Synthesis and characterisation of a novel tubulin-directed DO3A-colchicine conjugate with potential theranostic features

Article abstract of DOI:10.1016/j.bmcl.2011.04.014

Colchicine is a known tubulin binding agent enabling necrosis in tumors. A novel tubulin-directed DO3A-colchicine conjugate and its Gd(III) complex were prepared from N-deacetylcolchicine, coupling alkaloid and polyaza-alicyclic functions via a peptide coupling methodology. The longitudinal proton relaxivity of the Gd(III) complex in water at 4.7 T is 2.86 mM^-1 s^-1 and a similar efficacy as colchicine towards ovarian carcinoma cells in vitro.

Full text of DOI:10.1016/j.bmcl.2011.04.014

Bioorganic & Medicinal Chemistry Letters 21 (2011) 3346–3348
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and characterisation of a novel tubulin-directed DO3A–colchicine
conjugate with potential theranostic features
a
b
b
a,
⇑
Nicholas. J. Wardle , Tammy Kalber , Jimmy D. Bell , S. W. Annie Bligh
a
Institute for Health and Research Policy, London Metropolitan University, Holloway Road, London N7 8DB, UK
Metabolic and Molecular Imaging Group, Imaging Sciences Department, MRC Clinical Sciences Centre, Imperial College London, Hammersmith Hospital, London W12 0NN, UK
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Colchicine is a known tubulin binding agent enabling necrosis in tumors. A novel tubulin-directed DO3A–
Received 25 February 2011
Revised 26 March 2011
Accepted 3 April 2011
Available online 8 April 2011
colchicine conjugate and its Gd(III) complex were prepared from N-deacetylcolchicine, coupling alkaloid
and polyaza-alicyclic functions via a peptide coupling methodology. The longitudinal proton relaxivity of
ꢀ1 ꢀ1
the Gd(III) complex in water at 4.7 T is 2.86 mM
carcinoma cells in vitro.
s
and a similar efficacy as colchicine towards ovarian
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
Magnetic resonance imaging
Contrast agent
Gadolinium
Colchicine
Tubulin
9
0
There is increasing interest in the design of theranostics, com-
pounds with both diagnostic capability and therapeutic entity,
especially in cancer areas.1 Theranostics are more commonly
used in the design of imaging agents in nuclear medicine to mon-
itor the progress of therapeutic intervention. However, the use of
theranostic strategies in drug discovery and innovative medicine
development has prompted researchers in the magnetic resonance
imaging (MRI) field to couple MR contrast agents to molecular-
targeted therapeutic drugs for clinical use. Approaches to the de-
sign of MRI theranostics have to date focused on attaching the
therapeutic to iron oxide nanoparticles,4 or biopolymers.5 In such
approaches, the drug delivery mechanism of the therapeutic is
changed since it is no longer a drug-like molecule. The ideal ther-
anostic should elicit very little change in the efficacy of the drug,
precipitate little or no immune response and have an enhanced
therapeutic response. In this study we have chosen a potent tubu-
lin binding agent, colchicine, and conjugated it to a simple Gd(III)–
DO3A (DO3A: 1,4,7,10-tetraazacyclododecane-N ,N ,N -triacetic
acid) complex to demonstrate the feasibility of producing a drug-
like MRI theranostic.
Colchicine and its analogues are of interest as putative vectors
for radiological cancer therapeutics, and conjugation to a MRI con-
trast agent offers the facility to monitor their physiological distri-
bution and effectiveness as vascular disrupting agents. Efforts
have been made to achieve such conjugate structures, exemplified
in investigations of a
Y-labelled DOTA conjugate with tri-
9
9m
methylcolchicinic acid and a
conjugate with colchicine itself in nuclear medicine.
In recent years the mechanism by which colchicine derives its
potent antimitotic properties through disruption of microtubule
assembly has been extensively studied.
tionship studies and thermodynamics studies conducted on the
requisite colchicine–tubulin binding interactions have highlighted
the negligible latitude for modification of colchicine A-11 and C-
rings, respectively, tolerated for high-affinity tubulin binding.1
This has severely limited the site and manner of colchicine-derivat-
isation available to generate an effective derivative. However,
appropriate substitution of the B-ring N-site (Fig. 1) is not neces-
Tc-labelled ethylenedicysteine
–3
6,7
8
–10
Structure–activity rela-
2,13
1
4
sarily deleterious in this respect, and has been employed to mod-
1
5
ify physicochemical and cytotoxicity profiles.
Meanwhile, a conjugate derived from reaction between p-SCN-
Bn-DOTA–Gd(III) complex and N-(2-aminoethyl)carbamoyl-
derivative of colchicine has been investigated in the context of
a
0
0
000
1
6
MRI. This latter conjugate introduces a 4-atom linker between
alkaloid and thiourea-functionalised DOTA–Gd(III) (A, Fig 2) struc-
ture with the explicit aim of avoiding steric interactions in binding
with the desired physiological target. However, it seems reason-
able that such attenuation introduces conformational mobility,
which can compromise the diagnostic response in terms of relaxiv-
1
7
ity at low magnetic fields. However complex A did not show sig-
nificant cytotoxicity on HeLa and MCF-7 cell lines despite keeping
⇑
Corresponding author.
E-mail address: a.bligh@londonmet.ac.uk (S.W.A. Bligh).
the A- and C- rings intact, and gave a longitudinal relaxivities r
1
ꢀ1 ꢀ1
3.8 mM
s
at 2.37 T.
0
960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.04.014

Nicholas. J. Wardle et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3346–3348
3347
pended the colchicine to one of the pendant arms in the cyclen
to produce a DO3A derivative. We report here the synthesis of a
neutral gadolinium(III) complex of 1-(benzo[a]heptalen-9[5H]-
one-6,7-dihydro-1,2,3,10-tetramethoxy-7-carbamoyl)methyl-4,7,
1
0-tris(carboxymethyl)-1,4,7,10-tetraazacyclododecane, (5, Fig. 3),
characterisation of ligand and complex by NMR and microanalysis,
relaxivity (r ) of the complex, and in vitro cell studies in OVCAR-3
1
cells.
Figure 1. Structure of colchicine.
At the outset, generation of the Gd(III)–DO3A–colchicine-conju-
gate (5) was envisaged via derivatisation of the B-ring amine func-
1
6
tion of N-deacetylcolchicine (1, Scheme 1), necessitated by the
minimum modification of colchicine A- and C-rings for effective
tubulin-binding. A standard peptide-coupling methodology would
be employed for this purpose, presenting the polyaza-alicyclic sys-
tem as the known protected DOTA-based prochelator; 1-carboxy-
methyl-4,7,10-tris(carbo[1,1-dimethyl]ethoxymethyl)-1,4,7,10-
1
8
tetraazacyclododecane (DOTA–tris-t-Bu ester; 2).
The method of choice to generate the fully-protected conjugate
employed a benzotriazol-1-yloxytris(dimethyl-amino)phospho-
3
nium hexafluorophosphate (BOP)-mediated coupling methodology
1
9
in dichloromethane; a method selected to allow generation of
the active ester intermediate from deprotonated carboxylate, an
important consideration in the presence of four cyclen-ring N-
Figure 2. Structure of A.
1
13
sites. H and C NMR, and electrospray ionisation mass spectrom-
etry (ESI-MS) indicated successful coupling to generate 3 (41%),
(
2 2 f
TLC; eluent—CH Cl /MeOH 7:1 v/v, R = 0.66).
Mild acidolysis with TFA then revealed the requisite DO3A–col-
chicine conjugate proligand 4 without demethylation of the C-10
tropolone methoxy function (a potential outcome in acidic envi-
2
0
ronments), as indicated by retention of the associated resonances
1
in the H NMR spectrum (d
ation of 4 was carried out at 80 °C in an aqueous suspension of
Gd (CO , the reaction progress indicated by gradual disappear-
H
= 3.51, 3.58 ppm; 2ꢁ s, 3H). Complex-
2
3 3
)
ance of the suspension. Elemental analysis of 5 was consistent with
the anticipated 1:1 Gd(III):ligand stoichiometry confirming the
retention of C-10 tropolone methoxy functionality under the con-
ditions employed.
The longitudinal proton relaxivity (r
1
) of the neutral [Gd(DO3A–
Figure 3. Structure of 5.
ꢀ
1 ꢀ1
colchicine H O)] (5) in water was 2.86 mM
2
s
, lower than that
ꢀ
of the nonconjugated anionic complex [Gd(DOTA)(H
2
O)]
ꢀ
ꢀ
1
1
ꢀ1
ꢀ1
In this work we aim to produce a molecule with both diagnostic
and therapeutic capabilities. We kept the molecular weight of the
molecule as low as possible, and produced a neutral complex for
improved uptake and reduced immuno response. Hence we ap-
3.66 mM
4.82 mM
s
s
2 2
and the neutral complex [Gd(DO3A)(H O) ]
. However comparing the r1 of 5 with a conjugated
complex such as the ‘SMART’ contrast agent [Gd(GAL-DO3A)]
ꢀ1 ꢀ1
1.52 mM
s , it is almost double the value of the latter
Scheme 1. Synthesis of a DO3A–colchicine derivative 4 and its neutral Gd(III) complex 5. Reagents and conditions: (i) 1 (deacetylcolchicine), BOP (benzotriazole-1-yl-oxy-
tris-(dimethylamino)-phosphonium hexafluorophosphate), N-methylmorpholine, dichloromethane, rt, 24 h; (ii) TFA (trifluoroacetic acid), rt, 4 h; (iii) Gd (CO , H O, 80 °C,
0 h.
2
)
3 3
2
2

3
348
Nicholas. J. Wardle et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3346–3348
Figure 4. In vitro cell studies of the ligand 4 in OVCAR-3 cells. A histology picture of OVCAR-3 cells, (A) without and (B) with 1 nM of 4 after incubation for 24 h; (C) cell
counts for both colchicine (white) and the Gd(III) complex 5 after 24 h at various concentrations (black).
complex.21 The slight reduction in r
of 5 indicates that functional-
1
References and notes
ization of one arm of the DOTA hardly reduced the number of
water molecules (q) that were in fast exchange and that the amide
carbonyl was coordinated to the Gd(III) ion. The value of q for the
complex [Tb-(Gal-DO3A)] was reported to be 0.7 by fluorescence
study, and the enzymatically cleaved terbium complex [Tb-(R-
1
2
.
.
Cai, W. B.; Niu, G.; Chen, X. Y. Curr. Pharm. Des. 2008, 14, 2943.
Guthi, J. S.; Yang, S. G.; Huang, G.; Li, S. Z.; Khemtong, C.; Kessinger, C. W.;
Peyton, M.; Minna, J. D.; Brown, K. C.; Gao, J. M. Mol. Pharm. 2010, 7, 32.
3. Jain, T. K.; Foy, S. P.; Erokwu, B.; Dimitrijevic, S.; Flask, C. A.; Labhasetwar, V.
Biomaterials 2009, 30, 6748.
4. Chen, W.; Xu, N. F.; Xu, L. G.; Wang, L. B.; Li, Z. K.; Ma, W.; Zhu, Y. Y.; Xu, C. L.;
Kotov, N. A. Macromol. Rapid Commun. 2010, 31, 228.
5. Franchini, M. C.; Baldi, G.; Bonacchi, D.; Gentili, D.; Giudetti, G.; Lascialfari, A.;
Corti, M.; Marmorato, P.; Ponti, J.; Micotti, E.; Guerrini, U.; Sironi, L.; Gelosa, P.;
Ravagli, C.; Ricci, A. Small 2010, 6, 366.
2
2
DO3A)] was 1.2. Hence it is estimated the value of q to be about
ꢀ
1
as in the [Gd(DOTA)(H
2
O)] complex.
The efficiency of the ligand 4 for binding tubulin was assessed
in an ovarian carcinoma cell line, selected for its rapidly dividing
nature and sensitivity to colchicine.2 In fact the efficiency of 4 at
causing cell shape changes and cell death when compared to col-
chicine, produced very similar results. Cell counts showed that
both the Gd(III) complex 5 and colchicine were equally effective
at (significantly) reducing cell numbers (Fig. 4). In fact, there were
enough cells only at 1 nM which could still be harvested and pel-
leted for MRI scanning to assess the uptake of 5. However, at this
6.
Satpati, D.; Korde, A.; Pandey, U.; Dhami, P.; Banerjee, S.; Venkatesh, M. J.
Labelled Compd. Radiopharm. 2006, 49, 951.
3
7. Zareneyrizi, F.; Yang, D. J.; Oh, C. S.; Ilgan, S.; Yu, D. F.; Tansey, W.; Liu, C. W.;
Kim, E. E.; Podoloff, D. A. Anticancer Drugs 1999, 10, 685.
Hastie, S. B. Pharmacol. Ther. 1991, 51, 377.
Brossi, A.; Yeh, H. J. C.; Chrzanowska, M.; Wolff, J.; Hamel, E.; Lin, C. M.; Quin, F.;
Suffness, M.; Silverton, J. Med. Res. Rev. 1988, 8, 77.
8
9
.
.
1
1
0. Ravelli, R. B. G.; Gigant, B.; Curmi, P. A.; Jourdain, I.; Lachkar, S.; Sobel, A.;
Knossow, M. Nature 2004, 428, 198.
1. Andreu, J. M.; Perez-Ramirez, B.; Gorbunoff, M. J.; Ayala, D.; Timasheff, S. N.
Biochemistry 1998, 37, 8356.
1
concentration of 5, changes in r were too small to be detected.
A novel DO3A–colchicine tubulin-directed ligand and its Gd(III)
complex were thereby prepared, the former exhibiting an effi-
12. Mareel, M. M.; Demets, M. Int Rev Cytol: Surv. Cell Biol. 1984, 90, 125.
13. Muzaffar, A.; Brossi, A.; Lin, C. M.; Hamel, E. J. Med. Chem. 1990, 33, 567.
14. Ray, K.; Bhattacharyya, B.; Biswas, B. B. J. Biol. Chem. 1981, 256, 6241.
ciency similar to that of colchicine as an anti-cancer agent. The r
1
15. Bombuwala, K.; Kinstle, T.; Popik, V.; O Uppal, S.; Olesen, J. B.; Vina, J.;
Heckman, C. A. Beilstein J. Org. Chem. 2006, 2, ARTN 13.
of its Gd(III) complex is similar to that of the nonderivatised DOTA
complex. The synthetic strategy in producing a cyclen-based ligand
retaining its complexation property, that is, four nitrogen and four
oxygen donor atoms, and a drug-like molecule was achieved.
1
6. Efthimiadou, E. K.; Katsarou, M. E.; Fardis, M.; Zikos, C.; Pitsinos, E. N.;
Kazantzis, A.; Leondiadis, L.; Sagnou, M.; Vourloumis, D. Bioorg. Med. Chem. Lett.
2008, 18, 6058.
17. Caravan, P.; Farrar, C. T.; Frullano, L.; Uppal, R. Contrast Media Mol. Imaging
009, 4, 89.
2
1
8. Heppeler, A.; Froidevaux, S.; Macke, H. R.; Jermann, E.; Behe, M.; Powell, P.;
Supplementary data
Hennig, M. Chem. Eur. J. 1999, 5, 1974.
19. Castro, B.; Dormoy, J. R.; Evin, G.; Selve, C. Tetrahedron Lett. 1975, 1219.
20. Kocienski, P. J. Protecting Groups; Stuttgart: Georg Thieme Verlag, 2005.
21. Wardle, N. J.; Herlihy, A. H.; So, P. W.; Bell, J. D.; Bligh, S. W. A. Bioorg. Med.
Chem. 2007, 15, 4714.
Supplementary data (detailed synthetic procedures and spec-
troscopic data of the products 3–5, toxicity studies and relaxation
measurements are deposited) associated with this article can be
found, in the online version, at doi:10.1016/j.bmcl.2011.04.014.
2
2
2. Moats, R. A.; Fraser, S. E.; Meade, T. J. Angew. Chem. 1997, 36, 726.
3. Tozer, G. M.; Kanthou, C.; Baguley, B. C. Nat. Rev. Cancer 2005, 5, 423.