Synthesis and biological activity of mustard derivatives of combretastatins

Article abstract of DOI:10.1016/j.bmcl.2005.05.052

A series of chimeric compounds bearing the combretastatin and the nitrogen mustard cores were synthesized. All the compounds were cytotoxic and inhibited tubulin polymerization. When combretastatin was joined to chlorambucil via an ester linkage, the resu

Full text of DOI:10.1016/j.bmcl.2005.05.052

Bioorganic & Medicinal Chemistry Letters 15 (2005) 3551–3554
Synthesis and biological activity of mustard derivatives
of combretastatins
Beatrice Coggiola, Francesca Pagliai, Gianna Allegrone,
Armando A. Genazzani and Gian Cesare Tron*
`
DiSCAFF, Universita del Piemonte Orientale, Via Bovio, 6, 28100 Novara, Italy
Received 24 March 2005; revised 13 May 2005; accepted 17 May 2005
Available online 15 June 2005
Abstract—A series of chimeric compounds bearing the combretastatin and the nitrogen mustard cores were synthesized. All the
compounds were cytotoxic and inhibited tubulin polymerization. When combretastatin was joined to chlorambucil via an ester link-
age, the resultant compound proved to be significantly more potent than the two compounds put together. When combretastatin
was joined to nitrogen mustard via an ether linkage or when a true hybrid was synthesized, loss of potency was observed. None-
theless, these latter compounds appeared to be more efficacious and surprisingly were able to inhibit tubulin depolymerization at
high concentrations.
ꢀ 2005 Elsevier Ltd. All rights reserved.
1. Introduction
in rapid necrosis of the tumor, it has also been shown
that a resistant rim of cells remains after treatment. This
effect is most likely due to the preponderant contribu-
tion of normal vasculature in the blood flow
to the periphery of the tumoral mass.5 Therefore, the
tumoral mass is potentially capable of undergoing rapid
re-growth upon discontinuation of the drug,5,6 which
suggests that a multi-pharmacy approach would be
beneficial even with 1.2
It is generally accepted that most anticancer therapies
should be based on the use of more than one compound.
Such a strategy could improve therapeutic benefits both
because it reduces drug resistance and because it can
increase efficacy while decreasing side effects. Further-
more, it has been shown that some antitumoral agents
are not a substrate for multidrug resistance (MDR) pro-
teins and therefore might be ideal for this drug cocktail
approach. For instance, it has been shown that combre-
tastatin A4 (CA-4, 1) is able to act against MDR posi-
tive tumoral cell lines.1 CA-4 acts on microtubules and
prevents the polymerization of tubulin,2 binding to the
same site as colchicine on b-tubulin.3 Indeed, CA-4
phosphate, its water soluble pro-drug, has recently
entered clinical trials.4 Although tubulin is a ubiquitous
protein, it has been shown that 1 displays a remarkable
preference for neovasculature, and therefore, besides
being a cytotoxic agent, it can also be classified as an
angiostatic/toxic drug. Nonetheless, although in preclin-
ical studies, administration of CA-4 phosphate resulted
We hypothesized that the use of 1 in conjunction with a
classical anticancer agent might overcome the multidrug
resistance induced by the latter and the inability of the
former to inhibit blood flow to the periphery of the tu-
moral mass,5 thereby generating a more efficacious tool.
As a proof of principle, we chose alkylating agents (i.e.,
nitrogen mustards) and decided to generate novel chime-
ras. It has been shown that isosteric substitution of a
hydroxyl group in 1 with an amino group results in a
compound (AC 7739, 2) that displays an increase in
cytotoxicity, while maintaining an identical pharmaco-
dynamic profile.7 Similarly, compounds with only a
dimethylamino group in the para position of ring B of
1 maintain their cytotoxic potential, suggesting that
the presence of a primary amino group is not fundamen-
tal.8 We have therefore synthesized a nitrogen mustard
derivative on the amino group of AC7739 4. Alongside,
to explore further the potential of the combretastatin/
nitrogen mustard combination, we have also synthesized
Keywords: Alkylating agents; Antiproliferative activity; Combreta-
statin A-4; Chlorambucil; Tubulin polymerization.
*
Corresponding author. Tel.: +39 0321 375857; fax: +39 0321
357821; e-mail: tron@pharm.unipmn.it
These authors contributed equally to this work.
0960-894X/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.05.052

3552
B. Coggiola et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3551–3554
two chimeras that bear the basic skeleton of 1 joined
with 2-[bis(2-chloroethyl)amino]-1-ethanol via an ether
linkage (5) or with chlorambucil via an ester linkage (5).
3. Results and discussion
To analyze the relative cytotoxicity of the synthesized
compounds, SH-SY5Y neuroblastoma cell cultures12
were treated with increasing concentrations of com-
pounds and grown for 48 h. 1 displayed a concentra-
tion-dependent cytotoxicity, with an IC50 of approx.
1.5 0.28 nM (Table 1, Fig. 1). This is consistent with
previous data on the same cell type and demonstrates
that neuroblastoma cells are particularly susceptible to
this chemotherapeutic insult.13 When the hybrid mole-
cule 4 was tested, there was an approximately 500-fold
loss of toxicity. We then tested whether linking combre-
tastin to the nitrogen mustards via longer chains might
have restored cytotoxic activity. Indeed, 5, where an
ether chain was used to bridge the two functional
groups, was approximately 3-fold more potent than 4.
Last, an ester linkage was used to join combretastatin
to the nitrogen mustard (6). Surprisingly, 6 was signifi-
cantly more potent than 1, with an IC50 value of
0.64 0.11 nM. To establish whether this increase in po-
tency was due to an increased tubulin binding of 6 or
was attributable to the presence of two separate func-
tional entities, we incubated neuroblastoma cells with
chlorambucil, 1, or a combination of the two. In SH-
SY5Y cells, chlorambucil was a poor cytotoxic agent
up to concentrations of 10 lM (Fig. 1B). Nonetheless,
when this anticancer agent was added along with com-
2. Chemistry
Compounds9 were prepared starting from 1 and its
amino derivative 2 (Scheme 1), synthesized as previously
described.10,11
Compound 2 was double alkylated using chloro propa-
nol to yield the desired diol 3. Subsequent treatment
with methansulfonyl chloride gave the dimesylate deriv-
ative. The latter was converted in situ to dichloride by
treatment with lithium chloride in DMF to give 4.
Compound 5, characterized by the presence of an ethyl
spacer between the phenolic ring and the N-mustard res-
idue, was obtained, albeit in poor yield (10%), by react-
ing CA-4 with tris(2-chloroethyl)amine hydrochloride in
DMF using potassium tert-butylate as base.
Compound 6 was obtained by coupling 1 with chloram-
bucil using the classical EDCI/DMAP protocol. The
correct stereochemical assignment of the double bond
1
was determined by H NMR and, when required, UV
spectra (data not shown).
Attempts were also made to synthesize a chimeric deriv-
ative bearing only the amino group in the para position.
The attempt to obtain an active molecule, though, failed
since the compound displayed remarkable capacity to
isomerize from cis to trans during the alogenation
protocol.
Table 1. Cytotoxicity of combretastatin and mustard derivatives on
SH-SY5Y neuroblastoma cells
Compound 1
3
4
5
6
IC50 (nM) 1.5 0.28 >10,000 860 220 230 72 0.64 0.11
Scheme 1. Reagents and conditions: (a) CaCO3, 2-chloroethanol, water, reflux; (b) methansulfonyl chloride, TEA, DMF, rt. and then LiCl, 80ꢁC;
(c) tris(2-chloroethyl)amine, potassium tert-butylate, DMF, rt; (d) chlorambucil, EDCI, DMAP, CH2Cl2, rt.

B. Coggiola et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3551–3554
3553
Figure 1. Concentration–response curves of combretastin, chlorambucil, and their hybrids. n = 12–36 from at least three separate experiments,
bretastatin, a modest, but not statistically significant,
potentiation was observed. To test whether the relative
rank order of potency of 1 and 6 was maintained in
other cancerous cell lines, we performed concentra-
tion–response curves in mesothelioma (REN) and mas-
tocyte (RBL) cell lines.12 Surprisingly, in these latter
cell lines, 1 displayed cytotoxicity higher than that of
6, although 6 remained relatively potent (IC50 below
3 nM in both REN and RBL cells). These data hint at
the idea that increased potency observed with 6 in
SH-SY5Y cells is cell specific and is attributable to the
chlorambucil moiety. Although our data are not conclu-
sive, the difference between the cell lines might reside
either in their relative esterase activity or in the ability
of 6 to display, alongside tubulin binding, a different
mechanism of action.
forms compared to the concentrations required in cell
viability experiments15 and we therefore chose concen-
trations 50-fold higher than the respective IC50s. Indeed,
at these concentrations, compounds 1, 4, 5, and 6, but
not chlorambucil, induced a significant shift of tubulin
from the polymerized to the unpolymerized form
(Fig. 2). Such experiments would suggest that all the
synthesized compounds retain the original mechanism
of action of combretastatin. Yet, when cells were treated
for the same length of time with 5-fold higher concentra-
tions, compound 4 and compound 5 induced tubulin
polymerization. In other words, the addition of chlor-
ambucil moiety induced these compounds to change
from polymerization inhibitors (colchicine-like drugs)
to depolymerization inhibitors (taxol-like drugs).
Although we have not investigated further the mecha-
nism by which this shift in biological activity takes place,
we propose that compounds 4 and 5 are attracted to
To establish the mechanism of action of these com-
pounds, we performed a cytofluorimetric analysis using
propidium iodide to label DNA. Agents that act on
tubulin, such as 1, are known to induce a selective block
in G2/M, most likely due to the disruption of mitotic
spindle architecture. Indeed, 1, 4, 5, and 6 all induced
a selective increase of the G2/M phase, suggesting that
their cytotoxic nature is attributable to tubulin binding
(data not shown). Such a cytofluorimetric signature
was not observable with chlorambucil, which appears
to induce a modest cell cycle arrest in the G1 phase (data
not shown).
To establish conclusively whether all these agents act on
tubulin, an intact-cell polymerization assay was em-
ployed. To measure the degree of tubulin polymeriza-
tion, we adopted the method described by Minotti
et al.14 as modified by Tron et al.13 In brief, cells were
grown for 24 h in the presence of drugs and were then
lysed in the presence of paclitaxel, an agent that prevents
further depolymerization. Subsequently, polymerized
and unpolymerized tubulin were separated by centrifu-
gation. It has been previously reported that shorter incu-
bation times, but higher concentrations of drugs, are
required to observe changes in tubulin polymerization
Figure 2. Inhibition of tubulin polymerization or depolymerization by
combretastatin, chlorambucil, and their hybrids. P represents the
pelletable fraction (polymerized tubulin) and S represents the soluble
fraction (unpolymerized tubulin). Western blots are representative of
at least three experiments that gave similar results.

3554
B. Coggiola et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3551–3554
–NCH2CH2OH). Compound 4: yellow oil (49%). MS
(ESI) m/z: 440 (M+H)+; 1H NMR (CDCl3) d: 6.98 (dd,
1H, arom., J = 8.5/1.9 Hz), 6.86 (br s, 1H, arom.), 6.76 (d,
1H, arom., J = 8.5 Hz), 6.48 (s, 2H, arom.), 6.44 (s, 2H,
olefinic), 3.83 (s, 3H, OMe), 3.82 (s, 3H, OMe), 3.70 (s,
6H, OMe), 3.42 (br s, 8H, –NCH2CH2Cl); Compound 5:
deep yellow oil (10%). MS (ESI) m/z: 484 (M+H)+; 1H
NMR (CDCl3) d: 6.86 (dd, 1H, arom., J = 8.2/1.9 Hz),
6.80 (d, 1H, arom., J = 1.9 Hz), 6.76 (d, 1H, arom.,
J = 8.2 Hz), 6.51 (s, 2H, arom.), 6.47 (d, 1H, olefinic,
J = 12.1 Hz), 6.41 (d, 1H, olefinic, J = 12.1 Hz), 3.85 (t,
–OCH2CH2N, J = 5.8 Hz), 3.83 (br s, 6H, OMe), 3.69 (s,
6H, OMe), 3.51 (t, –NCH2CH2Cl, J = 7.1 Hz), 2.99–2.94
(m, –OCH2CH2N + NCH2CH2Cl); Compound 6: white
solid (89%). mp 118–119.5 ꢁC; MS (ESI) m/z: 624
(M+Na)+; 1H NMR (CDCl3) d: 7.10 (dd, 1H, arom.,
J = 8.0/1.9 Hz), 7.08 (d, 1H, J = 8.0 Hz), 6.99 (d, 1H,
arom., J = 1.9 Hz), 6.84 (d, 2H, arom., J = 8.2 Hz), 6.62
(d, 2H, arom., J = 8.5 Hz), 6.50 (s, 2H, arom.), 6.44 (s, 2H,
olefinic), 3.82 (s, 3H, OMe), 3.80 (s, 3H, OMe), 3.70 (s,
6H, OMe), 3.69–3.60 (m, –NCH2CH2Cl), 2.62 (t, COCH2-
CH2CH2Ph, J = 7.4 Hz), 2.53 (t, COCH2CH2CH2Ph, J =
7.4 Hz), (q, COCH2CH2CH2Ph, J = 7.7 Hz). Purity of
target compounds was checked by HPLC analysis using a
Phenomenex Luna (250 · 4.6 mm, 5 lm particle size)
column on a Shimadzu HPLC system. Solvents for the
separation were: solvent A: water; solvent B: acetonitrile
at a flow rate of 1 ml/min and a sample injection volume of
210 ll. For compound 6, eluants A and B were delivered
isocratically at a 5:95 ratio. For compounds 4 and 5, a
linear gradient was used from 70:30 to 30:70 in 30 min. All
three compounds displayed a purity of at least 95%.
10. Gaukroger, K.; Hadfield, J. A.; Hepworth, L. A.; Law-
rence, N. J.; McGown, A. T. J. Org. Chem. 2001, 66, 8135.
11. Pinney, K. G.; Mejia, M. P.; Villalobos, V. M.; Rosen-
quist, B. E.; Pettit, G. R.; Verdier-Pinard, P.; Hamel, E.
Biorg. Med. Chem. 2000, 8, 2417.
tubulin via their active combretastatin group. If high
enough concentrations are placed in the vicinity of a
microtubule, then the nitrogen mustard will alkylate
tubulin or an accessory protein to prevent depolymeriza-
tion. Indeed, it is surprising that although compounds 4
and 5 are significantly less potent than 1 and 6, their effi-
cacy appears to be even greater. In cells treated with
combretastatin, there appears to be a small, but signifi-
cant, residual cell viability (approx. 10%), while no cells
appear to resist at high concentrations of 4 or 5 (Fig. 1).
It is exciting to speculate that capacity of these com-
pounds to act either by increasing or decreasing depoly-
merization in a concentration-dependent manner may
be responsible for this increased efficacy. Indeed, it has
been previously suggested that antitumor therapy with
a combination of polymerizing and depolymerizing
agents provides an added value.2
References and notes
1. McGown, A. T.; Fox, B. W. Cancer Chemother. Pharma-
col. 1990, 26, 79.
2. Tozer, G. M.; Kanthou, C.; Parkins, S. A.; Hill, S. A.
Int. J. Exp. Pathol. 2002, 83, 21.
3. Lin, C.; Singh, S. B.; Chu, P. S.; Dempcy, R. O.; Schmidt, J.
M.; Pettit, G. R.; Hamel, E. Mol. Pharmacol. 1988, 34, 200.
4. Dowlati, A.; Robertson, K.; Cooney, M.; Petros, W. P.;
Stratford, M.; Jesberger, J.; Rafie, N.; Overmoyer, B.;
Makkar, V.; Stambler, B.; Taylor, A.; Waas, J.; Lewin, J.
S.; McCrae, K. R.; Remick, S. C. Cancer Res. 2002, 62,
3408.
5. (a) Pedley, R. B.; Hill, S. A.; Boxer, G. M.; Flynn, A. A.;
Boden, R.; Watson, R.; Dearling, J.; Chaplin, D. J.;
Begent, R. H. J. Cancer Res. 2001, 61, 4716; (b) Chaplin,
D. J.; Hill, S. A. Int. J. Radiat. Oncol. Biol. Phys. 2002, 54,
1491.
12. The SH-SY5Y human neuroblastoma cell line and the
RBL cell line were obtained from ATCC (LGC Promo-
chem Teddington, UK), The REN human mesothelioma
cells16 were a kind gift of Professor G. Gaudino
13. Tron, G. C.; Pagliai, F.; Del Grosso, E.; Genazzani, A. A.;
Sorba, G. J. Med. Chem. 2005, 48, 3260.
6. West, C. M. L.; Price, P. Anti Cancer Drugs 2004, 15, 179.
7. Ohsumi, K.; Nakagawa, R.; Fukuda, Y.; Hatanaka, T.;
Morinaga, Y.; Nihei, Y.; Ohishi, K.; Suga, Y.; Akiyama,
Y.; Tsuji, Y. J. Med. Chem. 1998, 41, 3022.
14. Minotti, A. M.; Barlow, S. B.; Cabral, F. J. Biol. Chem.
1991, 266, 3987.
15. Prinz, H.; Ishii, Y.; Hirano, T.; Stoiber, T.; Camacho, G.
8. Cushman, M.; Nagarathnam, D.; Gopal, D.; Chakraborti,
A. K.; Lin, C. M.; Hamel, E. J. Med. Chem. 1991, 34,
2579.
J. A.; Schmidt, P.; Dussman, H.; Burger, A. M.; Prehn, J.
¨
Chem. 2003, 46, 3382.
¨
H. M.; Gunther, E. G.; Unger, E.; Umezawa, K. J. Med.
9. Compound 3: yellow oil (35%). MS (ESI) m/z: 404
1
(M+H)+; H NMR (CDCl3) d: 7.10 (m, 2H, arom.), 6.82
(d, 1H, arom., J = 8.8 Hz), 6.51 (d, 1H, olefinic,
J = 11.5 Hz), 6.45 (d, 1H, olefinic, J = 11.5 Hz), 6.44 (s,
2H, arom.), 3.86 (s, 3H, OMe), 3.82 (s, 3H, OMe), 3.70 (s,
6H, OMe), 3.45 (br t, 2H, –NCH2CH2OH), 3.08 (br s, 2H,
16. Smythe, W. R.; Kaiser, L. R.; Hwang, H. C.; Amin, K.
M.; Pilewski, J. M.; Eck, S. J.; Wilson, J. M.; Albelda, S.
M. Ann. Thorac. Surg. 1994, 57, 1395.