Synthesis, in vitro, and in vivo evaluation of phosphate ester derivatives of combretastatin A-4

Article abstract of DOI:10.1016/S0960-894X(03)00206-3

Combretastatin A-4 disodiumphosphate (CA4P), a prodrug formulation of the natural product combretastatin A-4 (CA4), is currently in clinical investigation for the treatment of cancer. In vivo, CA4P is rapidly enzymatically converted to CA4, a potent inhibitor of tubulin polymerization (IC₅₀=1-2 μM), and rapidly causes bloodflow shutdown in tumor tissues. A variety of alkyl and aryl di- and triesters of CA4P have been synthesized and evaluated as potential CA4 prodrugs and/or stable CA4P analogues.

Full text of DOI:10.1016/S0960-894X(03)00206-3

Bioorganic & Medicinal Chemistry Letters 13 (2003) 1505–1508
Synthesis, In Vitro, and In Vivo Evaluation of Phosphate Ester
Derivatives of Combretastatin A-4
Mallinath B. Hadimani,a Jianyi Hua,b M. Devan Jonklaas,a Raymond J. Kessler,a
Yezhou Sheng,b Adrian Olivares,a Rajendra P. Tanpure,a Aimee Weiser,a
Jianxing Zhang,a Klaus Edvardsen,b,* Robert R. Kanea,* and Kevin G. Pinneya,*
aDepartment of Chemistry & Biochemistry and Center for Drug Discovery, Baylor University, Waco, TX 76798-7348, USA
bDepartment of Cell and Molecular Biology, University of Lund, Lund, Sweden
Received 22 January 2003; revised 21 February 2003; accepted 27 February 2003
Abstract—Combretastatin A-4 disodiumphosphate (CA4P), a prodrug formulation of the natural product combretastatin A-4
(CA4), is currently in clinical investigation for the treatment of cancer. In vivo, CA4P is rapidly enzymatically converted to CA4, a
potent inhibitor of tubulin polymerization (IC50=1–2 mM), and rapidly causes bloodflow shutdown in tumor tissues. A variety of
alkyl and aryl di- and triesters of CA4P have been synthesized and evaluated as potential CA4 prodrugs and/or stable CA4P ana-
logues.
# 2003 Elsevier Science Ltd. All rights reserved.
An emerging area of cancer chemotherapy research is
focused upon the development of agents, both small
molecules as well as biologics, that selectively destroy
tumor vasculature.1 These compounds are commonly
referred to as vascular targeting agents (VTAs). Several
small molecules that show anti-tumor effects based on
this novel mechanism have been identified and are in
various stages of clinical development.2 These com-
pounds are prodrug formulations of tubulin-binding
anilines or phenols that, in prodrug form, do not bind
to tubulin. These drugs function biologically by selec-
tively shutting down blood flow to neoplastic cells while
leaving the blood supply to healthy cells intact.3 The
selectivity of VTAs for the microvessels of tumors may
reflect, in part, variability in the cytoskeletal make-up of
rapidly proliferating endothelial cells inherent to micro-
vessels feeding tumor cells versus the normally pro-
liferating endothelial cells of microvessels serving
healthy cells.4 Combretastatin A-4 prodrug (CA4P) is a
remarkable VTA that is a simple phosphate monoester
of the phenolic natural product combretastatin A-4
(CA4), which is isolated from the Combretum caffrum
willow tree in South Africa.5
Preclinical and clinical studies have revealed an extre-
mely short circulation half-life for the CA4P phosphate
prodrug.6 Accordingly, the synthesis of longer-lived
CA4 prodrugs are attractive, as they could potentially
shed light on the role of pharmacokinetics in CA4P’s
VTA activity. Additionally, although CA4P is widely
recognized as a prodrug of CA4, it is possible that
CA4P itself might exhibit some biological activity, per-
haps against a different target than CA4. In order to
address these questions we have synthesized and char-
acterized a variety of phosphate esters of CA4. The
syntheses and activities of these compounds are pre-
sented herein.
Combretastatin A-4, which was the common starting
material for the synthesis of each of the derivatives
reported herein, was synthesized following the method
reported by Pettit et al.7 The CA4 dialkylphosphate
triesters 1–3 (Scheme 1) were then readily available via
the carbon tetrachloride-mediated oxidative phosphor-
ylation of CA4 with the appropriate dialkyl H-phos-
phonate.8,9 These reactions afforded the desired
products in good yield, with minimal isomerization of
the somewhat labile cis-stilbene bond.
A second set of CA4 derivatives synthesized were the
anionic alkyl and aryl phosphate diesters 4–8 (Scheme
2). Once again starting with CA4, these compounds
*Corresponding author. Tel.: +1-254-710-4556 Fax: +1-254-710-
4628; e-mail: bob_kane@baylor.edu
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00206-3

1506
M. B. Hadimani et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1505–1508
nium hydroxide was used to remove the 2-cyanoethyl
protecting groups, affording the phosphate diesters 4–8
in reasonable overall yields (as calculated from the
starting alcohols; 35–45% overall based on CA4 con-
sumption).
A special phosphate diester, the phosphate-linked CA4
dimer 9, was also synthesized (Scheme 3). This com-
pound was readily available via the condensation of two
equivalents of CA4 with POCl3, followed by hydrolysis
of the resulting acid chloride.
Scheme 1. Synthesis of phosphate triesters 1–3.
were accessible in reasonable yield using standard
phosphoramidite coupling chemistry.10
One additional compound prepared for comparison to
the phosphate esters was the ether-linked alkyl sulfonate
salt 10 (Scheme 4). This compound was isolated in good
yield from the ring opening of 1,3-propane sultone by a
phenoxide anion derived from CA4.
Each target alcohol was initially converted into its
2-cyanoethyl-N,N-diisopropylamino phosphoramidite.
The phosphoramidites were then coupled with CA4
using 1H-tetrazole as an activator, followed by MCPBA
oxidation in situ to produce the 2-cyanoethyl protected
phosphate triesters. In the final step, alcoholic ammo-
The biological activity of each of the compounds syn-
thesized was then assayed. A standard in vitro MTT
assay was used to determine the relative cytotoxicity of
the compounds (Table 1).11,12 Two different periods of
drug exposure—1 h and 5 days–were examined to
determine the rapidity of the cytotoxic activity of the
compounds.
In comparison to CA4 or CA4P the analogues described
in this paper exhibited unremarkable cytotoxicities,13
although the phosphate-linked dimer did exhibit rela-
tively good activity (IC50=0.2 mM) against a colon
cancer cell line (HT-29) that is resistant to other com-
bretastatins. Compound 10, a non-hydrolyzable anionic
ether derivative of CA4, was completely inactive in the
cytotoxicity assay. This is not altogether surprising as it
is known14 that, unlike CA4, this compound does not
interfere with tubulin polymerization.
Scheme 3. Synthesis of CA4 phosphate dimer 9.
Scheme 2. Synthesis of phosphate diesters 4–8.
Scheme 4. Synthesis of CA4 trimethylenesulfonate 10.

M. B. Hadimani et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1505–1508
1507
Table 1.
inherent in the in vivo nature of the assay. We were
therefore encouraged to discover that the diester deri-
vatives 7 and 8 (at 100 mg/kg) were able to shut down
tumor vasculature at levels comparable to CA4P. Fig-
ure 1 is an example from a vascular shutdown experi-
ment that demonstrates the dramatic difference in
vascular volume found in treated and control tumor
slices. In this experiment fluorescence is found in
actively vascularized tissue; the lack of fluorescence in
the treated tissue demonstrates the decrease in func-
tional vasculature.15
Compd
IC50
(1 hr; mM)
IC50
(5 day; mM)
Vascular shutdowna
(100 mg/kg)
CA4
CA4P
1
2
3
4
5
6
7
8
1.05
0.8
35
nt
8
38
10
>25
25
>25
>25
>25
0.004
0.002
0.13
nt
0.26
0.07
0.06
3.0
71
88
20
nt
0
0
0
0
0.07
3.0
89
90
0
Due to the activity compound 7 exhibited in this assay,
along with its limited cytotoxicity, it was subsequently
tested at 10 mg/kg, where it was found to cause a 46%
shutdown in tumor vasculature. This is significantly
greater activity than is found for CA4P at the same dose
(10% decrease in vascular volume at 10 mg/kg).
9
10
0.51
>25
10
aVascular shutdown is expressed as vessel density decrease in percent
of the control.
Our primary interest, however, lies in developing new
VTAs that destroy tumors indirectly, by cutting off
tumor blood flow rather than by being inherently cyto-
toxic. Accordingly, the ability of these CA4P analogues
to shut down blood flow in an endotheliaoma tumor
model was also examined.15 This in vivo assay quickly
provides a direct indication of the vascular targeting
ability of a given compound, and additional value is
In conclusion, we have synthesized nine novel ester
derivatives of the phosphate prodrug CA4P as well as a
non-hydrolyzable anionic CA4 derivative. Two of the
phosphate diesters (7 and 8) accomplish vascular shut-
down that is equivalent or greater than CA4P, while at
the same time exhibiting decreased in vitro cytotoxicity.
Additional studies exploring the vascular targeting
potential of CA4P diesters are currently underway, as
well as stability studies aimed at clarifying the source of
the desired vascular activity.
Acknowledgements
This research was supported by grants to K.E., R.R.K.,
and K.G.P. from OXiGENE, Inc. In addition, R.R.K.
and K.G.P. thank the Robert A. Welch Foundation
(grant Nos. AA-1355 and AA-1278 respectively) for
financial support of this project. K.E., R.R.K., and
K.G.P. contributed equally to this work.
References and Notes
1. Siemann, D. W. Int. J. Rad. Onc., Bio., Phys. 2002, 54,
1472.
2. Hori, K.; Saito, S.; Nihei, Y.; Suzuki, M.; Sato, Y. Jpn. J.
Cancer Res. 1990, 90, 1026. Blakey, D. C.; Ashton, S. E.;
Westwood, F. R.; Walker, M.; Ryan, A. J. Int. J. Rad. Onc.,
Bio., Phys. 2002, 54, 1497. Chaplin, D. J.; Hill, S. A. Int. J.
Rad. Onc., Bio., Phys. 2002, 54, 1491.
3. Tozer, G. M.; Prise, V. E.; Wilson, J.; Locke, R. J.; Voj-
novic, B.; Statford, M. R. L.; Dennis, M. F.; Chaplin, D. J.
Cancer Res. 1999, 59, 1626. Eikesdal, H. P.; Bjerkvig, R.;
Mella, O.; Dahl, O. Radiotherapy and Oncology 2001, 60, 147.
4. Kanthou, C.; Tozer, G. M. Blood 2002, 99, 2060. Tozer,
G. M.; Prise, V. E.; Wilson, J.; Cemazar, M.; Shan, S.;
Dewhirst, M. W.; Barber, R.; Vojnovic, B.; Chaplin, D. J.
Cancer Res. 2001, 61, 6413. Galbraith, S. M.; Chaplin, D. J.;
Lee, F.; Stratford, M. R. L.; Locke, R. J.; Vojnovic, B.; Tozer,
G. M. Anticancer Res. 2001, 21, 93.
5. Pettit, G. R.; Singh, S. B.; Hamel, E.; Lin, C. M.; Alberts,
D. S.; Garcia-Kendall, D. Experientia 1989, 45, 209.
6. Dowlatti, A.; Robertson, K.; Cooney, M.; Petros, W. P.;
Stratford, M.; Jesberger, J.; Rafie, N.; Overmoyer, B.; Mak-
Figure 1. Vascular shutdown by diester 7 at 100 mg/kg.

1508
M. B. Hadimani et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1505–1508
kar, V.; Stambler, B.; Taylor, A.; Waas, J.; Lewin, J. S.;
McCrae, K. R.; Remick, S. C. Cancer Res. 2002, 62, 3408.
7. Pettit, G. R.; Singh, S. B.; Boyd, M. R.; Hamel, E.; Pettit,
R. K.; Schmidt, J. M.; Hogan, F. J. Med. Chem. 1995, 38,
1666.
13. Relative cytoxicities were examined in several cell lines in
addition to the MHEC5-T endotheliaoma model (data not
shown).
14. Unpublished data, M. L. Trawick, Baylor University.
15. Experiments were performed in the MHEC5-T tumor
model established by sc injection of 0.5Â106 cultured
MHEC5-T cells into the right flank of SCID mice. When the
transplanted tumor reached the size of 300 mm3 (a size with-
out development of necrosis), mice received ip doses of either
10 mg/kg or 100 mg/kg of the various compounds. Twenty-
four h later 0.25 mL of diluted FluoSphere beads (1:6 in
physiological saline) was injected in the tail vein. The animals
were sacrificed after 3 min and the tumor was removed.
Cryosections (8 mm thick) were directly examined under a
fluorescent microscope. Blood vessels were indicated by blue
fluorescence from injected microbeads. For quantification,
three sections from three tumors treated in each group were
examined and in each section, more than 70% of the area was
automatically recorded with a microscopic digital camera at
Â100 magnification. The computer program Stage Pro (Media
Cybernetics, MD) was used to control the picture recording.
Image analysis was performed with Image Plus software
(Media Cybernetics, MD). The results were expressed as vessel
density decrease in percent of the control.
8. Pettit, G. R.; Rhodes, M. R. Anti-Cancer Drug Des. 1998,
13, 183.
9. Each new compound has been thoroughly characterized
(1H/13C/31P NMR, HPLC, HRMS). Supplemental material
containing a detailed experimental procedure and full char-
acterization details for each compound is available from the
authors upon request.
10. Oligonucleotides and Analogues: A Practical Approach;
Asseline, U.I., Ed.; IRL Press: Oxford, 1991.
11. Hansen, M. B.; Nielsen, S. E.; Berg, K. J. Immunol. Meth.
1989, 119, 203.
12. Mouse heart endotheliaoma cells (MHEC5-T) growing in
the exponential phase were plated at a final concentration of
1Â105 cells/mL. Cell lines were exposed to serial dilutions of
the drugs. Cell viability was determined after incubation for 1
h or 5 days at 37 ꢀC by the 3-(4,5-dimethylthiazol-2-yl)-2,5-
diplenyl-tetrazolium bromide (MTT) method. IC50 was
defined as the drug concentration that reduced cell viability
50% in comparison with control.