Synthesis and biological evaluation of cis-locked vinylogous combretastatin-A4 analogues: Derivatives with a cyclopropyl-vinyl or a cyclopropyl-amide bridge

Article abstract of DOI:10.1016/j.bmcl.2009.01.062

A series of novel combretastatin A4 analogues, in which the cis-olefinic bridge is replaced by a cyclopropyl-vinyl or a cyclopropyl-amide moiety, were synthesized and evaluated for inhibition of tubulin polymerization and antiproliferative activity. The derivative 9a with a (cis,E)-cyclopropyl-vinyl unit is the most promising compound. As expected, molecular docking of 9a has shown that only one of the cis-cyclopropyl enantiomers is a good ligand for tubulin.

Full text of DOI:10.1016/j.bmcl.2009.01.062

Bioorganic & Medicinal Chemistry Letters 19 (2009) 1318–1322
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and biological evaluation of cis-locked vinylogous combretastatin-A4
analogues: Derivatives with a cyclopropyl-vinyl or a cyclopropyl-amide bridge
Nancy Ty a,b, , Julia Kaffy a,b, , Alban Arrault c,d, Sylviane Thoret e, Renée Pontikis a,b,
,
*
Joelle Dubois e, Luc Morin-Allory c,d, Jean-Claude Florent a,b
a Institut Curie, Centre de Recherche, 26 rue d’Ulm, 75248 Paris, France
b CNRS, UMR 176, 26 rue d’Ulm, 75248 Paris, France
c ICOA, Université d’Orléans, BP 6759, 45067 Orléans Cedex 2, France
d CNRS, UMR 6005, 45067 Orléans Cedex 2, France
e ICSN, CNRS, UPR 2301, Avenue de la Terrasse, 91198 Gif-sur-Yvette, France
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of novel combretastatin A4 analogues, in which the cis-olefinic bridge is replaced by a cyclopro-
pyl-vinyl or a cyclopropyl-amide moiety, were synthesized and evaluated for inhibition of tubulin poly-
merization and antiproliferative activity. The derivative 9a with a (cis,E)-cyclopropyl-vinyl unit is the
most promising compound. As expected, molecular docking of 9a has shown that only one of the cis-
cyclopropyl enantiomers is a good ligand for tubulin.
Received 18 November 2008
Revised 17 January 2009
Accepted 22 January 2009
Available online 27 January 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
Tubulin polymerization
Colchicine binding site
Combretastatin A4
Bioisosteric replacement
Cyclopropane
Tumour vasculature represents an attractive target for antican-
cer drug discovery.1,2 Vascular disrupting agents (VDA), which ex-
ploit differences between normal and immature tumour blood
vessels, elicit selective shutdown of blood flow resulting in exten-
sive tumour-cell necrosis.3 The most promising compound, com-
bretastatin A4 (CA4) in the form of a water-soluble phosphate
prodrug (CA4P), is currently in phase II and III clinical trials for
the treatment of solid tumours.3–5 CA4 is a natural cis-stilbene
product isolated from the South African willow tree Combretum
caffrum, which strongly inhibits tubulin polymerization by binding
to the colchicine site.6 Interference with microtubule dynamics af-
fects the cell signaling pathways involved in regulating and main-
taining the cellular cytoskeleton of endothelial cells in tumour
vasculature.7,8 The result for the tumour’s blood vessels, is a rapid
morphological change of their endothelial cells leading to their
occlusion and interruption of blood flow.9,10
requirements for efficient binding to tubulin.11 Accordingly, many
structural variations have been investigated, but few acyclic
bridgehead analogues with three (or more)-atom linkers inserted
between the two aryl rings have been reported.12,13 We have pre-
viously shown with the vinylogous analogues of CA4, 1 and its
(E,Z)-dienic isomer, that a (E,Z)-butadiene moiety could replace
the cis-olefinic bond of CA4 without compromising tubulin poly-
merization properties.14 Moreover, derivative 2 with a simple phe-
nyl group as the B-ring, was found to be a more potent inhibitor
than CA4 (Fig. 1). But these dienic derivatives are prone to isomer-
ization to the more stable but inactive (E,E)-isomeric derivatives.
Thus, our attention was focused on the structural modifications
that maintained the cis-diaryl relationship of 2 (Fig. 2). The cyclo-
CH3O
CH3O
Extensive studies have been conducted to examine the struc-
ture–activity relationships of variously modified CA4 analogues.
The 3,4,5-trimethoxy substitutions on the A-ring and the cis-orien-
tation between the two aryl rings have been reported as essential
CH3O
CH3O
A
CH3O
B
CH3O
R
R1
OCH3
R2
1(Z,E): R1 = OH, R2 = OCH3
2: R1 = R2 = H
CA4: R = OH
CA4P: R = OPO3Na2
* Corresponding author. Tel.: +33(0)1 56 24 66 59; fax: +33(0)1 56 24 66 31.
E-mail address: renee.pontikis@curie.fr (R. Pontikis).
J.K. and N.T.: These authors contributed equally to this work.
Figure 1. The structures of CA4 and vinylogous analogues.
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.01.062

N. Ty et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1318–1322
1319
could not be obtained in a pure form after column chromatogra-
phy, it was submitted to a Suzuki–Miyaura coupling. Thus, treat-
ment of a mixture of vinyl iodides 3a and 3b (9:1) with phenyl
boronic acid, in the presence of Pd(PPh3)4 and EtONa as a base
(Scheme 2),14 afforded the adduct 9a, which retained the configu-
ration (cis,E) of the preponderant starting iodide (40% yield, purity
90%). The (trans,E)-analogue 9b was obtained, under the same con-
ditions, from a mixture containing mainly the trans-isomer 3b.22
These disappointing results led us to examine the formation of
the desired (cis,E)-cyclopropyl-styryl derivative 9a, via a Wittig
olefination with aldehyde 4. Therefore, by condensation of 4 with
phosphorus ylide generated from benzyl tributylphosphonium
bromide and n-BuLi, a mixture of Z- and E-olefins (9a/9c 1:1)
was obtained in 70% yield. Nevertheless, isomers 9a (cis,E) and 9c
(cis,Z) could be isolated in pure form for biological assays after nor-
mal phase HPLC purification.23 As expected, unlike the dienic com-
pound 2, the cyclopropyl-vinyl analogues 9a and 9c showed great
chemical stability.
Next, we investigated the replacement of the (E)-double bond
by an amide group. Type II CA4 analogues, bearing a cyclopro-
pane-carboxamide linker, would arise from the coupling reaction
of the cis-cyclopropane carboxylic acid 10a with a variety of ani-
lines (Scheme 3). In the presence of copper[II]acetylacetonate,
cyclopropanation of the known olefin 1124 with ethyl diazoacetate
(EDA)25 afforded a mixture of the cis- and trans-cyclopropane car-
boxylates 12a and 12b (46% conversion, ratio 3:7). When Cu(acac)2
was added in several batches, to avoid catalyst deactivation,26 the
yield in cyclopropane products was increased (69%). Under these
conditions cyclopropanation provided the cis-isomer 12a (18%
yield) and the trans-isomer 12b (46% yield) after chromatographic
separation.27 Hydrolysis of the ester trans-12b with 2 N NaOH at
room temperature gave the acid 10b. Saponification of the cis-
derivative 12a, had to be performed at 50 °C during 7 h, affording
the desired carboxylic acid 10a.19,27
Under standard EDCI/HOBt conditions, acid 10a was coupled to
aniline and two other amines bearing a phenol function protected
by a silyl group,28,29 to give the corresponding cis-amides 13a, 14a
and 15a. Removal of the hydroxyl protecting group by the action of
TBAF afforded the phenol derivatives 16a and 17a. In a similar
manner, the carboxamides 13b and 14b were obtained by coupling
reactions with the trans-cyclopropane carboxylic acid 10b. Treat-
ment of the acids 10a and 10b with oxalyl chloride followed by
coupling with 3-aminopyridine afforded the cis- and trans-deriva-
tives 18a and 18b in 71 and 63% yields, respectively.30
The synthesized cyclopropyl derivatives were evaluated (Table
1) for their ability to inhibit tubulin polymerization and their anti-
proliferative activities against two types of human cancer cell lines,
HCT-116 (colon carcinoma) and MCF-7 (hormone dependent
breast carcinoma). Against the MCF-7 breast cancer cells, only
CH3O
CH3O
OCH3
Type I
R
CH3O
CH3O
OCH3
H
N
CH3O
CH3O
O
2
OCH3
R
Type II
Figure 2. Design concept of new CA4 analogues.
propyl group is generally considered as an alkene bioisostere.15 In
this view, we targeted the synthesis of configurationally cis-locked
CA4 analogues of type I, which are characterized by a (cis,E)-cyclo-
propyl-vinyl unit.16 We were also interested to investigate the bio-
logical evaluation of cyclopropyl derivatives (type II) for which the
E-double bond was replaced by an amide group. Indeed, a second-
ary amide may be considered to have a similar spatial disposition
than a trans-olefin, despite its different electronic character.17 The
influence of the substitution on the B-ring was also examined.
Access to the type I derivatives, with a (cis,E)-cyclopropyl-vinyl
moiety, was envisioned via a Suzuki coupling reaction between
aromatic boronic acids and the vinyl iodide 3a which may be ob-
tained by a Takai homologation of the cyclopropane carbaldehyde
4 (Scheme 1). The Sonogashira coupling of aryliodide 5 with prop-
argyl alcohol afforded alkyne 6 which was reduced selectively to
the Z-olefin 7 by catalytic hydrogenation.18 According to Cooper19
,
the cyclopropanation furnished the cyclopropylcarbinol 8 which
was converted, by a ruthenium-catalyzed (TPAP/NMO) oxidation,
to the cis-cyclopropane carbaldehyde 4 (80% yield).20,21 Takai
homologation of aldehyde 4 was carried out using chromous chlo-
ride and iodoform in a ratio of 5.5:1 and was first investigated at
0 °C. A mixture of two E-vinyl iodides (J = 14.4 Hz), in a 8:2 ratio,
was isolated after chromatography (56% yield). At lower tempera-
ture (À10 °C) the ratio of these both isomers was inversed (3:7,
65% yield) while, at À78 °C, the reaction did not occur and the alde-
hyde was recovered in its trans-isomerized form. We hypothesized
that, at 0 °C, homologation occurred rapidly, so that the desired
(cis,E)-vinyl iodide 3a was predominantly formed, but the starting
cis-cyclopropane aldehyde 4 was in part isomerized into the trans-
isomer, leading to concomitant formation of the (trans,E)-product
3b after Takai reaction. Although the expected vinyl iodide 3a
I
OH
(c)
(a)
(b)
CH3O
CH3O
CH3O
CH3O
(a)
OCH3
OCH3
CH3O
CH3O
9a (cis-E) or 9b (trans-E)
CH3O
CH3O
3a or 3b
Ph
I
7
5
b
OCH3
OCH3
a
c
e
I
R
(e)
d
CH3O
CH3O
CH3O
CH3O
(b)
OCH3
CHO
OCH3
CH3O
CH3O
CH3O
CH3O
9a, 9c (cis-Z)
Ph
8 R = CH2OH
4 R = CHO
(d)
3a (cis-E), 3b (trans-E)
OCH3
OCH3
Scheme 1. Reagents and conditions: (a) CH„CCH2OH, Pd(PPh3)4, CuI, piperidine, rt,
1.5 h, 81%; (b) H2, cat. Lindlar, Et3 N, AcOEt, rt, 1 h, 71%; (c) Et2Zn, CH2I2, CH2Cl2, 0 °C
to rt, 5 h, 60%; (d) TPAP, NMO, CH2Cl2, rt, 30 min, 80%; (e) CrCl2, CHI3, THF, 0 °C,
1.5 h, 56% (3a/3b 8:2).
4
Scheme 2. Reagents and conditions: (a) PhB(OH)2, Pd(PPh3)4, EtONa, THF, reflux,
5 h; (b) PhCH2P+Bu3BrÀ, n-BuLi, THF, À30 °C, 30 min then À78 °C, 4, 30 min.

1320
N. Ty et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1318–1322
the (cis,E)-cyclopropyl-vinyl derivative 9a displayed modest anti-
proliferative activity (IC50 = 6.49 M). Thus, the discussion will fo-
cus on the results obtained with the HCT-116 cell line.
Of the two cis-cyclopropyl-vinyl compounds evaluated, 9a and
9c, only the (cis-E)-isomer 9a which retained the configuration
(Z,E) of the lead compound 2, exhibited some inhibitory activity
l
CH3O
CH3O
11
OCH3
(a)
COOR
against tubulin assembly (IC50 = 5.3
lM), although weaker than
COOR
those of CA4 (IC50 = 0.79 M) and diene 2. The fact that compound
l
CH3O
CH3O
CH3O
CH3O
+
9a is a racemic mixture and that only one enantiomer has the
structural features required to interact with tubulin, may ac-
count—at least in part—for its moderate potency (discussed below).
This result validates the hypothesis that a cis-cyclopropyl group
could efficiently replace a Z-olefinic bond in this dienic derivatives
series. It is not surprising that compound 9a showed a lower cyto-
OCH3
OCH3
12a R = C2H5
10a R = H
12b R = C2H5
10b R = H
(b)
(b)
(c) or
(c) or
(d)
(d)
O
toxic potency in comparison to CA4 (IC50 = 0.53 lM vs 0.003 lM),
Ar
N
H
since the same trend was observed with our previously reported
vinylogous CA4 analogues.14 Nevertheless, compounds which dis-
play potent ability to inhibit tubulin assembly while showing lim-
ited cytotoxicity may be of high interest in the search of new
potent VDAs.31
H
N
CH3O
CH3O
CH3O
CH3O
O
Ar
OCH3
OCH3
13a,14a,15a,18a
13b,14b,18b
OR
Replacement of the E-double bond of compound 9a by an amide
group (13a) was detrimental to tubulin interaction, nevertheless
did not affect cytotoxicity (IC50 = 0.68 lM). Additionally, attempts
OR
N
Ar
were directed at replacing the phenyl group by different aromatic
moieties to explore spatial and electronic requirements. Com-
pounds 16a and 17a, respectively, bearing a phenolic or an isovan-
illic ring, did not interact with tubulin and were devoid of cytotoxic
activity. The pyridyl analogue 18a, which might generate new
interactions with tubulin, exhibited weak cytotoxicity
OCH3
18a,18b
13a,13b 14a,14b
15a
R = TBDMS
R = TBDMS
(e)
(e)
16a,16b
R = H
17a
R = H
Scheme 3. Reagents and conditions: (a) N2CHCO2C2H5, Cu(acac)2 in several
batches, ClCH2CH2Cl, reflux, 7 h; (b) 2 N NaOH, MeOH: 50 °C, 7 h for 12a; rt, 7 h
for 12b; (c) ArNH2, EDCI, HOBt, CH2Cl2, rt, 3–4 h; (d) (COCl)2, DMF, CH2Cl2, rt, 3 h,
evaporation then 3-aminopyridine, Et3 N, CH2Cl2, rt, 2.5 h; (e) TBAF, THF, rt, 2 h.
(IC50 = 2.37 lM) but without any microtubule disrupting activity.
Amide compounds 13b and 18b bearing a trans-cyclopropyl moi-
ety were essentially inactive as tubulin inhibitors.
In order to investigate the possible binding modes of the sub-
ject compounds with tubulin, docking studies of two representa-
tive analogues (9a and 13a) were performed by using the
reported high-resolution crystal structure of the tubulin/DAMA-
colchicine complex.32 All the simulations were carried out using
the GOLD docking tool (Genetic Optimisation for Ligand Docking)
algorithm.33 We firstly docked the CA4 in the colchicine binding
site of tubulin and we ensured that a hydrogen bond was favored
with Cysb241, a key residue in the binding of colchicine and
CA4.34
First, the possible binding modes of the two enantiomeric
forms of the cyclopropyl derivative 9a with tubulin were investi-
gated. Docking of the (S,S)-isomeric form of 9a resulted in a single
conformational mode correctly mimicking the CA4 within the
tubulin binding site (Figs. 3A and 4A). On the other hand, docking
of the 9a-(R,R) isomer yielded two conformer clusters. In the lar-
ger group of conformers (8/10), while the trimethoxyphenyl
(TMP) ring overlaps well with that of CA4, a different orientation
was taken by the remaining part of the molecule, beginning from
the cyclopropyl ring (Figs. 3B and 4B). Then, the different poses of
both isomeric forms of 9a were assessed using quantitative crite-
ria. Thereby, we extracted the fitness score provided by GoldScore
for each of the ten first poses in both isomers. For the isomer 9a-
(S,S) the ten poses had similar scores (comprised between 44 and
48) higher than those of CA4 (35–41). For the 9a-(R,R) isomer,
while eight poses had almost the same score values (35–37),
the two other positions superimposing with CA4 markedly got
lower scores (30 and 32). This is an additional proof of the bad
binding of 9a-(R,R) in the colchicine site. These results suggest
that the isomer 9a-(S,S) docked in a correct manner in the colchi-
Table 1
Inhibition of tubulin polymerization (ITP) and cell proliferation by synthesized
cyclopropyl derivatives.
H
N
CH3O
CH3O
CH3O
CH3O
O
Ph
X
R
OCH3
OCH3
9a, 9c
13, 16, 17, 18
Compound
X
R
ITPa IC50
(l
M)
Cytotoxicity IC50 (l
M)b
HCT116
MCF7
9a (cis,E)
9c (cis,Z)
13a (cis)
13b (trans)
16a (cis)
17a (cis)
18a (cis)
18b (trans)
CA4
—
—
CH
—
—
H
H
H
5.3
38
>100
92
>100
82
>100
>100
0.79
0.5
0.53
nac
0.68
3.35
na
6.49
na
na
na
na
CH
C–OH
C–OH
N
OCH3
H
na
na
na
2.37
na
0.003
0.50
N
H
na
ndd
2 (Z,E)e
a
Compound concentration inhibiting 50% of about 2.0 mg/mL microtubular
protein assembly. For CA4 and derivatives 9a and 9c values are the average of three
experiments.
b
Compound concentration required to inhibit human tumor cells proliferation
by 50% after 72 h incubation. Values are the average of experiments in duplicate.
c
Na (not active): inhibition of cell growth did not exceed 20% at concentration of
cine binding site and, furthermore, with
interaction.
a stable binding
1
l
M (triplicate).
d
Nd: IC50 values could not be determined, as growth inhibition did not exceed
Docking of the isomer 9a-(S,S) revealed a consistent set of
recurring interactions with the tubulin (Fig. 3A). Two methoxy O
50%. 25% inhibitory concentration: IC25 = 0.004
l
M.
e
Values from Ref 14. ITP value for CA4: 1.2
lM.

N. Ty et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1318–1322
1321
Figure 3. Proposed binding mode of both cyclopropyl isomers of 9a in the colchicine binding site of tubulin. The ten best scored poses (A) for isomer 9a-(S,S), (B) for isomer
9a-(R,R), Yellow ribbons are b sheets, red ribbons are -helixes.
a
rently under investigation to identify new vascular disrupting
agents.
Acknowledgements
This work was financially supported by the Centre National de
la Recherche Scientifique (CNRS) and the Institut Curie (Pro-
gramme Incitatif et Coopératif: angiogenèse tumorale). The
authors thank Drs. Claude Monneret (CNRS, Institut Curie), Alain
Commerçon and Patrick Mailliet (Aventis Pharma, Medicinal
Chemistry Department, Vitry-sur-Seine, France) for helpful sugges-
tions and discussions. We acknowledge Mme Sylvie Dubruille for
HPLC analysis (CNRS, Institut Curie) and Dr Thierry Cresteil
(ICSN-CNRS UPE 2301, Gif sur Yvette, France) for accurate
in vitro cytotoxicity evaluation. Aventis Pharma is acknowledged
for a doctoral grant to J.K.
Figure 4. Superimposition of both cyclopropyl isomers of 9a (blue) and CA4 (red) in
the colchicine binding site of tubulin. (A) An overlay of CA4 and 9a-(S,S). (B) An
overlay of CA4 and the most representative binding position of 9a-(R,R).
Supplementary data
Experimental procedures, spectroscopic data for new com-
pounds and molecular modeling of derivative 13a, are available
on line. Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.bmcl.2009.01.062.
atoms of the A-ring participate in a hydrogen bond network with
the Cysb241. The TMP moiety occupies a hydrophobic pocket
bounded by Leu248, Val318, and Ile358 from the b sub-unit, while
the B-phenyl ring establishes hydrophobic contacts with Ala
a180
References and notes
and Val 181. Moreover, the lower part of the phenyl ring overlays
a
1. Denekamp, J. Int. J. Radiation Biol. 1991, 60, 401.
well with the methoxy methyl group of the CA4 B-ring reported as
an important hydrophobic center33 for the interaction of many col-
chicine site inhibitors within tubulin. It is noteworthy that the re-
gion at the vicinity of the cyclopropyl ring is an open space region
that should authorize steric extensions.35
In an effort to understand the reason of the discrepancy be-
tween the design of the amide derivatives and their poor biological
activity, we undertook the docking study of the benzanilide 13a
(see Supplementary data). As the amide function adopts a s-trans
conformation, we infer that the inability for 13a and the other
(cis)-cyclopropyl-amide derivatives to inhibit the tubulin assembly
is related to the electronic effects of the amide function.
2. (a) Arap, W.; Pasqualini, R.; Ruoslahti, E. Curr. Opin. Oncology 1998, 10, 560; (b)
Pasquier, E.; André, N.; Braguer, D. Curr. Cancer Drug Targets 2007, 7, 566.
3. (a) Patterson, D. M.; Rustin, G. J. S. Clin. Oncol. 2007, 19, 443; (b) Lippert, J. W., III
Bioorg. Med. Chem. 2007, 15, 605; (c) Hinnen, P.; Eskens, F. A. L. M. Br. J. Cancer
2007, 96, 1159; (d) Cai, S. X. Recent Patents Anticancer Drug Discov. 2007, 2, 79.
4. West, C. M. L.; Price, P. Anti-cancer Drugs 2004, 15, 179.
5. Horsman, M. R.; Siemann, D. W. Cancer Res. 2006, 66, 11520.
6. Pettit, G. R.; Singh, S. B.; Hamel, E.; Lin, C. M.; Alberts, D. S.; Garcia-Kendall, D.
Experientia 1989, 45, 209.
7. Kanthou, C.; Tozer, G. M. Blood 2002, 99, 2060.
8. Vincent, L.; Kermani, P.; Young, L. M.; Cheng, J.; Zhang, F.; Shido, K.; Lam, G.;
Bompais-Vincent, H.; Zhu, Z.; Hicklin, D. J.; Bohlen, P.; Chaplin, D. J.; May, C.;
Rafi, S. J. Clin. Invest. 2005, 115, 2992.
9. Galbraith, S. M.; Chaplin, D. J.; Lee, F.; Stratford, M. R.; Locke, R. J.; Vojnovic, B.;
Tozer, G. M. Anticancer Res. 2001, 21, 93.
In summary, we have synthesized and biologically evaluated a
novel series of CA4 analogues related to the 1,4-diarylbutadiene
2. Our results validate the feasibility of replacing a cis double
bond with a cis-cyclopropyl moiety. But, contrary to our expecta-
tion, the amide moiety was not a good surrogate for the trans
double bond of vinylogous CA4 analogues. We can presume that
the electronic nature, and not the geometry of the amide linkage,
is unsuitable for direct tubulin binding. The docking results ob-
tained for derivatives 9a and 13a correlate well with the observed
biological data. On the basis of the qualitative and quantitative
docking study of the racemic derivative 9a, we could assure that
the (S,S)-cyclopropyl enantiomeric form displays the highest
binding affinity for tubulin. Therefore, asymmetric synthesis of
both the pure enantiomers of 9a and other congeners are cur-
10. Tozer, G. M.; Kanthou, C.; Baguley, B. C. Nat. Rev. Cancer 2005, 5, 423. and
references cited therein.
11. For reviews on CA4 analogues see (a) Nam, N.-H. Curr. Med. Chem. 2003, 10,
1697; (b) Hsieh, H. P.; Liou, J. P.; Mahindroo, N. Curr. Pharm. Design 2005, 11,
1655; (c) Tron, G. C.; Pirali, T.; Sorba, G.; Pagliai, F.; Busacca, S.; Genazzani, A. J.
Med. Chem. 2006, 49, 3033.
12. (a) Lawrence, N. J.; Patterson, R. P.; Ooi, L.-L.; Cook, D.; Ducki, S. Bioorg. Med.
Chem. Lett. 2006, 16, 5844; (b) Romagnoli, R.; Baraldi, P. G.; Carrion, M. D.;
Lopez Cara, C.; Cruz-Lopez, O.; Preti, D.; Tolomeo, M.; Grimaudo, S.; Di Cristina,
A.; Zonta, N.; Balzarini, J.; Brancale, A.; Sarkar, T.; Hamel, E. Bioorg. Med. Chem.
2008, 16, 5367.
13. Provot, O.; Giraud, A.; Peyrat, J.-F.; Alami, M.; Brion, J.-D. Tetrahedron Lett. 2005,
46, 8547.
14. Kaffy, J.; Pontikis, R.; Florent, J. C.; Monneret, C. Org. Biomol. Chem. 2005, 3,
2657.
15. See, for example (a) Lagu, B.; Lebedev, R.; Pio, B.; Yang, M.; Pelton, P. Bioorg.
Med. Chem. Lett. 2007, 17, 3491; (b) Carpita, A.; Ribecai, A.; Rossi, R.; Stabile, P.
Tetrahedron 2002, 58, 3673.

1322
N. Ty et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1318–1322
16. Cis or trans refer to the cyclopropane ring stereochemistry, E or Z to the olefinic
bond configuration. All the synthesized cyclopropyl derivatives are cis or trans
racemates.
(s, 9H, 3Â OCH3), 240 (m, 1H, H-a), 2.27 (m, 1H, H-c), 1.39 (m, 1H, H-b),
1.03 (m, 1H, H-b); HRMS (DCI/NH3): calcd for C20H23O3 (MH+) 311.1647,
found 311.1643.
17. Konno, K.; Ojima, K.; Hayashi, T.; Tanabe, M.; Takayama, H. Chem. Pharm. Bull.
1997, 45, 185.
24. Ramacciotti, A.; Fiaschi, R.; Napolitano, E. Tetrahedron: Asymmetry 1996, 7,
1101.
18. Kato, M.; Kido, F.; Wu, M.-D.; Yoshikoshi, A. Bull. Chem. Soc. Jpn 1974, 47, 1516.
19. Cooper, P. D. Can. J. Chem. 1970, 48, 3882.
25. Doyle, M. P. Chem. Rev. 1986, 86, 919.
26. Majchrzak, M. W.; Kotelko, A. Synthesis 1983, 6, 469.
27. The structures of the cyclopropyl-carboxylic acids and esters 10 or 12 were
assigned on the basis of the values of the coupling constant between H-a and
H-c. For derivatives 10a and 12a, the high value of Jac (9.4 and 9.2) is consistent
with a cis-disposition of the substituents on the cyclopropyl ring.
28. Hennessy, E. J.; Buchwald, S. L. J. Am. Chem. Soc. 2003, 125, 12084.
29. Watanabe, T.; Suzuki, T.; Umezawa, Y.; Takeuchi, T.; Otsuka, M.; Umezawa, K.
Tetrahedron 2000, 56, 741.
20. The cis-cyclopropane structure (J = 8.7 Hz) for aldehyde 4 was confirmed by
NOESY analysis and by analogy with 1H NMR data of the phenyl analogue.21
Moreover, 1H NMR values for the aldehyde proton allowed easy attribution of
the cis or trans stereochemistry for the cyclopropane ring (d = 8.66 ppm,
J = 6.9 Hz/d = 9.33 ppm, J = 4.6 Hz, respectively).
21. Yamaguchi, K.; Kazuta, Y.; Abe, H.; Matsuda, A.; Shuto, S. J. Org. Chem. 2003, 68,
9255.
22. For the cyclopropyl-vinyl derivatives synthesized (vinyl iodides 3a and 3b,
styryl derivatives 9a and 9b) the stereochemistry cis or trans of the
cyclopropane ring was confirmed by 1H NMR chemical shifts analysis. For
derivatives 3a and 9a with a cis-cyclopropane unit, the cyclopropyl protons H-a
and H-c resonate at lower field than in the trans-compounds (ꢀ0.35 and
0.20 ppm, respectively), while the olefinic proton H-d is shielded (ꢀ0.35 ppm).
For atom numbering see Scheme 1.
30. The 1H NMR spectra of the cis-cyclopropyl-amide compounds show
a
downfield shift (approximately 0.30–0.35 ppm) of the H-c protons relative to
the same protons of the trans-isomers.
31. (a) Hadfield, J. A.; Gaukroger, K.; Hirst, N.; Weston, A. P.; Lawrence, N. J.;
McGown, A. T. Eur. J. Med. Chem. 2005, 40, 529; (b) Hadimani, M. B.; Hua, J.;
Jonklaas, M. D.; Kessler, R. J.; Sheng, Y.; Olivares, A.; Tanpure, R. P.; Weiser, A.;
Zhang, J.; Edvarsen, K.; Kane, R. R.; Pinney, K. G. Bioorg. Med. Chem. Lett. 2003,
13, 1505.
23. Characterization of compounds 9a and 9c. Compound 9a: colorless oil, 1H
NMR (300 MHz, CDCl3): d 7.22 (m, 5H, Ph), 6.53 (d, 1H, J = 15.7 Hz, H-e),
6.49 (s, 2H, H-2, H-6), 5.54 (dd, 1H, J = 15.7, 9.4 Hz, H-d), 3.84 (s, 6H, 2Â
OCH3), 3.82 (s, 3H, OCH3), 2.40 (m, 1H, H-a), 1.99 (m, 1H, H-c), 1.35 (m, 1H,
H-b), 1.04 (m, 1H, H-b); HRMS (DCI/CH4): calcd. for C20H23O3 (MH+)
311.1647, found 311.1642; compound 9c: colorless oil, 1H NMR (300 MHz,
CDCl3): d 7.45 (m, 2H, Ph), 7.36 (m, 2H, Ph), 7.23 (m, 1H, Ph), 6.47 (s, 2H, H-
2, H-6), 6.36 (d, 1H, J = 11.6 Hz, H-e), 5.01 (dd, 1H, J = 11.6, 9.6 Hz, H-e), 3.84
32. Ravelli, R. B.; Gigant, B.; Curmi, P. A.; Jourdain, I.; Lachkar, S.; Sobel, A.;
Knossow, M. Nature 2004, 428, 198.
33. Jones, G.; Willett, P.; Glen, R. C. J. Mol. Biol. 1995, 24(5), 43.
34. Nguyen, T. L.; McGrath, C.; Hermone, A. R.; Burnett, J. C.; Zaharevitz, D. W.; Day,
B. W.; Wipf, P.; Hamel, E.; Gussio, R. J. Med. Chem. 2005, 48, 6107.
35. Such cavity had already been reported. See, for example Bellina, F.;
Cauteruccio, S.; Monti, S.; Rossi, R. Bioorg. Med. Chem. Lett. 2006, 16, 5757.