] and aldehyde 26[28] were pre-
1
pared according to published procedures. H NMR spectra were re-
corded with a Bruker ACP 300 spectrometer at 300 MHz for
1H NMR, and at 75 MHz for 13C NMR spectroscopy. Chemical shift
values (d) are given in ppm relative to the residual solvent peak
(CHCl3) as the internal reference, and coupling constants (J) are
given in Hertz. All 13C NMR spectra were recorded with complete
proton decoupling. Peak assignment was unambiguously per-
formed using HMQC, HMBC, and NOESY techniques. Melting
points were determined by the capillary method using an Electro-
thermal 9200 apparatus and are uncorrected. Mass spectra were
recorded on a Waters ZQ 2000 system using electrospray ionization
(ESI). High-resolution ESIMS data were acquired from the “imagif”
service (CNRS-ICSN, 91198 Gif-sur-Yvette, France) on a Waters LCT
spectrometer. Reactions were monitored by thin-layer chromatog-
site (heterocyclic derivatives with
a
trimethoxybenzoyl
moiety[46] or a recently reported compound[47] discovered by
virtual screening).[48] Although the indole ring is positioned dif-
ferently, hydrophobic contacts of the methoxy group with
Metb259 could be maintained, as well as interactions with the
olefinic methyl substituent. The three A-ring-constrained com-
pounds 21c, 21d, and 22 dock in a very similar—albeit quite
unexpected—position, thus indicating less efficient binding to
tubulin (see Supporting Information, figure 1).
ChemMedChem 2011, 6, 1693 – 1705
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1699

R. Pontikis, G. G. Chabot, J.-C. Florent, et al.
MED
Figure 5. Proposed binding mode of CA4 and oxindole 9b in the colchicine binding site of tubulin. A) Superimposition of CA4 (pink) and the best-scored
pose of compound 9b (green). B) Superimposition of the two best-scored poses of compound 9b. Only relevant residues are indicated.
raphy (TLC) with Merck silica gel 60 F254. Flash chromatography
was carried out on Merck silica gel (320–400 mesh). Representative
structures with indicative numbering for NMR assignments are
shown:
Acetylenic derivatives
Preparation of amides 6b, 12, and 14: general procedure: A so-
lution of DCC (1 equiv) in CH2Cl2 was added dropwise at 08C to a
solution of the appropriate aniline and but-2-ynoic acid (1 equiv)
in dry CH2Cl2. After stirring for 3 h at room temperature, the mix-
ture was cooled to 08C and filtered. The residue was washed with
CH2Cl2. The filtrate was washed with 2n HCl, saturated aqueous
NaHCO3, and H2O. The organic layer was dried (MgSO4) and con-
centrated under reduced pressure.
N-(2-Iodo-5-methoxyphenyl)but-2-ynamide (6b): Aniline 3[50]
(375 mg, 1.5 mmol) afforded amide 6b as a white solid (155 mg,
36%) after chromatography (cyclohexane/EtOAc 8:2); mp: 136–
6-Iodo-2,3,4-trimethoxybenzoic acid (28). A solution of KMnO4
(380 mg, 2.40 mmol) in H2O (20 mL) was added dropwise to a sus-
pension of aldehyde 26[28] (500 mg, 1.55 mmol) in H2O (15 mL)
while stirring at 758C.[52] After stirring at 758C for 5 h, the reaction
mixture was cooled to room temperature, brought to pH 12 by ad-
dition of aqueous KOH (20%), then filtered on a pad of Celite. The
filtrate was acidified to pH 2 (2n HCl), and then extracted with
Et2O (50 mLꢂ3). The combined organic layers were dried (MgSO4)
and concentrated under reduced pressure. The residue was recrys-
tallized from CH2Cl2/hexane to give acid 28 as a white solid
(440 mg, 84%); mp: 100–1018C; 1H NMR (300 MHz, CDCl3): d=
8.91 (s, 1H, CO2H), 7.11 (s, 1H, H-5), 3.90 (s, 3H, OCH3), 3.85 ppm (s,
6H, OCH3); 13C NMR (75 MHz, CDCl3): d=171.6 (CO2H), 155.2 (C-4),
151.5 (C-2), 142.4 (C-3), 127.3 (C-1), 118.6 (CH-5), 84.3 (C-6), 62.2
(OCH3), 61.0 (OCH3), 56.4 ppm (OCH3); MS (ES+): m/z: 361 [M+
Na]+; Anal. calcd for C10H11IO5: C 35.52, H 3.28, found: C 35.37, H
3.21.
1
1388C; H NMR (300 MHz, CDCl3): d=7.93 (d, J=2.9 Hz, 1H, H-6),
7.61 (d, J=8.8 Hz, 1H, H-3), 7.60 (brs, 1H, NH), 6.49 (dd, J=8.8,
2.9 Hz, 1H, H-4), 3.79 (s, 3H, OCH3), 2.04 ppm (s, 3H, CH3); 13C NMR
(75 MHz, CDCl3): d=160.6 (C-5), 150.9 (CO), 138.7 (C-1), 138.6 (CH-
3), 113.2 (CH-4), 107.5 (CH-6), 85.2 (C-d), 77.2 (C-2), 75.2 (C-c), 55.5
(OCH3), 3.9 ppm (CH3); MS (ES+): m/z: 338 [M+Na]+; Anal. calcd
for C11H10INO2: C 41.93, H 3.20, N 4.45, found: C 41.89, H 3.18, N
4.33.
N-(3-Methoxyphenyl)-N-methylbut-2-ynamide (12). N-Methylani-
line 11 (500 mg, 3.65 mmol) afforded amide 12 as a colorless oil
(574 mg, 77%) after chromatography (cyclohexane/EtOAc 7:3);
1H NMR (300 MHz, CDCl3): d=7.29 (t, J=8.1 Hz, 1H, H-5), 6.87–
6.81 (m, 3H, H-2, H-4, H-6), 3.82 (s, 3H, OCH3), 3.29 (s, 3H, N-CH3),
1.76 ppm (s, 3H, CH3); 13C NMR (75 MHz, CDCl3): d=160.0 (C-3),
154.2 (CO), 144.3 (C-1), 129.7 (C-5), 119.3 (C-6), 113.2 (C-4), 112.9 (C-
2), 89.8 (C-d), 74.0 (C-c), 55.4 (OCH3), 36.3 (CH3), 3.9 ppm (CH3); MS
(ES+) m/z: 204 [M+H]+, 226 [M+Na]+; HRMS (DCI/CH4): m/z [M+
H]+ calcd for C12H14O2N: 204.1025, found: 204.1021.
N-(2,3,4-Trimethoxyphenyl)but-2-ynamide (14): Trimethoxyaniline
13[23,53] (1.16 g, 6.33 mmol) afforded amide 14 as cottony white
1700
www.chemmedchem.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1693 – 1705

Constrained Heterocyclic CA4 Analogues
crystals (1.40 g, 88%) after recrystallization (EtOH/hexane); mp:
130–1318C; 1H NMR (300 MHz, CDCl3): d=7.96 (d, J=9.0 Hz, 1H,
H-6), 7.82 (s, 1H, NH), 6.62 (d, J=9.0 Hz, 1H, H-5), 3.96 (s, 3H,
OCH3), 3.86 (s, 3H, OCH3), 3.83 (s, 3H, OCH3), 2.08 ppm (s, 3H, CH3);
13C NMR (75 MHz, CDCl3): d=150.7 (CO), 150.1 (C-4), 142.5 (C-2),
141.6 (C-3), 124.9 (C-1), 115.2 (CH-6), 106.9 (CH-5), 84.0 (C-d), 75.6
(C-c), 61.3 (OCH3), 60.9 (OCH3), 56.1 (OCH3), 3.8 ppm (CH3); MS (ES+
) m/z: 250 [M+H]+; Anal. calcd for C13H15NO4: C 62.64, H 6.07, N
5.62, found: C 62.70, H 6.01, N 5.61.
2,3,4-Trimethoxybenzylbut-2-ynoate (18): A solution of DCC
(1.87 g, 9.08 mmol) in dry CH2Cl2 (50 mL) was added dropwise to a
solution of 2,3,4-trimethoxybenzyl alcohol obtained by reduction[54]
of the corresponding benzaldehyde (1.5 g, 7.57 mmol), but-2-ynoic
acid (763 mg, 9.08 mmol), and DMAP (85 mg, 0.75 mmol) in dry
CH2Cl2 (10 mL) at 08C. After stirring for 16 h at room temperature,
the reaction mixture was cooled to 08C. The precipitate was fil-
tered off and eluted with CH2Cl2 (50 mLꢂ2). The filtrate was con-
centrated under reduced pressure, and the residue was subjected
to flash chromatography (cyclohexane/EtOAc 8:2) to give ester 18
N-[(2-(Trimethylsilyl)ethoxy)methyl]-N-(2,3,4-trimethoxyphenyl)-
but-2-ynamide (15): NaH (60% in mineral oil, 138 mg, 3.44 mmol)
was added to a stirred solution of amide 14 (660 mg, 2.65 mmol)
in dry THF (30 mL) at 08C. After 30 min at 08C, SEMCl (780 mL,
3.97 mmol) was added, and the mixture was stirred at room tem-
perature for 65 h before being quenched with saturated aqueous
NaHCO3 and extracted with Et2O (100 mLꢂ2). The combined or-
ganic layers were washed with brine, dried (MgSO4), and concen-
trated under reduced pressure. The residue was purified by flash
chromatography (cyclohexane/EtOAc 75:25) to give amide 15 (two
1
as a colorless oil (1.4 g, 70%); H NMR (300 MHz, CDCl3): d=7.04
(d, J=8.5 Hz, 1H, H-6), 6.64 (d, J=8.5 Hz, 1H, H-5), 5.12, (s, 2H,
CH2), 3.92 (s, 3H, OCH3), 3.86 (s, 3H, OCH3), 3.84 (s, 3H, OCH3),
1.95 ppm (s, 3H, CH3); 13C NMR (75 MHz, CDCl3): d=153.6 (CO),
152.4 (C-4), 152.6 (C-2), 142.1 (C-3), 125.7 (CH-6), 121.0 (C-1), 107.0
(CH-5), 85.6 (C-e), 72.4 (C-d), 63.0 (CH2), 61.4 (OCH3), 60.7 (OCH3),
55.9 (OCH3), 3.7 ppm (CH3); MS (ES+): m/z: 287 [M+Na]+.
6-Iodo-2,3,4-trimethoxybenzylbut-2-ynoate (25): Ester 25, pre-
pared by following the above procedure (reaction time: 65 h) em-
ploying acid 28 (350 mg, 1.03 mmol) and butynol (115 mL,
1.54 mmol) was obtained after flash chromatography (cyclohex-
1
rotamers at a ratio of 7:3) as a colorless oil (745 mg, 75%); H NMR
(300 MHz, CDCl3), major/minor: d=6.89/6.83 (d, J=8.8 Hz, 1H, H-
6), 6.63/6.66 (d, J=8.8 Hz, 1H, H-5), 5.40 (d, J=10.3 Hz, 1H, N-
CHHa), 4.68 (d, J=10.3 Hz, 1H, N-CHHb), 3.90 (s, 3H, OCH3), 3.87 (s,
3H, OCH3), 3.85 (s, 3H, OCH3), 3.59–3.68 (m, 2H, O-CH2), 1.73/2.05
(s, 3H, CH3), 0.95–0.88 (m, 2H, Si-CH2), 0.01 ppm (s, 9H, CH3);
13C NMR (75 MHz, CDCl3), major/minor: d=155.9/155.1 (CO), 154.0/
153.7 (C-4), 150.5/149.7 (C-2), 142.4/142.8 (C-3), 127.4/125.6 (C-1),
124.9/123.7 (CH-6), 106.4/106.9 (CH-5), 89.8/90.1 (C-d), 76.1 (CH2-N),
74.1/73.5 (C-c), 66.2/65.6 (CH2-O), 61.2/61.1 (OCH3), 60.8 (OCH3),
55.9 (OCH3), 18.1/17.9 (CH2-Si), 3.8/4.1 (CH3), ꢀ1.4 ppm (CH3); MS
(ES+) m/z: 380 [M+H]+, 402 [M+Na]+; HRMS (DCI/CH4) m/z [M+
H]+ calcd for C19H30O5NSi: 380.1893, found: 380.1899.
1
ane/EtOAc 8:2) as a colorless oil (271 mg, 67%); H NMR (300 MHz,
CDCl3): d=7.05 (s, 1H, H-5), 4.88 (q, J8,11 =2.4 Hz, 2H, O- CH2), 3.90
(s, 3H, OCH3), 3.85 (s, 3H, OCH3), 3.84 (s, 3H, OCH3), 1.87 ppm (t,
J
8,11 =2.4 Hz, 3H, CH3); 13C NMR (75 MHz, CDCl3): d=166.4 (CO),
155.1 (C-4), 151.5 (C-2), 142.4 (C-3), 127.9 (C-1), 118.1 (CH-5), 84.0
(C-6), 83.8 (C-e), 72.7 (C-d), 54.2 (CH2), 61.9 (OCH3), 60.8 (OCH3),
56.4 (OCH3), 3.7 ppm (CH3); MS (ES+): m/z: 413 [M+Na]+; HRMS
(DCI/CH4) m/z [M+H]+ calcd for C14H16O5I: 391.0043, found:
391.0045.
General procedure for the tandem Heck–Suzuki–Miyaura reac-
tion (path A).[21,27] A flask, flame-dried under high vacuum, was
charged with the appropriate alkyne (1 equiv), boronic acid
(1.1 equiv), CsF (3.3 equiv, flame-dried under high vacuum prior to
use) and purged with argon three times. Anhydrous THF (0.03m)
was added, and the resulting mixture was degassed (bubbling
argon, 15 min) before Pd(OAc)2 (5 mol%) and PPh3 (10 mol%) were
added. After heating at reflux under argon for the appropriate
time, the reaction was cooled to room temperature, quenched
with H2O, and extracted twice with Et2O. The combined organic
layers were washed with H2O, brine, dried over MgSO4, concentrat-
ed under reduced pressure, and purified by flash chromatography
(cyclohexane/EtOAc 8:2 or 6:4) to provide the expected cycload-
ducts.
2-Iodophenylbut-2-ynoate (6d):
2.27 mmol) in dry CH2Cl2 (15 mL) was added dropwise to a solution
of phenol (500 mg, 2.27 mmol), but-2-ynoic acid (186 mg,
A solution of DCC (467 mg,
4
2.27 mmol) and DMAP (25 mg, 0.22 mmol) in dry CH2Cl2 (10 mL) at
08C. After stirring at room temperature for 3 h, the mixture was
cooled to 08C and filtered. The filtrate was then washed with 2n
HCl, saturated aqueous NaHCO3, and H2O. The organic layer was
dried (MgSO4) and concentrated under reduced pressure. The resi-
due was purified by flash chromatography (cyclohexane/EtOAc
95:5) to give ester 6d as a colorless oil (435 mg, 67%); 1H NMR
(300 MHz, CDCl3): d=7.83 (d, J=7.7 Hz, 1H, H-3), 7.37 (t, J=
7.7 Hz, 1H, H-5), 7.12 (d, J=7.7 Hz, 1H, H-6), 6.99 (t, J=7.7 Hz, 1H,
H-4), 2.08 ppm (s, 3H, CH3); 13C NMR (75 MHz, CDCl3): d=151.0
(CO), 150.5 (C-1), 139.6 (CH-3), 129.5 (CH-5), 128.1 (CH-4), 122.9
(CH-6), 90.0 (C-2), 88.9 (C-d), 71.9 (C-c), 4.1 ppm (CH3); MS (ES+) m/
z: 287 [M+H]+, 309 [M+Na]+; HRMS (DCI/CH4) m/z [M+H]+ calcd
for C10H8O2I: 286.9569, found: 286.9563.
(E)-6-Methoxy-3-(1-(3,4,5-trimethoxyphenyl)ethylidene)indolin-2-
one (9b): Oxindole 9b was obtained from alkyne 6b (44 mg,
0.14 mmol) and boronic acid 7 as pale-yellow crystals (39 mg,
80%); mp: 192–1938C; 1H NMR (300 MHz, CDCl3): d=7.90 (brs,
1H, NH), 6.50 (s, 2H, H-2, H-6), 6.37 (d, J=1.9 Hz, 1H, H-7’), 6.22
(m, 2H, H-4’, H-5’), 3.92 (s, 3H, OCH3), 3.83 (s, 6H, OCH3), 3.74 (s,
3H, OCH3), 2.74 ppm (s, 3H, CH3); 13C NMR (75 MHz, CDCl3): d=
170.4 (C-2’), 160.1 (C-6’), 153.8 (C-3, C-5), 151.8 (C-a), 139.1 (C-7’a),
138.5 (C-4), 137.7 (C-1), 124.3 (C-4’), 123.2 (C-3’), 116.2 (C-3’a), 106.3
(C-5’), 103.4 (CH-2, CH-6), 96.1 (CH-7’), 61.1 (OCH3), 56.2 (OCH3 ꢂ2),
55.4 (OCH3), 22.4 ppm (CH3); MS (ES+): m/z: 378 [M+Na]+; HRMS
(ESI) m/z [M+H]+ calcd for C20H22O5N: 356.1498, found: 356.1490.
2-Iodo-5-methoxyphenylbut-2-ynoate (6e): Ester 6e, prepared by
following the above procedure employing phenol 5 (500 mg,
2 mmol) and but-2-ynoic acid (168 mg, 2 mmol), was obtained
after recrystallization (CH2Cl2/hexane) as white crystals (610 mg,
96%); mp: 64–668C; 1H NMR (300 MHz, CDCl3): d=7.66 (d, J=
8.9 Hz, 1H, H-3), 6.70 (d, J=2.8 Hz, 1H, H-6), 6.61 (dd, J=8.8,
2.8 Hz, 1H, H-4), 3.77 (s, 3H, OCH3), 2.08 ppm (s, 3H, CH3); 13C NMR
(75 MHz, CDCl3): d=160.9 (C-5), 151.2 (C-1), 150.9 (CO), 139.3 (CH-
3), 114.6 (CH-4), 109.1 (CH-6), 88.9 (C-d), 78.3 (C-2), 72.0 (C-c), 56.6
(OCH3), 4.1 ppm (CH3); MS (ES+) m/z: 317 [M+H]+; Anal. calcd for
C11H9IO3: C 41.80, H 2.87, found: C 42.36, H 3.09.
(E)-1-Benzyl3-(3,4,5-trimethoxybenzylidene)indolin-2-one (9c):
Oxindole 9c was obtained from alkyne 6c (200 mg, 0.55 mmol)
and boronic acid 7 as an orange oil (88 mg, 50%); 1H NMR
(300 MHz, CDCl3): d=7.85 (s, 1H, H-a), 7.80 (d, J=7.8 Hz, 1H, H-4’),
7.31–7.34 (m, 5H, Ph-H), 7.16 (t, J=7.8 Hz, 1H, H-6’), 6.92 (s, 2H, H-
ChemMedChem 2011, 6, 1693 – 1705
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1701

R. Pontikis, G. G. Chabot, J.-C. Florent, et al.
MED
2, H-6), 6.88 (t, J=7.8 Hz, 1H, H-5’), 6.74 (d, J=7.8 Hz, 1H, H-7’),
5.00 (s, 2H, CH2), 3.93 (s, 3H, OCH3), 3.87 ppm (s, 6H, OCH3);
13C NMR (75 MHz, CDCl3): d=168.6 (CH-2’), 153.2 (C-3, C-5), 143.4
(C-7’a), 139.2 (C-4), 137.7 (CH-a), 136.0 (C-Ph), 130.1 (C-1), 129.6
(CH-6’), 128.7 (CH-Phꢂ2), 127.6 (CH-Phꢂ2), 127.5 (CH-Ph), 126.3 (C-
3’a), 122.9 (CH-4’), 121.7 (CH-5’), 121.3 (C-3’), 109.3 (CH-7’), 106.6
(CH-2, CH-6), 61.0 (OCH3), 56.2 (OCH3 ꢂ2), 43.7 ppm (CH2); MS (ES+
): m/z: 402 [M+H]+, 424 [M+Na]+; HRMS (ESI): m/z [M+H]+ calcd
for C25H23O4N: 402.1705, found: 402.1704
(E)-6,7,8-Trimethoxy-4-[1-(4-methoxyphenyl)ethylidene]isochro-
man-1-one (27): Isochromanone 27 was obtained from alkyne 25
(50 mg, 0.128 mmol) and 4-methoxyphenylboronic acid as beige
crystals (17 mg, 36%); mp: 99–1018C; 1H NMR (300 MHz, CDCl3):
d=7.10 (d, J=8.7 Hz, 2H, H-2’, H-6’), 6.84 (d, J=8.7 Hz, 2H, H-3’,
H-5’), 5.90 (s, 1H, H-5), 4.93 (s, 2H, CH2-3), 3.97 (s, 3H, OCH3), 3.83
(s, 3H, OCH3), 3.79 (s, 3H, OCH3), 3.27 (s, 3H, OCH3), 2.24 ppm (s,
3H, CH3); 13C NMR (75 MHz, CDCl3): d=162.1 (C-1), 159.2 (C-4’),
156.0 (C-6), 155.0 (C-8), 142.1 (C-7), 138.7 (C-a), 136.4 (C-4a), 134.5
(C-1’), 129.9 (CH-2’, CH-6’), 123.4 (C-4), 114.1 (CH-3’, CH-5’), 112.1
(C-8a), 107.2 (CH-5), 67.2 (CH2-3), 62.2 (OCH3), 61.1 (OCH3), 55.3
(OCH3 ꢂ2), 21.7 ppm (CH3); MS (ES+): m/z: 393 [M+Na]+; HRMS
(ESI) m/z [M+H]+ calcd for C21H23O6: 371.1495, found: 371.1496
(E)-6-Methoxy-3-(1-(3,4,5-trimethoxyphenyl)ethylidene)benzofur-
an-2-one (9e): Benzofuranone 9e was obtained from alkyne 6e
(150 mg, 0.47 mmol) and boronic acid 7 as a yellow crystalline
solid (73 mg, 44%); mp: 147–1488C; 1H NMR (CDCl3, 300 MHz):
d=6.59 (s, 1H, H-7’), 6.51 (s, 2H, H-2, H-6), 6.37 (m, 2H, H-4’, H-5’),
3.92 (s, 3H, OCH3), 3.83 (s, 6H, OCH3), 3.76 (s, 3H, OCH3), 2.70 ppm
(s, 3H, CH3); 13C NMR (75 MHz, CDCl3): d=167.9 (C-2’), 160.9 (C-6’),
154.7 (C-a), 153.9 (C-3, C-5, C-7’a) 138.2 (C-4), 137.3 (C-1), 123.7 (C-
4’), 118.9 (C-3’), 116.2 (C-3’a), 109.3 (C-5’), 103.4 (CH-2, CH-6), 96.7
(CH-7’), 61.1 (OCH3), 56.6 (OCH3 ꢂ2), 55.6 (OCH3), 22.7 ppm (CH3);
MS (ES+) m/z: 357 [M+H]+, 379 [M+Na]+; HRMS (ESI) m/z [M+
H]+ calcd for C20H21O6: 357.1338, found: 357.1337.
Compounds obtained through CꢀH activation (path B)
(E)-6-Methoxy-1-methyl-3-[1-(3,4,5-trimethoxyphenyl)ethylide-
ne]indolin-2-one (9 f): Propynamide 12 (28 mg, 0.137 mmol), tri-
methoxyiodobenzene (46 mg, 0.157 mmol), NaOAc (23 mg,
0.280 mmol), and dry DMF (2 mL) were added under argon to a
flame-dried flask. The solution was purged with argon for 20 min,
and Pd(OAc)2 (1.5 mg, 0.0066 mmol) was added. After stirring at
808C for 2 h, the reaction mixture was cooled to room tempera-
ture, quenched with H2O, and extracted twice with EtOAc (10 mLꢂ
2). The combined organic layers were washed with brine (10 mLꢂ
2), dried (MgSO4), and concentrated under reduced pressure. The
residue was subjected to flash chromatography (cyclohexane/
EtOAc 75:25) to give 9 f as a pale-yellow solid which was recrystal-
lized from Et2O/hexane (23 mg, 46%); mp: 155–1588C; 1H NMR
(300 MHz, CDCl3): d=6.48 (s, 2H, H-2, H-6), 6.33 (s, 1H, H-7’), 6.23
(m, 2H, H-4’, H-5’), 3.92 (s, 3H, OCH3), 3.82 (s, 6H, OCH3), 3.76 (s,
3H, OCH3), 3.23 (s, 3H, N-CH3), 2.75 ppm (s, 3H, C-CH3); 13C NMR
(75 MHz, CDCl3): d=168.7 (C-2’), 160.3 (C-6’), 153.9 (C-3, C-5),
151.3 (C-a), 143.7 (C-7’a), 138.6 (C-1), 137.7 (C-4), 124.0 (CH-4’),
123.0 (C-3’), 115.5 (C-3’a), 105.5 (CH-5’), 103.6 (CH-2, CH-6), 95.1
(CH-7’), 61.1 (OCH3), 56.2 (OCH3 ꢂ2), 55.4 (OCH3), 25.7 (N-CH3),
22.4 ppm (C-CH3); MS (DCI/CH4): m/z: 370 [M+H]+; HRMS (DCI/
CH4) m/z [M+H]+ calcd for C21H24O5N: 370.1654, found: 370.1654.
(E)-3,(E)-4-(3,4,5-Trimethoxyphenyl)but-3-en-2-ylidene)indolin-2-
one (10a): Oxindole 10a was obtained from alkyne 6a (50 mg,
0.175 mmol) and styrylboronic acid 8 as a bright orange powder
1
(43 mg, 70%); mp: 218–2208C; H NMR (300 MHz, CDCl3): d=7.90
(brs, 1H, NH), 7.73 (d, J=15.9 Hz, 1H, H-b), 7.61 (d, J=7.8 Hz, 1H,
H-4’), 7.20 (t, J=7.8 Hz, 1H, H-6’), 7.13 (d, J=15.9 Hz, 1H, H-a), 7.03
(t, J=7.8 Hz, 1H, H-5’), 6.88 (d, J=7.8 Hz, 1H, H-7’), 6.80 (s, 2H, H-
2, H-6), 3.93 (s, 6H, OCH3), 3.90 (s, 6H, OCH3), 2.76 ppm (s, 3H,
CH3); 13C NMR (75 MHz, CDCl3): d=170.0 (C-2’), 153.6 (C-3, C-5),
149.3 (C-c), 139.6 (C-4), 139.4 (C-7’a), 137.9 (CH-a), 132.2 (C-1), 128.0
(CH-6’), 124.5 (CH-4’), 124.0 (C-3’a), 123.7 (C-3’), 121.7 (CH-5’), 109.5
(CH-7’), 104.6 (CH-2, CH-6), 61.0 (OCH3), 56.2 (OCH3 ꢂ2), 15.4 ppm
(CH3); MS (ES+) m/z: 352 [M+H]+, 374 [M+Na]+; HRMS (ESI) m/z
[M+H]+ calcd for C21H22O4N: 352.1549, found: 352.1549. The (E)-
3,(Z)-4 isomer trans-10a was obtained when the coupling reaction
occurs with NaOH as additive[21]
:
mp: 220–2218C; 1H NMR
(300 MHz, CDCl3): d=9.18 (d, J=16 Hz, 1H, H-b), 7.77 (brs, 1H
NH), 7.62 (d, J=7.6 Hz, 1H, H-4’), 7.19 (t, J=7.6 Hz, 1H, H-6’), 7.13
(d, J=16 Hz, 1H, H-a), 7.03 (t, J=7.6 Hz, 1H, H-5’), 6.87 (s, 2H, H-2,
H-6), 6.84 (d, J=7.6 Hz, 1H, H-7’), 3.93 (s, 6H, OCH3), 3.88 (s, 3H,
OCH3), 2.55 ppm (s, 3H, CH3); 13C NMR (75 MHz, CDCl3): d=169.3
(C-2’), 153.4 (C-3, C-5), 148.8 (C-c), 139.7 (C-7’a), 139.1 (C-4), 136.9
(CH-a), 132.7 (C-1), 128.0 (CH-6’), 126.9 (CH-b), 124.9 (C-3’a), 124.7
CH-4’), 122.8 (C-3’), 121.7 (CH-5’), 109.3 (CH-7’), 104.8 (CH-2, CH-6),
61.0 (OCH3), 56.1 (OCH3 ꢂ2), 16.8 ppm (CH3); MS (ES+) m/z: 352
[M+H]+, 374 [M+Na]+.
(E)-3-[1-(3-Hydroxy-4-methoxyphenyl)ethylidene]-5,6,7-trime-
thoxy-1-[(2-(trimethylsilyl)ethoxy)methyl]indoline-2-one
(17):
Employing propynamide 15 (116 mg, 0.306 mmol) and aryliodide
16 (88 mg, 0.351 mmol) and following the same procedure as
above, but after a heating at 1508C for 5 min, the cycloadduct 17
was obtained after flash chromatography (cyclohexane/EtOAc
1
75:25) as a yellow oil (70 mg, 46%); H NMR (300 MHz, CDCl3): d=
6.95 (d, J=8.1 Hz, 1H, H-5’), 6.88 (s, 1H, H-2’), 6.79 (d, J=8.1 Hz,
1H, H-6’), 5.78 (s, 1H, H-4), 5.71 (s, 1H, OH), 5.38 (s, 2H, N-CH2),
3.93 (s, 3H, OCH3), 3.90 (s, 3H, OCH3), 3.82 (s, 3H, OCH3), 3.38 (s,
3H, OCH3), 3.63 (t, J=8.1 Hz, 2H, O-CH2), 2.74 (s, 3H, CH3), 0.94 (t,
J=8.1 Hz, 2H, Si-CH2), ꢀ0,01 ppm (s, 9H, CH3); 13C NMR (75 MHz,
CDCl3): d=168.9 (C-2), 154.1 (C-a), 148.2 (C-5), 146.6 (C-4’), 146.2
(C-3’), 143.2 (C-6), 139.3 (C-7), 136.1 (C-1’), 127.2 (C-7a), 123.3 (C-3),
118.6 (CH-6’), 118.3 (C-3a), 113.3 (CH-2’), 111.1 (CH-5’), 103.7 (CH-4),
70.2 (N-CH2), 65.7 (O-CH2), 61.4 (OCH3), 60.9 (OCH3), 56.1 (OCH3),
55.7 (OCH3), 22.9 (CH3-a), 18.0 (Si-CH2), ꢀ1,4 ppm (CH3); MS (ES+):
m/z: 524 [M+Na]+; HRMS (DCI/CH4) m/z [M+H]+ calcd for
C26H36O7NSi: 502.2261, found: 502.2243.
(E)-3,(E)-4-(3,4,5-Trimethoxyphenyl)but-3-en-2-ylidene)benzofur-
an-2-one (10d): Benzofuranone 10d was obtained from alkyne 6d
(64 mg, 0.22 mmol) and styrylboronic acid 8 as an orange powder
1
(38 mg, 47%); mp: 172–1738C; H NMR (300 MHz, CDCl3): d=7.68
(d, J=16.0 Hz, 1H, H-b), 7.64 (d, J=7.8 Hz, 1H, H-4’), 7.30 (t, J=
7.8 Hz, 1H, H-6’), 7.27 (d, J=16.0 Hz, 1H, H-a), 7.16 (t, J=7.8 Hz,
1H, H-5’), 7.13 (d, J=7.8 Hz, 1H, H-7’), 6.81 (s, 2H, H-2, H-6), 3.94 (s,
6H, OCH3), 3.91 (s, 3H, OCH3), 2.72 ppm (s, 3H, CH3); 13C NMR
(75 MHz, CDCl3): d=167.9 (C-2’), 153.6 (C-3, C-5), 152.7 (C-7’a),
152.2 (C-c), 139.9 (C-4), 139.6 (CH-a), 131.7 (C-1), 128.9 (CH-6’),
126.8 (CH-b), 124.4 (C-3’), 123.8 (CH-4’), 123.7 (CH-5’), 118.5 (C-3’a),
110.9 (CH-7’), 105.0 (CH-2, CH-6), 61.1 (OCH3), 56.3 (OCH3 ꢂ2),
15.7 ppm (CH3); MS (ES+): m/z: 353 [M+H]+, 375 [M+Na]+; HRMS
(ESI) m/z [M+H]+ calcd for C21H21O5: 353.1389, found: 352.1383.
2,3,4-Trimethoxybenzyl
3-(5-iodo-2-methoxyphenoxy)but-2-
enoate (19). Employing propynamide 18 (100 mg, 0.378 mmol)
and aryliodide 16 (108 mg, 0.434 mmol) and following the same
procedure as for 9 f, but after a heating at 708C for 20 min, ether
19 was obtained after flash chromatography (toluene/EtOAc 97:3)
1702
www.chemmedchem.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1693 – 1705

Constrained Heterocyclic CA4 Analogues
1
as a yellow oil (191 mg, 98%); H NMR (300 MHz, CDCl3): d=7.45
were grown in Dulbecco’s modified Eagle’s medium (DMEM) con-
taining 2 mm l-glutamine, 10% fetal bovine serum (FBS),
100 UmLꢀ1 penicillin, and 100 mgmLꢀ1 streptomycin (378C, 5%
CO2). EAhy 926 cells in exponential growth were plated onto 96-
well plates at 5000 cells per 100 mL per well. Twenty-four hours
after plating, the medium was aspirated, and 100 mL medium con-
taining the test compound was added to the well containing the
cells (in triplicate) in 10-fold dilutions, and incubated for 2 h. After
the 2 h incubation period, digital micrographs were taken of repre-
sentative center areas of each well at a magnification of 100ꢂ and
200ꢂ. CA4 was routinely included in the experiments as internal
standard.
(dd, J=8.6, 2.1 Hz, 1H, H-4’), 7.27 (d, J=2.1 Hz, 1H, H-6’), 7.02 (d,
J=8.5 Hz, 1H, H-6), 6.71 (d, J=8.6 Hz, 1H, H-3’), 6.65 (d, J=8.5 Hz,
1H, H-5), 5.05 (s, 2H, CH2-a), 4.83 (s, 1H, H-d), 3.87 (s, 3H, OCH3),
3.85 (s, 3H, OCH3), 3.84 (s, 3H, OCH3), 3.77 (s, 3H, OCH3), 2.48 ppm
(s, 3H, CH3); 13C NMR (75 MHz, CDCl3): d=172.1 (C-e), 167.4 (CO),
154.1 (C-4), 152.5 (C-2), 151.4 (C-2’), 142.4 (C1’), 142.1 (C-3), 135.5
(CH-4’), 131.7 (CH-6’), 125.2 (CH-2), 122.2 (C-1), 114.7 (CH-3’), 107.1
(CH-5), 95.3 (CH-d), 81.6 (C-5), 61.4 (OCH3), 60.7 (CH2-a), 60.8
(OCH3), 55.9 (OCH3), 55.8 (OCH3), 18.1 ppm (CH3); MS (ES+): m/z:
536 [M+Na]+.
(E)-3-(1-(3,4,5-Trimethoxyphenyl)ethylidene)indolin-2-one (9g).
Pyrrolidine (1.8 mL, 21 mmol) was added dropwise to a well-stirred
mixture of 2-indolinone (2.0 g, 15 mmol) and 3,4,5-trimethoxyace-
tophenone (3.15 g, 15 mmol) in MeOH (20 mL). After heating at
reflux for 16 h, the mixture was cooled, and the orange solid prod-
uct was filtered to give, after recrystallization in MeOH, pure 9g
(2.8 g, 57%); mp: 215–2168C; 1H NMR (300 MHz, CDCl3): d=7.96
(brs, 1H, NH), 7.09 (t, J=7.7 Hz, 1H, H-6’), 6.80 (d, J=7.7 Hz, 1H,
H-7’), 6.68 (t, J=7.7 Hz, 1H, H-5’), 6.50 (s, 2H, H-2, H-6), 6.33 (d, J=
7.7 Hz, 1H, H-4’), 3.93 (s, 3H, OCH3), 3.83 (s, 6H, OCH3), 2.78 ppm
(s, 3H, CH3); 13C NMR (75 MHz, CDCl3): d=169.6 (C-2’), 155.2 (C-a),
153.9 (C-3, C-5), 139.3 (C-7’a), 138.3 (C-4), 137.9 (C-1), 128.2 (CH-6’),
123.6 (C-3’a), 123.4 (CH-4’), 123.3 (C-3’), 121.5 (CH-5’), 109.2 (CH-7’),
103.3 (CH-2, CH-6), 61.1 (OCH3), 56.2 (OCH3 ꢂ2), 22.8 ppm (CH3); MS
(ES+): m/z: 326 [M+H]+, 348 [M+Na]+; Anal. calcd for C19H19NO4:
C 70.14, H 5.89,N 4.31, found: C 70.12, H 5.85, N 4.31.
MTT cell proliferation assays: Murine B16 melanoma cells, murine
3LL Lewis lung carcinoma cells, and EAhy 926 endothelial cells
were grown in DMEM medium containing 2 mm l-glutamine, 10%
FBS, 100 UmLꢀ1 penicillin, and 100 mgmLꢀ1 streptomycin (378C,
5% CO2). All compounds were initially dissolved in DMSO at a
stock concentration of 2.5 mgmLꢀ1 and were further diluted in cell
culture medium. For comparative purposes, CA4 was routinely in-
cluded in the experiments as reference compound. Cells in expo-
nential growth were plated onto 96-well plates at 5000 cells per
well in 100 mL culture medium. Twenty-four hours after plating,
100 mL medium containing the compound of interest at final con-
centrations ranging from 0.01 to 30 mm were added to the wells
(in triplicate) containing the cells, and incubated for 48 h at 378C
and 5% CO2. After the 48 h exposure period to the test com-
pounds, cell viability was assayed using the MTT test,[59] and ab-
sorbance was read at l 562 nm in a microplate reader (BioKinetics
Reader, EL340). Appropriate controls with DMEM only and MTT
were run to subtract background absorbance. The concentration
of compound that inhibited cell viability by 50% (inhibitory con-
centration for 50% of cells, or IC50) was determined using Graph-
Pad Prism software. Results are presented as the mean of three in-
dependent experiments each run in triplicate.
Biological evaluations
Inhibition of tubulin polymerization: Microtubule disruption was
monitored by using the fluorescent dye 4’,6-diamidino-2-phenylin-
dole (DAPI)[55] in a 96-well plate format as described by Barron
et al.[29b] and Bane et al.[56] The standard assay was performed as
follows: wells were charged with tubulin (Cytoskeleton, 97% pure,
final concentration 1 mgmLꢀ1) in PME buffer (100 mm 1,4-piperazi-
nebis(ethanesulfonic acid (PIPES), 1 mm MgSO4, 2 mm EGTA) with
10 mm DAPI and varying concentrations of the test compounds,
using colchicine as an internal control. After pre-incubation at
room temperature for 45 min, 1 mm GTP (5 mL) was added to each
well to initiate tubulin polymerization, and the plate was then
transferred to a thermostated Victor plate reader at 378C for an ad-
Antiproliferative assays on human carcinoma cell lines: These assays
were monitored at the laboratory “Ciblothꢄque Cellulaire”, ICSN-
CNRS, Gif-sur-Yvette, France. HCT116 (colon carcinoma) and MCF7
(hormone-dependent breast carcinoma) cells were grown in RPMI
1640 medium supplemented with 10% fetal calf serum, streptomy-
cin, and amphotericin B (Fungizoneꢅ), and maintained at 378C in a
humidified atmosphere containing 5% CO2. Cells (500–800 per
well) were seeded in 96-well microplates containing 200 mL
growth medium. After culturing for 24 h, cells were exposed to
varying concentrations (0.5 nm–10 mm) of the test compound dis-
solved in DMSO (<1% in each preparation). After 72 h of incuba-
tion, the MTS reagent (40 mL, Promega) was added 2 h before the
absorbance at l 490 nm was recorded. IC50 corresponds to the con-
centration of the test compound that elicits 50% inhibition of cell
growth. Experiments were performed in triplicate.
ditional 2 h. Fluorescence was then read (lexcitation: 360 nm, lemission
:
450 nm. Percent inhibition was determined as follows: 1ꢀ(DFsample
/
DFcontrol)ꢂ100, for which DFcontrol =F(no inhibition)ꢀF(complete in-
hibition), and DFsample =F(sample)ꢀF(complete inhibition with col-
chicine). The IC50 for compound-induced inhibition of tubulin poly-
merization is the concentration of compound at which the extent
of inhibition of polymerization is 50% of the maximum value, as
determined from the semi-logarithmic plot of percent inhibition as
a function of drug concentration.
Cord disruption assay: HUVEC were prepared and cultured in EGM-
2MV microvascular endothelial cell growth medium (Lonza, Wal-
kersville, MD, USA), as previously described.[60] In brief, HUVEC were
plated in 96-well plates on a thick layer of Matrigel (Becton Dickin-
son, 75 mL per well) and allowed to form tube-like structures over
24 h. Test compounds were dissolved in DMSO (<0.1% in each
preparation, 100 mL per well) and were added to the formed cords
and incubated for 2.5 h. The effects of the compounds on capillary
tube disruption were evaluated by light microscopy (40ꢂ magnifi-
cation). Experiments were done in triplicate.
Endothelial cell morphology: Effect on the morphology of trans-
formed HUVEC (EAhy 926) cells. To assess the effects of the com-
pounds on the morphology of endothelial cells, we used the EAhy
926 endothelial cell line, which is derived from the fusion of
human umbilical vein endothelial cells (HUVEC) with the perma-
nent human cell line A549.[57] The EAhy 926 cell line is considered
to be one of the best immortalized HUVEC lines, because these
cells express most of the biochemical markers of parental HUVEC
EAhy 926 cells,[58] originally obtained from Dr. Cora-Jean S. Edgell
(Pathology Department, University of North Carolina, Chapel Hill,
NC 27599-7525, USA); these were used with her permission, and
Docking protocol: The reported 3D structure of tubulin was re-
trieved from the Protein Data Bank (http://www.rcsb.org/, PDB ID:
ChemMedChem 2011, 6, 1693 – 1705
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1703

R. Pontikis, G. G. Chabot, J.-C. Florent, et al.
MED
1SA0).[39] Subunits C, D, and E were removed. Only subunits A, B
(colchicine binding site), and small molecules DAMA-colchicine
(CN2), GTP, and the Mg2+ ion in this site were conserved. We en-
sured that correct ionization states were established for Asp, Glu,
and His residues. The orientations of the hydroxy group hydrogen
atoms from the Ser, Thr, and Tyr amino acids were not considered,
as these orientations are optimized during the docking runs. Final-
ly, no charge was added to the whole protein, because the algo-
rithm used processes it automatically. All docking runs were per-
formed into the colchicine binding site of tubulin, applying GOLD
(Genetic Optimization for Ligand Docking), with its default parame-
ters. This algorithm uses a stochastic genetic strategy for the con-
formational search and is also able to take into account conforma-
tional freedom of the residues. GoldScore is the GOLD fitness scor-
ing function, which is made of four components (external hydro-
gen bond, van der Waals and external van der Waals, internal
van der Waals, and internal torsion). This scoring function was opti-
mized for the prediction of ligand binding positions. Consequently,
it remains adapted in our comparison study. Ligands were built
within the builder module of Sybyl[61] and next submitted to
Corina,[62] a 3D structure generator. Obtained conformations were
next minimized within Sybyl using the Tripos force field.
[11] G. G. Dark, S. A. Hill, V. E. Prise, G. M. Tozer, G. R. Petitt, D. J. Chaplin,
Cancer Res. 1997, 57, 1829–1834.
[12] For reviews on CA4 analogues, see: a) N.-H. Nam, Curr. Med. Chem.
2003, 10, 1697–1722; b) H. P. Hsieh, J. P. Liou, N. Mahindroo, Curr.
Pharm. Des. 2005, 11, 1655–1677; c) G. C. Tron, T. Pirali, G. Sorba, F. Pa-
gliai, S. Busacca, A. A. Genazzani, J. Med. Chem. 2006, 49, 3033–3044;
d) T. Brown, H. Holt, Jr., M. Lee, Top Heterocycl. Chem. 2006, 2, 1–51;
e) R. Singh, H. Kaur, Synthesis 2009, 15, 2471–2491.
[13] For recent reports on CA4 analogues with a stereochemically defined
one-atom linker (1,1-ethylene bridge), see: a) A. Hamze, A. Giraud, S.
Messaoudi, O. Provot, J.-F. Peyrat, J. Bignon, J.-M. Liu, J. Wdzieczaek-
Bakala, S. Thoret, J. Dubois, J.-D. Brion, M. Alami, ChemMedChem 2009,
4, 1912–1924; b) R. ꢆlvarez, C. ꢆlvarez, F. Mollinedo, B. Sierra, M. Med-
arde, R. Pelꢇez, Bioorg. Med. Chem. 2009, 17, 6422–6431.
[14] For recent reports on heterocyclic CA4 derivatives, see: a) K. Bonezzi, G.
Taraboletti, P. Borsotti, F. Bellina, R. Rossi, R. Giavazzi, J. Med. Chem.
2009, 52, 79066–7910; b) R. Romagnoli, P. G. Baraldi, O. Cruz-Lopez, C.
Lopez Cara, M. D. Carrion, A. Brancale, E. Hamel, L. Chen, R. Bortolozzi,
G. Basso, G. Viola, J. Med. Chem. 2010, 53, 4248–4258; c) R. Schobert, B.
Biersack, A. Dietrich, K. Effenberger, S. Knauer, T. Mueller, J. Med. Chem.
2010, 53, 6595–6602; d) B. Biersack, K. Effenberger, R. Schobert, M.
Ocker, ChemMedChem 2010, 5, 420–427.
[15] For recent reports on polycyclic derivatives derived from CA4, see:
a) M.-J. Lai, C.-C. Kuo, T.-K. Yeh, H.-P. Hsieh, L.-T. Chen, W.-Y. Pan, K.-Y.
Hsu, J.-Y. Chang, J.-P. Liou, ChemMedChem 2009, 4, 588–593; b) A.
Putey, F. Popowycz, Q.-T. Do, P. Bernard, S. K. Talapatra, F. Kozielski, C. M.
Galmarini, B. Joseph, J. Med. Chem. 2009, 52, 5916–5925; c) N. Ty, G.
Dupeyre, G. G. Chabot, J. Seguin, L. Quentin, A. Chiaroni, F. Tillequin, D.
Scherman, S. Michel, X. Cachet, Eur. J. Med. Chem. 2010, 45, 3726–3739;
d) V. Colombel, A. Joncour, S. Thoret, J. Dubois, J. Bignon, J. Wdziec-
zaek-Bakala, O. Baudouin, Tetrahedron Lett. 2010, 51, 3127–3129; e) C.
Bailly, C. Bal, P. Barbier, S. Combes, J.-P. Finet, M.-P. Hildebrand, V. Peyrot,
N. Wattez, J. Med. Chem. 2003, 46, 6347–5444.
Acknowledgements
This work was supported financially by the Centre National de la
Recherche Scientifique (CNRS), the Institut Curie, the Institut Na-
tional de la Santꢀ et de la Recherche Mꢀdicale (INSERM), and by
a grant from the Institut National du Cancer (INCa, Boulogne Bill-
ancourt, France). M.A. is grateful to the Ministꢁre de l’Enseigne-
ment Supꢀrieur et de la Recherche for a PhD Fellowship Grant.
[16] Some functionalized 2-arylbenzylidene derivatives displayed cytotoxic
activity and act as tubulin polymerization inhibitors: a) N. J. Lawrence,
D. Rennison, A. T. McGown, J. A. Hadfield, Bioorg. Med. Chem. Lett. 2003,
13, 3759–3763; b) E. Rizzi, S. Dallavalle, L. Merlini, G. L. Beretta, G. Prate-
si, F. Zunino, Bioorg. Med. Chem. Lett. 2005, 15, 4313–4316; c) A. An-
dreani, S. Burnelli, M. Granaiola, A. Leoni, A. Locatelli, R. Morigi, M. Ram-
baldi, L. Varoli, M. W. Kunkel, J. Med. Chem. 2006, 49, 6922–6924; d) B.
Pandit, Y. Sun, P. Chen, D. L. Sackett, Z. Hu, W. Rich, C. Li, A. Lewis, K.
Schaefer, P.- K. Li, Bioorg. Med. Chem. 2006, 14, 6492–6501; e) A. Zuse,
P. Schmidt, S. Baasner, K. J. Bçhm, K. Mꢈller, M. Gerlach, E. G. Gꢈnther, E.
Unger, H. Prinz, J. Med. Chem. 2006, 49, 7816–7825.
Keywords: antivascular agents · combretastatin · heterocyclic
compounds · palladium-catalyzed reactions · tubulin
[1] a) J. Folkman, N. Engl. J. Med. 1971, 285, 1182–1186; b) J. Denekamp, Br.
J. Radiol. 1993, 66, 181–196.
[17] J. Kaffy, R. Pontikis, J.-C. Florent, C. Monneret, Org. Biomol. Chem. 2005,
3, 2657–2660.
[2] a) G. M. Tozer, C. Kanthou, B. C. Baguley, Nat. Rev. Cancer 2005, 5, 423–
435; b) C. Kanthou, G. M. Tozer, Int. J. Exp. Pathol. 2009, 90, 284–294;
c) D. W. Siemann, Cancer Treat. Rev. 2011, 37, 63–74, and references
therein.
[3] For reviews on VDAs, see: a) P. Hinnen, F. A. L. M. Eskens, Br. J. Cancer
2007, 96, 1159–1165; b) J. W. Lippert III, Bioorg. Med. Chem. 2007, 15,
605- 615; c) S. X. Cai, Recent Pat. Anti-Cancer Drug Discovery 2007, 2,
79–101; d) E. L. Schwartz, Clin. Cancer Res. 2009, 15, 2594–2601; e) M. J.
McKeage, B. C. Baguley, Cancer 2010, 116, 1859–1871.
[18] For pioneering works in this field, see: a) B. Burns, R. Grigg, V. Sridharan,
P. Stevenson, S. Sukirthalingam, T. Worakun, Tetrahedron Lett. 1989, 30,
1135–1138; b) C. H. Oh, Y. M. Lim, Tetrahedron Lett. 2003, 44, 267–270;
c) S. Couty, B. Liegault, C. Meyer, J. Cossy, Org. Lett. 2004, 6, 2511–2514;
d) W. S. Cheung, R. J. Patch, M. R. Player, J. Org. Chem. 2005, 70, 3741–
3744; e) R. Yanada, S. Obika, T. Inokuma, K. Yanada, M. Yamashita, S.
Ohta, Y. Takemoto, J. Org. Chem. 2005, 70, 6972–6975.
[19] A. Pinto, L. Neuville, P. Retailleau, J. Zhu, Org. Lett. 2006, 8, 4927–4930.
[20] For a recent review on the synthesis of 3-alkenyloxindoles, see: A. Mille-
maggi, R. J. K. Taylor, Eur. J. Org. Chem. 2010, 4527–4547.
[21] M. Arthuis, R. Pontikis, J.-C. Florent, Tetrahedron Lett. 2007, 48, 6397–
6400.
[4] a) M. A. Jordan, L. Wilson, Nat. Rev. Cancer 2004, 4, 253–265; b) S.
Honorꢀ, E. Pasquier, D. Braguer, Cell. Mol. Life Sci. 2005, 62, 3039–3056.
[5] For reviews on CA4, see: a) C. M. L. West, P. Price, Anti-Cancer Drugs
2004, 15, 179–187; b) D. W. Siemann, D. J. Chaplin, P. A. Walicke, Expert
Opin. Invest. Drugs 2009, 18, 189–197.
[22] Li and co-workers recently reported the synthesis of 3-(1-arymethyle-
ne)oxindoles by a Pd-catalyzed CꢀH functionalization of N-arylpropio-
namides with aryliodonium salts in the presence of Pd(OAc)2 and Et3N:
S. Tang, P. Peng, P. Zhong, J.-H. Li, J. Org. Chem. 2008, 73, 5476–5480.
[23] The 2,3,4-trimethoxyaniline 13 was prepared from 2,3-dimethoxyphenol
in a three-step process (nitration,[24] O-methylation, hydrogenation).
[24] C. L. Dwyer, C. W. Holzapfel, Tetrahedron 1998, 54, 7843–7848.
[25] N. J. Lawrence, F. A. Ghani, L. A. Hepworth, J. A. Hadfield, A. T. McGown,
R. G. Pritchard, Synthesis 1999, 1656–1660.
[26] Removal of the SEM moiety was unsuccessful under the investigated
conditions: a) J. M. Muchowski, D. R. Solas, J. Org. Chem. 1984, 49, 203–
205; b) G. Luo, L. Chen, G. Dubowchik, J. Org. Chem. 2006, 71, 5392–
5395.
[6] C. J. Mooney, G. Nagaiah, P. Fu, J. K. Wasman, M. M. Cooney, P. S. Savvi-
des, J. A. Bokar, A. Dowlati, D. Wang, S. S. Agarwala, S. M. Flick, P. H.
Hartman, J. D. Ortiz, P. N. Lavertu, S. C. Remick, Thyroid 2009, 19, 233–
240.
[7] G. J. Rustin, G. Shreeves, P. D. Nathan, A. Gaya, T. S. Ganesan, D. Wang, J.
Boxall, L. Poupard, D. J. Chaplin, M. R. L. Stratford, J. Balkissoon, M.
Zweifel, Br. J. Cancer 2010, 102, 1355–1360.
[8] ClinicalTrials.gov, A service of the U.S. National Institutes of Health:
http://clinicaltrials.gov/ct2/home (accessed June 6, 2011).
[9] L. Vincent, P. Kermani, L. M. Young, J. Cheng, F. Zhang, K. Shido, G. Lam,
H. Bompais-Vincent, Z. Zhu, D. J. Hicklin, P. Bohlen, D. J. Chaplin, C. May,
S. Rafii, J. Clin. Invest. 2005, 115, 2992–3006.
[10] C. Kanthou, G. M. Tozer, Blood 2002, 99, 2060–2069.
1704
www.chemmedchem.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1693 – 1705

Constrained Heterocyclic CA4 Analogues
[27] M. Arthuis, R. Pontikis, J.-C. Florent, J. Org. Chem. 2009, 74, 2234–2237.
[28] M. Z. Cherkaoui, G. Scherowsky, New J. Chem. 1997, 21, 1203–1210.
[29] a) C. Heusꢄle, D. Bonne, M.-F. Carlier, Eur. J. Biochem. 1987, 165, 613–
620; b) D. M. Barron, S. K. Chatterjee, R. Ravindra, R. Roof, E. Baloglu,
D. G. Kingston, S. Bane, Anal. Biochem. 2003, 315, 49–56.
[30] a) K. Grosios, S. E. Holwell, A. T. McGown, G. R. Pettit, M. C. Bibby, Br. J.
Cancer 1999, 81, 1318–1327; b) S. M. Galbraith, D. J. Chaplin, F. Lee,
M. R. Stratford, R. J. Locke, B. Vojnovic, G. M. Tozer, Anticancer Res. 2001,
21, 93–102.
[31] Compound 9g was synthesized by a Knoevenagel reaction.
[32] Compound 9b exhibits anti-tubulin activity similar to that of the re-
ported analogue,[16] which did not bear a methyl group on the olefinic
bond (comparison with podophyllotoxin; in our assay: IC50 =0.27 mm).
Thus, we suggest that substitution of the double bond with a methyl
group, which prevents olefinic isomerization, does not perturb the in-
teraction with tubulin.
[45] A. Tripathi, M. Fornabaio, G. E. Kellogg, J. T. Gupton, D. A. Gewirtz, W. A.
Yeudall, N. E. Vega, S. L. Mooberry, Bioorg. Med. Chem. 2008, 16, 2235–
2242.
[46] a) R. Romagnoli, P. G. Baraldi, M. D. Carrion, C. Lopez Cara, O. Cruz-
Lopez, D. Preti, M. Tolomeo, S. Grimaudo, A. Di Cristina, N. Zonta, J. Bal-
zarini, A. Brancale, T. Sarkar, E. Hamel, Bioorg. Med. Chem. 2008, 16,
5367–5376; b) R. Romagnoli, P. G. Baraldi, M. D. Carrion, O. Cruz-Lopez,
C. Lopez Cara, G. Basso, G. Viola, M. Khedr, J. Balzarini, S. Mahboobi, A.
Sellmer, A. Brancale, E. Hamel, J. Med. Chem. 2009, 52, 5551–5555.
[47] Y.-K. Chiang, C.-C. Kuo, Y.-S. Wu, C.-T. Chen, M. S. Coumar, J.-S. Wu, H.-P.
Hsieh, C.-Y. Chang, H.-Y. Jseng, M.-H. Wu, J.-S. Leou, J.-S. Song, J.-Y.
Chang, P.-C. Lyu, Y.-S. Chao, S.-Y. Wu, J. Med. Chem. 2009, 52, 4221–
4233.
[48] This different spatial arrangement could arise from the use, for current
docking studies,[39] of tubulin in a curved conformation, affording more
space for the binding of compounds in the colchicine site.[47]
[49] R. Yanada, S. Obika, Y. Kobayashi, T. Inokuma, M. Oyama, K. Yanada, Y.
Takemoto, Adv. Synth. Catal. 2005, 347, 1632–1642.
[33] Decreases in tubulin polymerization inhibition and cytotoxicity were
also observed, but to a lesser extent, by Li and co-workers if the amido
hydrogen atom of the corresponding non-stabilized oxindole was sub-
stituted with a methyl group.[16d]
[50] C. Ma, X. Liu, X. Li, J. Flippen-Anderson, S. Yu, J. M. Cook, J. Org. Chem.
2001, 66, 4525–4542.
[34] Within the dienic series, the derivative with o-methoxyphenol as the B-
ring is less potent than the phenyl derivative 1.[17]
[51] M. Tsukayama, H. Utsumi, A. Kunugi, H. Nozaki, Heterocycles 1997, 45,
1131–1142.
[35] This isomer was obtained when the tandem Heck–Suzuki–Miyaura reac-
tion was carried out with sodium hydroxide as additive instead of
cesium fluoride.[21]
[36] X. Ren, M. Dai, L.-P. Lin, P.-K. Li, J. Ding, Br. J. Pharmacol. 2009, 156,
1228–1238.
[37] G. Micheletti, M. Poli, P. Borsotti, M. Martinelli, B. Imberti, G. Taraboletti,
R. Giavazzi, Cancer Res. 2003, 63, 1534–1537.
[38] a) A. Cirla, J. Mann, Nat. Prod. Rep. 2003, 20, 558–564; b) M. B. Hadima-
ni, J. Hua, M. D. Jonklaas, R. J. Kessler, Y. Sheng, A. Olivares, R. P. Tanpure,
A. Weiser, J. Zhang, K. Edvardsen, R. R. Kane, K. G. Pinney, Bioorg. Med.
Chem. Lett. 2003, 13, 1505–1508.
[39] R. B. Ravelli, B. Gigant, P. A. Curmi, I. Jourdain, S. Lachkar, A. Sobel, M.
Knossow, Nature 2004, 428, 198–202.
[40] T. L. Nguyen, C. McGrath, A. R. Hermone, J. C. Burnett, D. W. Zaharevitz,
B. W. Day, P. Wipf, E. Hamel, R. Gussio, J. Med. Chem. 2005, 48, 6107–
6116.
[41] G. Jones, P. Willett, R. C. Glen, J. Mol. Biol. 1995, 245, 43–53.
[42] a) Y. Kong, J. Grembecka, M. C. Edler, E. Hamel, S. L. Mooberry, M. Sabat,
J. Rieger, M. L. Brown, Chem. Biol. 2005, 12, 1007–1014; b) C. Rappl, P.
Barbier, V. Bourgarel-Rey, C. Grꢀgoire, R. Gilli, M. Carre, S. Combes, J.-P.
Finet, V. Peyrot, Biochemistry 2006, 45, 9210–9218; c) S. Ducki, G. Mack-
enzie, B. Greedy, S. Armitage, J. Fournier Dit Chabert, E. Bennett, J. Net-
tles, J. P. Snyder, N. J. Lawrence, Bioorg. Med. Chem. 2009, 17, 7711–
7722.
[52] V. Rys, A. Couture, E. Deniau, P. Grandclaudon, Eur. J. Org. Chem. 2003,
1231–1237.
[53] A. Maiti, P. V. N. Reddy, M. Cushman, M. Sturdy, S. Pegan, A. D. Mesecar,
L. Marler, J. M. Pezzuto, J. Med. Chem. 2009, 52, 1873–1884.
[54] M. M. Taafroot, Y. Landais, J.-P. Robin, D. Davoust, Tetrahedron Lett.
1986, 27, 1781–1784.
[55] D. Bonne, C. Heusele, C. Simon, D. Pantaloni, J. Biol. Chem. 1985, 260,
2819–2825.
[56] S. L. Bane, R. Ravindra, A. A. Zaydman in Microtubule Protocols (Ed.: J.
Zhou), Humana Press, Totowa, 2007, pp. 281–288.
[57] C. J. Edgell, C. C. McDonald, J. B. Graham, Proc. Natl. Acad. Sci. USA
1983, 80, 3734–3737.
[58] D. Bouis, G. A. Hospers, C. Meijer, G. Molena, N. H. Mulder, Angiogenesis
2001, 4, 91–102.
[59] D. A. Scudiero, R. H. Shoemaker, K. D. Paull, A. Monks, S. Tierney, T. H.
Nofziger, M. J. Currens, D. Seniff, M. R. Boyd, Cancer Res. 1988, 48,
4827–4833.
[60] F. Zemani, D. Benisvy, I. Galy-Fauroux, A. Lokajczyk, S. Colliec-Jouault, G.
Uzan, A. M. Fischer, C. Boisson-Vidal, Biochem. Pharmacol. 2005, 70,
1167–1175.
[61] Sybyl X Suite, Tripos International, St. Louis MO (USA), 2009.
[62] a) J. Sadowski, J. Gasteiger, Chem. Rev. 1993, 93, 2567–2581; b) CORINA
–
Fast Generation of High-Quality 3D Molecular Models: http://www.
molecular-networks.com/products/corina (accessed June 6, 2011).
[43] A hydrogen bond with Asnb258 has already been observed with com-
pounds bearing an aniline[44] or a nitrogen heterocycle[45] as B-ring.
[44] D. Simoni, F. P. Invidiata, M. Eleopra, P. Marchetti, R. Rondanin, R. Baru-
chello, G. Grisolia, A. Tripathi, G. E. Kellogg, D. Durrant, R. M. Lee, Bioorg.
Med. Chem. 2009, 17, 512–522.
Received: March 23, 2011
Revised: May 24, 2011
Published online on July 5, 2011
ChemMedChem 2011, 6, 1693 – 1705
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1705