Synthesis and biological evaluation of B-ring modified colchicine and isocolchicine analogs

Article abstract of DOI:10.1016/j.bmcl.2006.02.010

A series of modified colchicine and isocolchicine analogs (C-7 substituent) were synthesized and evaluated in vitro against a PC3 cancer cell line and for inhibition of microtubule polymerization. The colchicine analogs all displayed strong inhibition of

Full text of DOI:10.1016/j.bmcl.2006.02.010

Bioorganic & Medicinal Chemistry Letters 16 (2006) 2761–2764
Synthesis and biological evaluation of B-ring modified
colchicine and isocolchicine analogs
Michael Cifuentes,a Brett Schilling,b Rudravajhala Ravindra,b
Jacquelyn Wintera and Mark E. Janika,*
aDepartment of Chemistry, SUNY Fredonia, Fredonia, NY 14063, USA
bDepartment of Chemistry, Binghamton University, Binghamton, NY 13902-6016, USA
Received 8 December 2005; revised 2 February 2006; accepted 2 February 2006
Available online 28 February 2006
Abstract—A series of modified colchicine and isocolchicine analogs (C-7 substituent) were synthesized and evaluated in vitro against
a PC3 cancer cell line and for inhibition of microtubule polymerization. The colchicine analogs all displayed strong inhibition of
tubulin polymerization, while compounds 6 and 20 also possessed an increased cytotoxic activity as compared to colchicine. More
importantly, isocolchicine analogs 7, 15, and 17 showed inhibition of microtubule polymerization with IC50 values ranging from 58
to 68 lM. In addition, 7 displayed strong cytotoxic activity with an IC50 = 93 nM which was more potent than colchicine analog 12.
Ó 2006 Elsevier Ltd. All rights reserved.
The compound colchicine (1) (Fig. 1) is a highly potent
antimitotic agent that derives its therapeutic benefit by
binding to the protein tubulin.1,2 The mechanistic under-
standing of colchicine–tubulin binding has been highly
investigated using both structure–activity relationship
(SAR) studies and thermodynamic analyses.3–5 These
studies suggest that the A- and C-rings of the parent
molecule comprise the minimum structural feature
necessary for high affinity drug–tubulin binding. An
example illustrating the importance of the C-ring is re-
vealed by isocolchicine (2), which is virtually inactive
in binding to tubulin.6,7 Recently, however, the discov-
ery of the first active isocolchicine analog (3) (B-ring
substituent modified) has prompted renewed interest in
the role of the B-ring in colchicine–tubulin binding.8
Therefore, in a continuing effort to identify potent tubu-
lin binding isocolchicine analogs and to investigate the
SAR of the B-ring substituent we decided to synthesize
and biologically evaluate a series of colchicine and iso-
colchicine analogs modified at the C-7 position of the
B-ring.
Deacetylcolchicine (4) and deacetylisocolchicine (5) ob-
tained from colchicine (Scheme 1) 9,10 were reacted with
lithium perchlorate and propylene oxide in acetonitrile
to afford their respective b-amino alcohols in 95%
O
MeO
MeO
O
MeO
MeO
O
S
MeO
MeO
A
B
NHCCH3
NH
O
NHCCH3
OMe
C
OMe
OMe
O
OMe
OMe
2
1
OMe
3
N(CH3)2
O
O
Figure 1.
Keywords: Colchicine; SAR; Tubulin; Isocolchicine; Antimitotic.
*
Corresponding author. Tel.: +1 716 673 3508; fax: +1 716 673 3347; e-mail: janikm@fredonia.edu
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.02.010

2762
M. Cifuentes et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2761–2764
O
MeO
MeO
MeO
MeO
NHCCH3
NH2
NH2
OM
b,c
+
a
MeO
MeO
OMe
OMe
OMe
O
5
4
1
O
O
e
OMe
OMe
O
O
MeO
MeO
MeO
+
NHCH2CC H3
NHCH2CCH3
MeO
OMe
OMe
O
7
OMe
6
OMe
O
Scheme 1. Reagents and conditions: (a) 1—2 N HCl/CH3OH, reflux; 2—CH2N2/CH2Cl2; (b) propylene oxide, LiClO4, CH3CN; (c) DMSO, (COCl)2,
CH2Cl2, TEA, À60 °C.
NH2
NH2
NH
N
NH2
NH2
a
b
c
O
S
F
F
NHCCH3
O
F
NHCCH3
S
F
C
10
11
8
9
H3C
MeO
S
MeO
S
NHCCH3
+
NHCCH3
OMe
d
MeO
MeO
13
OMe
OMe
O
12
OMe
O
Scheme 2. Reagents and conditions: (a) (CH3CO)2O, CH2Cl2, 0 °C; (b) P4S10, Na2CO3, THF, 0 °C; (c) COCl2, THF; (d) 4 and 5, TEA, CH3CN.
yield.11 It should be noted that various metal salts
(LiOTf, ZnCl2, and CoCl2) were initially used prior to
lithium perchlorate; however, they all resulted in non-
regioselective attack of the amine on the epoxide
ring.12,13 Initial attempts at alcohol oxidation utilized
both pyridinium chlorochromate and pyridinium
dichromate, however, no product formation was ob-
served.14 Thus, we alternatively employed Swern oxida-
tion whereby the b-amino alcohols were added to a
solution of DMSO/oxalyl chloride at À60 °C to afford
compounds 6 and 7 in 88% yield.15 The two isomers
were separated using radial chromatography.
gene for 4 h at room temperature to yield thiobenzimi-
dazolone 11 (71%). Thiobenzimidazolone 11 was then
reacted with a mixture containing isomers 4 and 5 in
acetonitrile. The reaction was monitored by TLC and
after stirring for 48 h at room temperature compounds
12 and 13 were afforded in a combined 76% yield. The
two isomeric compounds were separated using radial
chromatography.
Additional investigation of the amide carbonyl was
sought through conversion of the carbonyl moiety into
a methylene (CH2) group. Initial attempts at the syn-
thesis of compounds 18 and 19 (Scheme 3) focused
on the use of reducing agents such as LiAlH4 and bor-
ane-dimethyl sulfide; however, no formation of the de-
sired secondary amine occurred.18 Thus, we instead
attempted to synthesize the target amines utilizing a
Mitsunobu alkylation of a 2-nitrobenzenesulfonamide
intermediate (14 and 15).19 The alkylation reactions
proceeded in moderate yield but reaction times tended
to be of the order of 2–3 days. We therefore alterna-
tively employed the use of compounds 16 and 17
(2,4-dinitrobenzenesulfonamide), as shown in (Scheme
3), which were more efficiently alkylated using Mitsun-
obu conditions in 8 h.20 The dinitrobenzenesulfona-
mide-protecting group was removed under mild
reaction conditions to yield compounds 18 and 19
(82%) which were separated using radial chromatogra-
phy. It is of interest to note that an attempt to form
To investigate the role of the amide carbonyl group, we
sought to synthesize analogs of 1 and 2 whereby the
amide was replaced with a thioamide functionality.
Our initial attempts at thioamide formation involved the
use of Lawesson’s reagent;16 however, this resulted in
conversion of the C-ring carbonyl into a thiocarbonyl.
To synthesize our thioamide analogs (12 and 13), we in-
stead utilized a procedure that relied on the use of thio-
acylating reagent 11, which was synthesized as shown in
(Scheme 2).17 Compound 8 was chosen as our benzimi-
dazolone source and it was reacted with acetic anhy-
dride to afford compound 9 in 89% yield. Conversion
of the amide in 9 to a thioamide was accomplished using
phosphorus pentasulfide to give 10 in 72% yield.
Compound 10 was reacted with two equivalents of phos-

M. Cifuentes et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2761–2764
2763
MeO
MeO
O
S
MeO
MeO
O
S
a or b
NH
R
4
+
5
NH
R
O
NO2
+
O
NO2
OMe
OMe
O
R= H, 14
OMe a)
b)
a)
b)
R= H, 15
R= NO , 17
OMe
R= NO , 16
2
O
2
MeO
MeO
MeO
NHCH2CH3
NHCH2CH3
OMe
+
c,d
MeO
16 & 17
OMe
OMe
O
19
18
OMe
O
MeO
MeO
NHR
O
NHR
OMe
e
+
R =
NO2
4
+
5
MeO
MeO
OMe
OMe
NO2
20
21
OMe
O
Scheme 3. Reagents and conditions: (a) 2-nitrobenzenesulfonyl chloride, TEA, CH2Cl2; (b) 2,4-dinitrobenzenesulfonyl chloride, 2,6-lutidine,
CH2Cl2; (c) CH3CH2I, K2CO3, CH3CN; (d) HSCH2CO2H, TEA, CH2Cl2; (e) 2,4-dinitrobenzenesulfonyl chloride, TEA, CH2Cl2.
Table 1. Inhibition of microtubule polymerization and cytotoxic activity against PC3 cell lines for compounds 6, 7, and 12–21
Colchicine isomer IC50, MTP assemblya (lM) IC50, PC3b (nM) Isocolchicine isomer IC50, MTP assembly (lM) IC50, PC3b (nM)
1
6
1.7 0.04c
8.3 0.37
2.2 0.10
5.6 0.14
9.7 0.27
5.1 0.16
4.6 0.17
11 0.65c
7.6 2.1
2
7
Inactived
68 2.4
4,600 0.19
93 10.3
12
14
16
18
20
97.5 28.7
19.8 11.8
50 16.6
13
15
17
19
21
Inactivee
58 2.4
1,752 435
905 117
59 4.8
Inactivee
Not soluble
10,250 1,639
1,722 819
5,700 754
17.6 13.2
9.0
2
a Concentration of ligand required to effect a 50% reduction in microtubule polymerization.
b Human prostate cancer cells were used for cytotoxicity measurements.
c Values represent means SD of at least three determinations (assay procedure described in supplementary materials section).
d Inhibition of polymerization measured up to a concentration of 300 lM drug.
e Inhibition of polymerization measured up to a concentration of 150 lM drug.
16 and 17 using triethylamine (2 equiv.) rather than
lutidine as the base resulted in the formation of
compounds 20 and 21, respectively, as the major reac-
tion products.
colchicine into a methylene group resulted in the more
active compound 18 (IC50 = 5.1 lM) as compared
to 6.
Of significant importance was the antimicrotubule
activity displayed by select isocolchicine analogs.
The most active isocolchicine analog was compound
15 with an IC50 value of 58 lM. This analog con-
tained an aromatic nitrobenzenesulfonamide substitu-
ent at the C-7 position. Addition of a second nitro
group on the C-7 aromatic ring resulted in com-
pound 17, which displayed similar activity
(IC50 = 59 lM) as compared to 15. Removal of the
sulfonyl group, however, resulted in poor compound
solubility beyond 30 lM as seen in 21. Insertion of
a methylene unit between the amide nitrogen and
carbonyl in inactive isocolchicine 2 (Table 1) resulted
in the active compound 7 (IC50 = 68 lM). Interesting-
ly, the thioamide isocolchicine analog 13 displayed no
activity, which is in contrast to 12 the most active
analog in the colchicine series. Compound 19, conver-
sion of the carbonyl into a methylene group, was
inactive as well.
The antimicrotubule activities of compounds 6, 7, and
12–21 were assayed by their ability to inhibit in vitro
assembly of microtubule protein (Table 1) using col-
chicine and isocolchicine as reference compounds.21
As expected, the nature of the C-7 substituent did
not greatly affect the in vitro polymerization activity
of colchicine analogs (6, 12, 14, 16, 18, and 20) which
all displayed strong microtubule inhibition. The most
active compound was the thioamide analog 12
(IC50 = 2.2 lM) possessing an activity nearly identical
to that of colchicine (IC50 = 1.7). Concerning the
aromatic analogs (14, 16, and 20), compound 20 pos-
sessed the strongest inhibition with an IC50 = 4.6 lM.
Introduction of a methylene group between the amide
nitrogen and carbonyl of colchicine resulted in com-
pound 6 which displayed nearly a 5-fold reduction
in activity (IC50 = 8.3 lM) as compared to colchicine.
Interestingly though, conversion of the carbonyl in

2764
M. Cifuentes et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2761–2764
In Table 1, we have also reported the cytotoxic activi-
ties of the analogs against a human prostate cancer cell
line (PC3).22 The most active colchicine analogs were
compounds 6 (IC50 = 7.6 nM) and 20 (IC50 = 9.0 nM)
which were both more potent than the parent com-
pound colchicine (IC50 = 11 nM). Additional colchicine
analogs (12, 14, 16, and 18) all displayed strong cyto-
toxic activities as well. Concerning the isocolchicine
analogs, compounds 7, 13, 15, and 19 were significantly
more active than the parent isocolchicine. Most note-
worthy was compound 7 (IC50 = 93 nM) which dis-
played strong antiproliferative activity. Remarkably,
compound 7 possessed stronger activity than the colchi-
cine analog 12. The remaining isocolchicines (17 and
21) possessed cytotoxicities of 10,250 and 5700 nM,
respectively.
Acknowledgment
We thank Dr. Susan Bane for her helpful insight and
guidance concerning this research and in the preparation
of the manuscript.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmcl.2006.02.010.
References and notes
1. Capraro, H. G.; Brossi, A.. In The Alkaloids; Brossi, A.,
Ed.; Academic: New York, 1984; Vol. 23, p 1.
2. Ravelli, R. B. G.; Gigant, B.; Curmi, P. A.; Jourdain, I.;
Iachkar, S.; Sobel, A.; Knossow, M. Nature 2004, 428, 198.
3. Lee, K. H. Med. Res. Rev. 1999, 19, 569.
4. Quinn, F. R.; Neiman, Z.; Beisler, J. A. J. Med. Chem.
1981, 24, 636.
5. Timasheff, S. N.; Andreu, J. M.; Na, G. C. Pharmacol.
Ther. 1991, 52, 191.
6. Hastie, S. B.; Williams, R. C., Jr.; Puett, D.; McDonald,
T. L. J. Biol. Chem. 1989, 264, 6682.
7. Dumortier, C.; Yan, Q.; Bane, S.; Engelborghs, Y.
Biochem. J. 1997, 327, 685.
8. Das, L.; Datta, A. B.; Gupta, S.; Poddar, A.; Sengupta, S.;
Janik, M. E.; Bhattacharyya, B. Biochemistry 2005, 44,
3249.
9. Wilson, L.; Friedkin, M. Biochemistry 1966, 5, 2463.
10. Capraro, H. G.; Brossi, A. Helv. Chim. Acta 1979, 62, 965.
11. Chini, M.; Crotti, P.; Macchia, F. Tetrahedron Lett. 1990,
31, 4661.
12. Auge, J.; Leroy, F. Tetrahedron Lett. 1996, 37, 7715.
13. Iqbal, J.; Pandey, A. Tetrahedron Lett. 1990, 31, 575.
14. Corey, E. J.; Suggs, J. W. Tetrahedron Lett. 1975, 16, 2647.
15. Mancuso, A.; Huang, S. L.; Swern, D. J. Org. Chem. 1978,
43, 2480.
16. Cava, M. P.; Levinson, M. I. Tetrahedron 1985, 41, 5061.
17. Zacharie, B.; Lagraoui, M. D.; Penney, C. L.; Gagnon, L.
J. Med. Chem. 1999, 42, 2046.
18. Brown, H. C.; Weissman, P. M.; Yoon, N. M. J. Am.
Chem. Soc. 1966, 88, 1458.
19. Kurosawa, W.; Kan, T.; Fukuyama, T. Org. Synth. 2002,
79, 186.
20. Fukuyama, T.; Cheung, M.; Jow, C. K.; Hidai, Y.; Kan,
T. Tetrahedron Lett. 1997, 38, 5831.
21. Williams, R. C., Jr.; Lee, J. C. Methods Enzymol. 1982, 85,
376.
22. Chang, M. C.; Uang, B. J.; Wu, H. L.; Lee, J. J.; Hahn, L.
J.; Jeng, J. H. Br. J. Pharmacol. 2002, 135, 619.
23. Pyles, E. A.; Hastie, S. B. Biochemistry 1993, 32, 2329.
In conclusion, a series of B-ring C-7 substituent mod-
ified colchicine and isocolchicine analogs were synthe-
sized and evaluated for antimicrotubule and cytotoxic
activity. The colchicine derivatives all displayed strong
potencies with 12 possessing the strongest antimicrotu-
bule inhibition and 6 and 20 displaying a more potent
cytotoxicity against PC3 than colchicine. Interestingly,
compound 6 was 5-fold less active in the MTP assay as
compared to colchicine. Previous studies have indicat-
ed that the association rate of colchicine analogs bind-
ing to tubulin is affected by the C-7 substituent.23
Thus, 6 may bind slower to tubulin than colchicine
which could explain its lower potency in the MTP as-
say (fixed/short incubation time) as compared to the
cytotoxicity experiments (3 days incubation time)
where kinetics should not be a factor. A most notewor-
thy finding was the potency displayed by isocolchicine
compounds 15 and 17. Previous studies have indicated
that select isocolchicine analogs (aromatic C-7 side
chains) may be able to interact with a-tubulin through
their B-ring substituent thereby increasing their overall
binding affinity for the protein.8 Compounds 15 and 17
could likely interact with tubulin in a similar manner
which would explain their improved activity as com-
pared to inactive isocolchicine. A most unexpected re-
sult in the iso series was the relatively high activity of 7
in both the MTP and cytotoxic assays. In fact, com-
pound 7 surprisingly displayed a stronger inhibition
against PC3 cell lines as compared to colchicine deriv-
ative 12. In further studies, we will focus on determin-
ing the association rate of 6 with tubulin as compared
to colchicine. In addition, we will synthesize analogs of
7 (–NHCH2CH2COCH3, etc.) to further explore the
SAR of isocolchicinoids of this type.