Synthesis and SAR requirements of adamantane-colchicine conjugates with both microtubule depolymerizing and tubulin clustering activities

Article abstract of DOI:10.1016/j.bmc.2011.07.040

A series of analogues of conjugate 1, combining an adamantane-based paclitaxel (taxol) mimetic with colchicine was synthesized and tested for cytotoxicity in a cell-based assay with the human lung carcinoma cell line A549. The most active compounds (10 EC₅₀ 2 ± 1.0 nM, 23 EC ₅₀ 6 ± 1.4 nM, 26 EC₅₀ 5 ± 1.8 nM, 28 EC₅₀ 11 ± 1.7 nM, 30 EC₅₀ 4.8 ± 0.5 nM) were found to interfere with the microtubule dynamics in an interesting manner. Treatment of the cells with these compounds promoted disassembly of microtubules followed by the formation of stable tubulin clusters. Structure-activity relationships for the analogues of 23 revealed the sensitivity of both cytotoxicity and tubulin clustering ability to the linker length. The presence of adamantane (or another bulky hydrophobic and non-aromatic moiety) in 23 was found to play an important role in the formation of tubulin clusters. Structural requirements for optimal activity have been partially explained by molecular modeling.

Full text of DOI:10.1016/j.bmc.2011.07.040

Bioorganic & Medicinal Chemistry 19 (2011) 5529–5538
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Synthesis and SAR requirements of adamantane–colchicine conjugates
with both microtubule depolymerizing and tubulin clustering activities
Olga N. Zefirova a,b, , Evgeniya V. Nurieva a, Dmitrii V. Shishov a, Igor I. Baskin a, Fabian Fuchs c,
⇑
Heiko Lemcke c, Fabian Schröder c, Dieter G. Weiss c, Nikolay S. Zefirov a,b, Sergei A. Kuznetsov c
a Department of Chemistry, M. V. Lomonosov Moscow State University, 119992 Moscow, Russian Federation
b Institute of Physiologically Active Compounds, Noginski District, Moscow Area, 142432 Chernogolovka, Russian Federation
c Institute of Biological Sciences, Cell Biology and Biosystems Technology, University of Rostock, D-18059 Rostock, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of analogues of conjugate 1, combining an adamantane-based paclitaxel (taxol) mimetic with
Received 24 June 2011
Revised 21 July 2011
Accepted 21 July 2011
Available online 28 July 2011
colchicine was synthesized and tested for cytotoxicity in a cell-based assay with the human lung carci-
noma cell line A549. The most active compounds (10 EC50
2
1.0 nM, 23 EC50
6
1.4 nM, 26 EC50
5
1.8 nM, 28 EC50 11 1.7 nM, 30 EC50 4.8 0.5 nM) were found to interfere with the microtubule
dynamics in an interesting manner. Treatment of the cells with these compounds promoted disassembly
of microtubules followed by the formation of stable tubulin clusters. Structure–activity relationships for
the analogues of 23 revealed the sensitivity of both cytotoxicity and tubulin clustering ability to the lin-
ker length. The presence of adamantane (or another bulky hydrophobic and non-aromatic moiety) in 23
was found to play an important role in the formation of tubulin clusters. Structural requirements for opti-
mal activity have been partially explained by molecular modeling.
Keywords:
Adamantane derivatives
Colchicine derivatives
Hybrid molecules
Microtubule depolymerization
Tubulin clustering
Cytotoxicity
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction
Fig. 1),4 which possessed very high cytotoxicity against A549
human lung carcinoma cells. Moreover the preliminary test indi-
Natural compounds such as paclitaxel (taxol), colchicine and
vinca-alkaloids (vincristine and vinblastine) possess high antitu-
mour activity due to their interaction with the intracellular protein
tubulin. Although these compounds have the same target protein,
their binding sites and their biological activities differ consider-
ably. Thus, paclitaxel causes spontaneous tubulin polymerization
into stable microtubules and stabilize preformed microtubules
whereas colchicine and vinca-alkaloids inhibit tubulin polymeriza-
tion and microtubule formation.1,2
Among the different approaches to the design of new anti-
tubulin agents, studies of various hybrid ligands recently attracted
much attention.3 The rationale behind the combination of two or
more tubulin ligands in a single molecule is that the hybrid ligand
may have higher activity and/or better pharmacological profile
than the equivalent combination of individual compounds. Besides,
this approach leads sometimes to a discovery of molecules with
unusual types of activity.
cated that hybrid 1 had an unusual cytotoxicity profile, namely
both microtubule destabilizing activity and ability to promote
the formation of stable tubulin structures resembling the action
of paclitaxel with microtubule bundling or vinblastine-like activity
with tubulin paracrystal formation.
In the present work we carried out additional biological tests to
verify the effect of 1 on tubulin and studied the structure–activity
relationships for a number of analogues of the lead compound.
Several aspects were suggested for the investigation: the linker
length, the possibility of amino acid side chain removal, the posi-
tion of the linker attachment and the role of adamantane moiety.
Molecular modeling was performed for a better understanding
SAR results.
2. Results and discussion
2.1. Biological assays for the lead compound
Earlier we synthesized a hybrid ligand combining an adaman-
tane-based paclitaxel mimetic with colchicine (1, trans/cis 2:1,
The hybrid 1 was first tested using in vitro tubulin polymeriza-
tion assay.5 Purified bovine brain tubulin was polymerized in the
presence of 10% DMSO and supplemented with 10
10 M colchicine, as positive control. After 1 h of incubation the
resulting preparations were analyzed by Allen Video-Enhanced
lM of 1 or
⇑
l
Corresponding author. Tel.: +7 495 939 48 78; fax: +7 495 939 02 90.
E-mail address: olgaz@org.chem.msu.ru (O.N. Zefirova).
0968-0896/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2011.07.040

5530
O. N. Zefirova et al. / Bioorg. Med. Chem. 19 (2011) 5529–5538
O
Ph
NH
O
7
4
1
O
O
OCH3
O
BocNH
O
OH
OCH3
CH3O
CH3O
Figure 1. Adamantane-based paclitaxel mimetic—colchicine hybrid.
Table 1
Cytotoxicity and effect of the hybrid compounds on A549 cellular microtubules
N
Compound
Ph
Cytotoxicitya EC50 SD (nM)
Effect on cellular microtubulesb at 1
lM
OCH3
OCH3
O
(CH2)5
NH
O
O
BocNH
BocNH
BocNH
BocNH
O
O
O
1
73
2.9
Depolymerization and tubulin clustering ++
Depolymerization and tubulin clustering +
Depolymerization and tubulin clustering +++
ND
OCH3
OH
CH3O
OCH3
Ph
O
(CH2)5
NH
OCH3
OCH3
O
O
4
41 9.8
OH
O
O
CH3O
NH
OCH3
Ph
(CH2)6
O
O
OCH3
OCH3
O
O
O
O
O
10
11
2
1.0
OH
O
CH3O
OCH3
OCH3
Ph
O
(CH2)3
NH
O
220 12
OCH3
OH
O
CH3O
Colchicine
Paclitaxel (Taxol)
27 1.5
4.6 0.7
Depolymerization only
Microtubule bundling
a
The average of three to six experiments.
The number of ‘+’ symbolizes the relative strength of the effect.
b
Figure 2. Effect of conjugate 1 on the microtubule network in A549 lung carcinoma cells analyzed by indirect immunofluorescence microscopy. Cells were treated: A—with
0.6% DMSO (negative control); B—with 1 M of colchicine (positive control); C and D—with 1 M and 5 M of compound 1 respectively. Destabilizing effect of 1 on the
microtubule network and tubulin clustering effect are clearly seen in C and D. Bar 20 m.
l
l
l
l

O. N. Zefirova et al. / Bioorg. Med. Chem. 19 (2011) 5529–5538
5531
Figure 3. Typical patterns of tubulin distribution in treated cells used for evaluation of the effect of tested compounds on microtubules as indicated in Tables 1 and 2.
A—‘depolymerization only’; B—tubulin clustering ‘+’; C—tubulin clustering ‘++’; D—tubulin clustering ‘+++’. Bar 20 m.
l
Differential Interference Contrast light microscopy.6 According to
the images obtained hybrid 1 completely inhibited the microtu-
bule assembly in vitro. This effect was more pronounced than that
of colchicine, in the presence of which few small aggregates and
short microtubules have been detected (see Supplementary data).
Compound 1 was previously found to inhibit strongly the cell
growth of human lung carcinoma cell line A549 in a cell prolifera-
tion test.4 In this study cytotoxic properties of 1 were further
evaluated by A549 cell-based viability assay with the standard
MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide]
assay protocol.7 The EC50 value was found to be close to that of
of 5 lM no intact microtubules were observed and the clustering
effect of 1 was more prominent (Fig. 2D).
These experiments proved the ability of hybrid 1 to cause tubu-
lin clustering in addition to the colchicine-like activity. We named
the former ability ‘tubulin clustering’ because the morphology of
these clusters is different from paclitaxel-induced microtubule
bundles and vinblastine-induced tubulin paracrystals (see also
Fig. 3). Interestingly, the ‘clustering effect’ cannot be caused by
the simultaneous cell treatment with 5 lM paclitaxel and 5 lM
colchicine for eight hours, thus being a specific property of the hy-
brid molecule.
colchicine (see Table 1)
.
The effect of hybrid 1 on microtubule dynamics in A549 cells
2.2. Chemistry
was also investigated and compared to that of colchicine. Cells
were incubated with 1
formaldehyde and stained with primary mouse monoclonal anti-
l
M and 5
l
M of 1 for 8 h, fixed in 4% para-
To determine the structural requirements for the tubulin clus-
tering effect and cytotoxicity a structure–activity investigation
was performed for analogues of 1. The derivative of hybrid 1
without adamantane moiety (structure 4) was synthesized from
12-hydroxydodecanoic acid and N-deacetylcolchicine (prepared in
three steps from colchicine9) as shown in Scheme 1. Esterification
of the obtained amide 2 by an oxazolidine-type protected N-tert-
butoxycarbonylphenylisoserine10 led to the conjugate 3 and depro-
tection of the latter afforded the purpose compound 4.
The synthesis of two hybrids 10, 11 with different linker lengths
was analogous to that of 1 (Scheme 2). The esterification of 1-hy-
droxy-adamantantane-4-one by protected amino acid and subse-
quent reduction of the resulting keto-ester gave a mixture of
isomeric alcohols 5.4 Further esterification of 5 by either suberic
acid polyanhydride11 or glutaric acid anhydride led to isomeric
esters 6 and 7 respectively. Finally, N-deacetylcolchicine was
attached to the carboxylic group of compounds 6 and 7. The
following deprotection of amino acid in the products 8 and 9 affor-
ded hybrids 10 (trans/cis 7:3) and 11 (trans/cis 1:1).
bodies against
a-tubulin followed by incubation with ALEXA-
Fluor488 labelled goat anti-mouse IgG as the secondary antibody.8
Images of the samples were analyzed by fluorescence microscopy
and are shown on Figure 2 A–D.
At a concentration of 1 lM hybrid 1 had a clearly visible desta-
bilizing effect on the microtubule network: the samples displayed
only very few small microtubule fragments. Although the effect of
the tested compound was slightly weaker than that of colchicine,
contrary to the latter, hybrid 1 in addition caused a formation of
tubulin clusters which were detected among short microtubules
and diffusely labeled tubulin (Fig. 2C). Notably, at a concentration
It should be mentioned that this value was higher than the IC50 value for
compound 1 determined earlier by a different method (i.e. the proliferation assay,
where the cells were counted directly with hemocytometer cell count calculator).4
Actually it is not clear, why the two methods give so different absolute values, while
the relative values obtained for the different compounds are in a good agreement.
However in the present investigation MTT experiments were carried out under
identical conditions for all the tested compounds. The differences in the effects were
also confirmed by immunofluorescence microscopy.
All derivatives of 1 without amino acid side chain (23–29) and
their analogues with hydrophobic substituents other than

5532
O. N. Zefirova et al. / Bioorg. Med. Chem. 19 (2011) 5529–5538
OCH3
(CH2)11
NH
HO
(CH2)11
OH
OCH3
HO
a
O
CH3O
O
2
O
Ph
O
CH3O
Ph
O
OCH3
OCH3
OH
BocN
O
(CH2)11
NH
p-MeO-Ph
b
O
BocN
O
O
CH3O
p-MeO-Ph
3
O
CH3O
CH3O
Ph
O
OCH3
OCH3
(CH2)11
NH
O
c
BocNH
O
O
OH
4
CH3O
Scheme 1. Reagents and conditions: (a) EEDQ, CH2Cl2, rt, 60 h; (b) DCC, DMAP, CH2Cl2, rt, 12 h; (c) pTSA, MeOH, rt, 2.5 h.
Ph
(CH2)n
Ph
OH
OH
O
O
O
a
O
O
O
O
BocN
BocN
O
O
6 n=6
7 n=3
5
p-MeO-Ph
p-MeO-Ph
(CH2)n
Ph
HN
O
O
b
O
O
OCH3
O
BocN
O
O
OCH3
8 n=6
CH3O
p-MeO-Ph
9 n=3 CH3O
O
Ph
(CH2)n
O
NH
c
O
O
OCH3
O
BocNH
O
OH
OCH3
10 n=6
11 n=3
CH3O
CH3O
Scheme 2. Reagents and conditions: (a) (OOC–(CH2)n–CO)m, DMAP, CH2Cl2, rt, 24-48 h; (b) N-deacetylcolchicine, EEDQ, CH2Cl2, rt, 24 h; (c) pTSA, MeOH, rt, 2.5 h.
adamantane (30–33) were prepared from the corresponding alco-
hols as shown in Scheme 3.
cytotoxicity than the parent compound and a strongly less pro-
nounced clustering effect. Thus the adamantane fragment was
maintained in the next two series of compounds.
2.3. Biotests and structure–activity relationships
The data for 1, 10 and 11 clearly indicate that the elongation of
the spacer length improves cytotoxity, the hybrid 10 with a six-
methylene linker being one order of magnitude more potent than
colchicine and as potent as paclitaxel (see Table 1).
To check the role of amino acid substituent in the parent struc-
ture we synthesized compound 23, which is an analogue of 1 with-
out N-tert-butoxycarbonyl-(2R,3S)-phenylisoseryl moiety. This
amino acid represents a C13 side chain of paclitaxel derivative tax-
oter and is known to provide the most important contribution to
the paclitaxel-binding site interaction.1,2 As seen in Table 2 conju-
gate 23 exhibits very high cytotoxicity and its clustering activity is
The inhibitory effect of the synthesized compounds on prolifer-
ation of human lung carcinoma A549 cells was evaluated using
MTT test.7 Effect on microtubule dynamics was investigated as de-
scribed in section 2.1 (the typical patterns of tubulin clusters in
cells treated with different compounds are shown in Fig. 3). The re-
sults of both tests are presented in Tables 1 and 2.
As evidenced from Table 1 an analogue of hybrid 1 without ada-
mantane moiety and with the similar distance between colchicine
and the amino acid (hybrid 4) exhibited only slightly higher

O. N. Zefirova et al. / Bioorg. Med. Chem. 19 (2011) 5529–5538
(CH2)n
5533
HN
RO
O
O
OCH3
(CH2)n
OH
O
b
a
R
R
OH
O
O
O
OCH3
CH3O
CH3O
12 R=2-Ad, n=5
13 R=2-Ad, n=4
14 R=2-Ad, n=3
15 R=2-Ad, n=6
16 R=2-Ad, n=7
17 R=1-Ad, n=5
23 R=2-Ad, n=5
24 R=2-Ad, n=4
25 R=2-Ad, n=3
26 R=2-Ad, n=6
27 R=2-Ad, n=7
28 R=1-Ad, n=5
18 R=1-Ad, n=7
29 R=1-Ad, n=7
19 R= cyclohexyl, n=5
20 R=Me, n=5
30 R= cyclohexyl, n=5
31 R=Me, n=5
21 R=i-Pr, n=5
32 R=i-Pr, n=5
22 R=Ph, n=5
33 R=Ph, n=5
Scheme 3. Reagents and conditions: (a) (OOC–(CH2)n–CO)m, DMAP, CH2Cl2, rt, 24–36 h; (b) N-deacetylcolchicine, EEDQ, CH2Cl2, rt, 12–24 h, (c) TFA, CH2Cl2, rt, 24 h.
Table 2
Cytotoxicity and effect of the colchicine conjugates on A549 cellular microtubules
OCH3
OCH3
(CH2)n
RO
NH
O
O
O
OCH3
CH3O
Compound
R
Cytotoxicitya EC50 SD (nM)
Effect on cellular microtubulesb at 1
lM
N
n
23
5
6
1.4
Depolymerization and tubulin clustering +++
Depolymerization and tubulin clustering ++
depolymerization only
24
25
26
27
28
29
4
3
6
7
5
7
830 140
200 62
5
1.8
Depolymerization and tubulin clustering +++
Depolymerization and tubulin clustering +
Depolymerization and tubulin clustering +++
ND
3800 630
11 1.7
29 1.2
30
31
5
5
4.8 0.5
32 2.1
Depolymerization and tubulin clustering +++
Depolymerization only
H3C
H3C
32
33
5
5
30 2.7
Depolymerization and tubulin clustering +
H3C
5700 820
Weak depolymerization only
Colchicine
Paclitaxel (Taxol)
27 1.5
4.6 0.7
Depolymerization only
Microtubule bundling
a
The average of three to six experiments.
The number of ‘+’ symbolizes the relative strength of the effect.
b
improved in comparison with 1. This result is remarkable and sig-
nifies that the clustering effect does not depend upon the presence
of taxoter amino acid side chain and, consequently, is hardly con-
nected to an interaction with the paclitaxel binding site of tubulin.
Thus, these data are in accordance with the results of our study on
the morphology of the clusters (see Section 2.1).
As though the structure of conjugate 23 is much simpler than
that of 1, we have chosen 23 as a new lead compound for subse-
quent structure–activity investigations.
A SAR study in a series of analogues of 23 with different linker
length 23–27 (Table 2) shows a maximum of cytotoxicity for
ligands with five- and six-methylene linkers (23, 26). Further
elongation of the spacer (27) or its shortening (24, 25) leads to a
considerable decrease of activity. Interestingly, all compounds in
the series are in fact rather simple colchicine derivatives with a
modified side chain at C7 (it is generally accepted that substitution
at this position is highly tolerated (see e.g.12)). Nevertheless a
strong dependence of both cytotoxicity and clustering effect on

5534
O. N. Zefirova et al. / Bioorg. Med. Chem. 19 (2011) 5529–5538
Figure 4. Location of compound 23 at the interface of
a/b tubulin subunits (b-subunit is colored in red, a-subunit in blue, hydrogen atoms are not shown for clarity,
molecular surface mesh is indicated).
the length of this substituent was observed. Thus, three-methylene
ligand 25 maintains only the microtubule depolymerizing activity
while its four-methylene analogue 24 possesses also a noticeable
clustering effect.
The structural modification of the lead 23 next studied was a
change of the position of linker attachment to adamantane. Conju-
gate 28 with the linker attached to a bridgehead atom possessed
high cytotoxicity and clustering effect close to that of 23, thus
suggesting that the movement of the linker was tolerated.
Interestingly, the two-methylene elongation of the spacer length
in 28—structure 29—only slightly reduced the cytotoxicity, while
in the pair 23/27 it was detrimental for the activity.
2.4. Molecular modeling
To gain more knowledge about the mechanism of action of the
synthesized compounds we conducted a molecular modeling study
using a three-dimensional model of the structure of the tubulin
dimer–colchicine–vinblastine complex (PDB ID: 1Z2B). Conjugate
23 was chosen for the modeling among the synthesized com-
pounds with highest cytotoxicity and maximal clustering effect
(10, 23, 26, 28, 30). As though the clustering ability of the conju-
gates does not depend upon the presence of taxoter side chain
(see Section 2.3), molecular docking was performed to the colchi-
cine binding site located at b-subunit at the interface of a/b tubulin
The final structure–activity relationship was established for a
series of conjugates similar to 23 but with lipophilic moieties
distinct from adamantane. Replacement of a bridgehead core
for cyclohexane led to compound 30 with the same high cytotox-
icity and close tubulin clustering effect. Unexpectedly weak
activity was observed for ligand 33 with a phenyl substituent,
moreover its clustering ability was lost. Cytotoxicity of methyl
(31) and isopropyl (32) derivatives was almost equal to that of
colchicine and for the former only the colchicine-like depolymer-
izing effect was observed. These results indicate, that the pres-
ence of a bulky hydrophobic and non-aromatic moiety in the
studied conjugates plays an important role in the tubulin-cluster-
ing effect.
Since all the most cytotoxic compounds synthesized (10, 23, 26,
28, 30) possess the strongest tubulin clustering ability (see Tables 1
and 2), we suppose that this property might be an important factor
in the increment of mitostatic activity. This proposition is also
confirmed by comparison of the activities in the triad of structur-
ally close compounds colchicine—31–23. A noticeably higher cyto-
toxicity is observed for the compound 23 with high clustering
effect, while compounds with depolymerizing ability only have
equal lower cytotoxicity.
subunits. The study was carried out with the help of the AutoDock
Vina program, which takes ligands flexibility into account.13
Ligand-tubulin complexes with the best values of the corresponding
scoring function calculated by AutoDock Vina were chosen. The
obtained binding mode of 23 is presented in Figure 4.
According to the modeling results, while colchicine fragment of
23 binds to the corresponding site of b-subunit (formed by the res-
idues of Cys241, Leu248, Leu255, Asn 258, Met 259 etc.), the ada-
mantane moiety of the conjugate is located in a binding site
formed by hydrophobic residues Tyr224 and Val177 in the
unit. Thus, compound 23 connects two binding sites located on
two opposite surfaces of
and b tubulin subunitsà
a-sub-
a
.
Each of oxygen atoms of the linker ester group of 23 may form a
hydrogen bond with the hydroxyl of Tyr224. This hydrogen bond-
ing seems to be important for providing the location of the ada-
mantane moiety at the interface of two subunits and can be the
reason for the observed dependence of the activity on spacer
length.
A similar molecular modeling study was carried out for com-
pound 31, which possesses no tubulin clustering ability. The result
à
It should be mentioned that if the docking study is carried out using the model of
b-subunit only, then in the best binding mode of 23 its adamantane moiety is located
in an additional binding site formed by hydrophobic residues Ile347 and Pro348 along
with hydrophobic parts of Thr314 and Asn 349. Both carbonyl oxygens of the linker in
23 may form hydrogen bonds with Asn 258 and Met 259 residues and the importance
of these bonds might be the reason for the observed dependence of activity on spacer
length. However this docking study does not explain the difference in binding of
compounds with and without tubulin clustering ability as it does the docking to both
It should be mentioned, that the effect of the lead compound 23
on microtubule dynamics was also studied using cancer cell line,
different from A549, namely HeLa human cervical carcinoma cells.
In these cells we also observed the intensive formation of tubulin
clusters, which proves the fact that this ability of 23 is not only
limited to A549 cancer cells.
a- and b-subunits of the tubulin dimer (see below).

O. N. Zefirova et al. / Bioorg. Med. Chem. 19 (2011) 5529–5538
5535
Figure 5. The best-binding mode of compound 31 with tubulin (b-subunit is colored in red,
a-subunit in blue, hydrogen atoms are not shown for clarity, molecular surface
mesh is indicated). The direction of the linker of 31 towards the inner space of the b-subunit is clearly seen.
indicates that its best binding mode is different from that of 23
(Fig. 5). The linker and methyl group of 31 are not exposed to
the interface of tubulin subunits but are directed towards the inner
space of the b-subunit protein globe (Fig. 5). The analogues binding
mode (i.e. the direction of the side chain to the internal area of
b-subunit) was observed for conjugate 33 with phenyl substituent
(data not shown).
Since compounds 31 and 33 do not exhibit the tubulin-cluster-
ing effect it is logical to propose that this property might be con-
nected to the ability of the adamantane moiety of 23 to interact
4. Experimental section
4.1. Chemistry
Reaction control was carried out by thin-layer chromatography
on ‘Silufol’ plates. 1H NMR and 13C NMR spectra were recorded in
CDCl3 at 400 and 100 MHz correspondingly and are referenced to
residual chloroform (d 7.26 ppm 1H; d 77.0 ppm 13C). Chemical
shifts of the second isomer are given in square brackets. Attached
proton test (APT) technique was used for signal assignment in
some 13C NMR spectra. Electron impact mass spectra were
obtained with a typical voltage of 70 eV. GC-EIMS spectra were
recorded on ‘JMS-D300’ mass spectrometer with ‘HP-5890’
chromatograph (150 °C, 70 eV). Elemental analysis of the synthe-
sized compounds was performed on CNH analyser ‘Carlo-Erba’
ER-20. Infrared spectra (IR) were registered on ‘Thermo Nicolet
IR200’ apparatus in KBr plates and reported in cmÀ1. Melting
points were measured in a block with sealed capillaries and are
uncorrected.
with the hydrophobic site on the
a-subunit, providing a kind of
‘crosslinking’ of the tubulin dimer subunits. Binding of the conju-
gate 23 to other tubulin areas is also possible. Further experi-
ments are ongoing to understand the details of the mechanisms
by which the compounds effect microtubule dynamics and induce
tubulin clustering in cultured cells and to evaluate an anti-tumor
effect and an overall toxicity of the most active conjugates in
vivo.
Compounds were purified by flash and column chromatogra-
3. Conclusion
phy, which were performed on silica gel Acros (40–60 lm). Some
of the oily compounds obtained can be solidified into glassy solids
on standing in refrigerator.
In summary, a hybrid of adamantane-based paclitaxel mimetic
with colchicine (1) was proved to interfere with the microtubule
dynamics in an unusual manner, i.e. to promote disassembly of
microtubules followed by the formation of stable tubulin clusters.
This ability was found to be sensitive to the length and structure
of the linker. More pronounced tubulin clustering effect was ob-
served for the derivative of 1 without amino acid side chain—
conjugate 23. Structure—activity relationships for the analogues
of 23 indicate that both cytotoxicity and clustering ability are
very sensitive to the linker length and much less sensitive to
the shift of the position of linker attachment to adamantane.
Replacement of the bridgehead core in 23 for other lipophilic
groups revealed, that the presence of a bulky hydrophobic and
non-aromatic moiety in the studied conjugates plays an impor-
tant role in bringing about the tubulin-clustering effect. These
structure—activity data were partially explained by molecular
modeling studies.
4.2. Synthesis and characteristics of the final compounds
(synthetic protocols and analytical data for compounds 2, 3,
6–9, 12–21 are provided within Supplementary data)
4.2.1. Opening of the oxazolidine-type protected N-tert-
butoxycarbonylphenylisoserine (general procedure A)
The solution of a conjugate with oxazolidine-type protected
N-Boc-phenylisoserine and p-toluenesulfonic acid (pTSA) in
methanol was stirred 2.5 h at room temperature. After neutraliza-
tion by 5% solution of NaHCO3 methanol was evaporated and the
residue was extracted by CH2Cl2. The organic layer was dried over
Na2SO4 and subjected to column chromatography (methanol/
CH2Cl2 1:50).
4.2.1.1. (2R,3S)-N-(tert-Butoxycarbonyl)phenylisoserine 12-oxo-
The most cytotoxic compounds synthesized in this work (10, 23,
26, 28, 30) possess activity in nanomolar concentrations and are
several times more cytotoxic than colchicine.
12-(N-deacetylcolchicin-7-N-yl)dodecyl ester (4).
It was
prepared from (0.077 g, 0.082 mmol) and pTSA (0.012 g,
3

5536
O. N. Zefirova et al. / Bioorg. Med. Chem. 19 (2011) 5529–5538
0.115 mmol) according to procedure A. The product 4 was isolated
as white semisolid foam, 0.039 g, yield 59%). C46H62N2O11 requires:
C, 67.46; H, 7.63; N, 3.42. Found: C, 67.22; H, 7.68; N, 3.40. 1H
NMR: 1.21 (9H, s, tBu), 1.13–1.38 (14H, m), 1.54 (2H, m), 1.65
(2H, m), 1.81 (1H, m, J = 6.7, 6.0 Hz, H6colch), 2.17–2.36 (3H, m),
2.34 (1H, m, J = 13.2, 6.7, Hz, H5colch), 2.48 (1H, dd, J = 13.2,
6.0 Hz, H5colch), 3.63 (3H, s, OMe), 3.81 (1H, br s, OH), 3.88 (3H, s,
OMe), 3.92 (3H, s, OMe), 3.97 (3H, s, OMe), 4.18 (2H, t, J = 6.7 Hz,
CH2O(O)C), 4.43 (1H, m, H2isoserinyl), 4.65(1H, m, J = 6.6, 6.0 Hz,
H7colch), 5.18 (1H, m, H3isoserinyl), 5.40–5.51 (1H, br s, NH), 6.50
(1H, s, H4colch), 6.83 (1H, d, J = 10.8 Hz, H11colch), 7.04 (1H, m,
H4Ph), 7.27 (1H, d, J = 10.8 Hz, H12colch), 7.24–7.36 (5H, m), 7.46
(1H, s, H8colch); 13C NMR: 25.4, 25.67, 28.31, 28.42, 29.14, 29.25,
29.35 (C(Me)3), 29.46, 29.98, 36.29, 36.84, 52.11 (C7colch), 56.07
(OMe), 56.14 (OMe), 56.41 (OMe), 61.44 (OMe), 61.66 (C3isoserinyl),
66.55 (OMe), 73.64, 79.77 (C(Me)3), 107.37, 112.6, 125.69,
126.75, 127.64, 128.56, 130.71, 134.25, 135.34, 136.63, 139.36,
141.69, 151.27, 151.91, 153.51, 155.1, 164.02, 171.39 (C@O),
172.92 (C@O), 179.46 (C@O); IR: 3438, 3303(NN), 3060, 2929–
2854 (C–H), 1741 (C@O), 1716, 1678, 1655, 161 4, 1589, 1560,
1535, 1489, 1460, 1433, 1400, 1365, 1350, 1323, 1282, 1252,
1194, 1171, 1140, 1095, 1049, 1018, 987, 924, 904, 843, 800,
777, 702, 671, 592, 484.
4.2.2. Preparation of amides from dicarboxylic acids mono-
esters (general procedure B)
The solution of dicarboxylic acid mono-ester in CH2Cl2, 2-
ethoxy-1-ethoxycarbonyl-1,2-dihydroquinoline (EEDQ) and N-deac-
etylcolchicine were stirred together at room temperature for
12–60 h. The mixture was concentrated and the residue was
purified by column chromatography.
4.2.2.1.
etylcolchicine (23).
N-(7-Adamant-2-yloxy-7-oxoheptanoyl)-N-deac-
It was prepared by method B from 12
(0.094 g, 0.319 mmol), EEDQ (0.085 g, 0.344 mmol) and N-deac-
etylcolchicine (0.095 g, 0.266 mmol). Chromatographic purifica-
tion (ethyl acetate/petroleum ether 2:3, then CH2Cl2/methanol
20:1) gave 23 as yellowish oil (0.092 g, yield 54%). C37H47NO8 re-
quires: C, 70.12; H, 7.47; N, 2.21; Found: C, 69.98; H, 7.52; N,
2.21; MS (MALDI) m/z found for (M+Na)+ 656.8. 1H NMR:
1.24–1.34 (2H, m), 1.48–1.55 (6H, m), 1.68–1.95 (13H, m),
2.11–2.38 (6H, m), 2.47 (1H, dd, J = 11.4, 7.0 Hz, H5colch), 3.62
(3H, s, OMe), 3.86 (3H, s, OMe), 3.90 (c, 3H, MeO), 3.97 (c, 3H,
MeO), 4.63 (1H, m, J = 6.3, 5.8 Hz, H7colch), 4.85 (1H, m, H2adam),
6.50 (1H, s, H4colch), 6.84 (1H, d, J = 10.7 Hz, H11colch), 7.30 (1H,
d, J = 10.7 Hz, H12colch), 7.53 (1H, s, H8colch), 7.66 (1H, br s, NH);
13C NMR: 24.81, 25.00, 27.17, 31.73, 31.82, 34.61, 35.86, 36.30,
36.79, 37.18, 37.35, 52.19 (C7colch), 56.09 (OMe), 56.41 (OMe),
61.38 (OMe), 61.62 (OMe), 76.73 (C2adam), 107.35, 112.96,
125.77, 130.70, 134.29, 135.63, 136.97, 141.60, 151.13, 151.50,
152.3, 164.04, 171.11 (C@O), 172.70 (C@O), 179.13 (C@O); IR:
3286 (NH), 2925–2854 (C–H), 1726 (C@O), 1675, 1657, 1616,
1589, 1562, 1487, 1460, 1431, 1400, 1350, 1323, 1282, 1252,
1194, 1174, 1140, 1095, 1045, 1018, 985, 964, 923, 904, 841,
484.
4.2.1.2.
nyl]oxy}-4-adamantyl 8-oxo-8-(N-deacetylcolchicin-7-N-yl)octa-
noate (10). It was prepared from 8 (0.043 g, 0.04 mmol) and
1-{[(2R,3S)-N-(tert-Butoxycarbonyl)phenylisoseri-
pTSA (0.009 g, 0.047 mmol) according to procedure A. Column
chromatography of the product did not lead to the separation of
individual isomers and compound 10 was isolated as mixture of
trans- and cis-stereoisomers (ca. 7:3) (0.027 g, yield 70%, white
solid). Mp = 32–34 °C. C52H66N2O13 requires: C, 67.37; H, 7.18; N,
3.02. Found: C, 67.40; H, 7.16; N, 3.03. MS (MALDI) m/z found
for (M+Na)+ 950.1. 1H NMR: 1.43 (9H, s, tBu), 1.60–2.43 (28H, m),
2.52–2.58 (1H, m), 3.16 (1H, br s, OH), 3.67 [3.63] (3H, s,OMe),
3.92 [3.90] (3H, s, OMe), 3.97 [3.96] (3H, s, OMe), 4.02 [3.99]
(3H, s, OMe), 4.39 (1H, br, H2isoserinyl), 4.67 (1H, m, H7colch), 4.82
(0.3H, t, J4eq-3/5eq = 3.1 Hz, H4adam), 4.95 (0.7H, t, J4ax-3/5eq = 4.0 Hz,
H4adam), 5.26 (1H, m, H3isoserinyl), 5.41 (0.7H, d, J. = 8.5 Hz, NH),
5.52 (0.3H, d, J = 8.5 Hz, NH), 6.09–6.12 (1H, br s, NH), 6.55
[6.53] (1H, s, H4colch), 6.81–6.85 (1H, m, H11colch), 7.28–7.39 (7H,
m, arom.).
4.2.2.2.
etylcolchicine (24).
N-(6-Adamant-2-yloxy-6-oxohexanoyl)-N-deac-
It was prepared by method B from 13
(0.03 g, 0.170 mmol), EEDQ (0.03 g, 0.120 mmol) and N-deac-
etylcolchicine (0.034 g, 0.095 mmol). Chromatographic purifica-
tion (ethyl acetate/petroleum ether 1:2, then CH2Cl2/methanol
20:1) gave 24 as yellowish oil (0.054 g, yield 92). C36H45NO8 re-
quires: C 69.77, H 7.32, N 2.26. Found: C, 69.80; H, 7.30; N, 2.19.
1H NMR: 1.51–1.57 (2H, m), 1.63–1.88 (13H, m), 1.95–1.98 (4H,
m), 2.22–2.45 (6H, m), 2.52 (1H, dd, J = 13.2, 6.1 Hz, H5colch), 3.65
(3H, s, OMe), 3.90 (3H, s, OMe), 3.94 (3H, s, OMe), 3.99 (3H, s,
OMe), 4.65 (1H, m, J = 6.6, 6.0 Hz, H7colch), 4.90 (1H, m, H2adam),
6.53 (1H, s, H4colch), 6.81 (1H, br s, NH), 6.83 (1H, d, J = 10.8 Hz,
H11colch), 7.32 (1H, d, J = 10.8 Hz, H12colch), 7.44 (1H, s, H8colch); 13C
NMR: 24.59, 24.82, 26.97, 27.19, 29.94, 31.75, 31.82, 34.4, 35.7,
36.3, 36.81, 37.36, 52.15 (C7colch), 56.09 (OMe), 56.36 (OMe),
61.39 (OMe), 61.59 (OMe), 76.87 (C2adam), 107.33, 112.5, 125.66,
130.73, 134.21, 135.24, 136.53, 141.64, 151.22, 151.77, 153.46,
164.0, 172.27 (C@O), 172.94 (C@O), 179.48 (C@O).
4.2.1.3.
nyl]oxy}-4-adamantyl 5-oxo-5-[N-deacetylcolchicino]pentano-
ate (11). It was prepared from 9 (0.015 g, 0.08 mmol) and
1-{[(2R,3S)-N-(tert-Butoxycarbonyl)phenylisoseri-
pTSA (0.067 g, 0.0668 mmol) according to procedure A. Column
chromatography of the product did not lead to the separation of
individual isomers and compound 11 was isolated as mixture of
trans- and cis-stereoisomers (ca. 1:1) (0.032 g, yield 48%, white
semisolid foam). C49H60N2O13 requires: C, 66.50; H, 6.83; N 3.17.
Found: C, 66.52; H, 6.78; N, 3.15.1H NMR: 1.42 (9H, s, tBu),
1.69–2.55 (23, m), 3.24 (1H, s, OH), 3.67 [3.64] (3H, s, OMe), 3.91
[3.87] (3H, s, OMe), 3.95 [3.93] (3H, s, OMe), 4.00 (3H, s, OMe),
4.40 (1H, br s, H2isoserinyl), 4.62 (1H, m, H7colch), 4.92 [4.81] (1H,
two br s 1:1, H4adam, 5.24 (1H, d, J = 8.84 Hz, H3isoserinyl), 5.43
[5.38] (1H, d, J = 9.35 Hz, NH), 6.55 [6.45] (1H, s), 6.80–6.86 (1H,
two d, J = 11.2 Hz, H11colch), 7.20–7.48 (7H, m, arom.).13C NMR:
20.82 [20.70], 28.33 (C(Me)3), 29.70 [26.56], 30.26, 33.63, 34.65,
34.77, 35.12, 39.06, 39.32, 40.73, 52.11 (C7colch), 56.05 (CHNHBoc),
56.11(OMe), 56.37 (OMe), 61.40 (OMe), 61.57 (OMe), 74.16 [73.65]
(C2isoserinyl), 75.21 [75.00] (C4adam), 79.71 (C(Me)3), 82.04 [81.86]
(C1adam), 107.32. 112.43, 125.69 [125.62], 126.73 [126.49], 127.55
[127.36], 128.50 [128.40], 130.85, 134.29, 134.85, 135.29, 136.35,
141.55, 151.18, 151.36, 153.48, 154.97, 164.00, 171.85(C@O),
172.34(C@O), 172.42(C@O), 179.51(C@O).
4.2.2.3.
etylcolchicine (25).
N-(5-Adamant-2-yloxy-5-oxopentanoyl)-N-deac-
It was prepared by method B from 14
(0.027 g, 0.101 mmol), EEDQ (0.027 g, 0.109 mmol) and N-deac-
etylcolchicine (0.030 g, 0.084 mmol). Chromatographic purification
(ethyl acetate/petroleum ether 1:2, then CH2Cl2/methanol 20:1)
gave 25 as yellowish oil (0.033 g, yield 64%). C35H43NO8 requires:
C, 69.40; H, 7.16; N, 2.31. Found: C, 69.51; H, 7.11; N, 2.46. 1H
NMR: 1.51–1.54 (2H, m), 1.72-1.76 (4H, m), 1.81–1.98 (9H, m),
2.25–2.37 (5H, m), 2.41 (1H, m, J = 13.2, 6.7, Hz, H5colch), 2.53
(1H, dd, J = 13.2, 6.3 Hz, H5colch), 3.67 (3H, s, OMe), 3.91 (3H, s,
OMe), 3.95 (3H, s, OMe), 4.00 (3H, s, OMe), 4.65 (1H, m, J = 6.6,
6.0 Hz, H7colch), 4.90 (1H, m, H2adam), 6.52 (1H, s, H4colch), 6.85
(1H, d, J = 10.8 Hz, H11colch), 7.32 (1H, d, J = 10.8 Hz, H12colch), 7.33
(1H, br s, NH), 7.48 (1H, s, H8colch); 13C NMR: 20.86, 26.97, 27.16,
29.93, 31.74, 31.83, 34.04, 35.16, 36.31, 36.87, 37.34, 52.11

O. N. Zefirova et al. / Bioorg. Med. Chem. 19 (2011) 5529–5538
5537
(C7colch), 56.1 (OMe), 56.36 (OMe), 61.39 (OMe), 61.58 (OMe), 77.06
(C2adam), 107.33, 112.42, 125.65, 130.73, 134.19, 135.24, 136.42,
141.65, 151.22, 151.53, 153.47, 164.01, 171.86 (C@O), 172.58
(C@O), 179.45 (C@O).
4.2.2.7.
etylcolchicine (29).
N-(9-Adamant-1-yloxy-9-oxononanoyl)-N-deac-
It was prepared by method B from 18
(0.060 g, 0.186 mmol), EEDQ (0.048 g, 0.194 mmol) and N-deac-
etylcolchicine (0.038 g, 0.106 mmol). Chromatographic purifica-
tion (ethyl acetate/petroleum ether 2:3, then CH2Cl2/methanol
20:1) gave 29 as yellowish oil (0.037 g, yield 53%). C39H51NO8 re-
quires: C, 70.78; H, 7.77; N, 2.12. Found: C, 70.85; H, 7.80; N,
2.12.1H NMR: 1.23 (6H, m), 1.45–1.69 (10H, m), 1.84 (1H, m,
J = 11.8, 6.4 Hz, H6colch), 2.03–2.32 (14H, m), 2.37 (1H, m, J = 13.0,
6.4, Hz, H5colch), 2.49 (1H, dd, J = 13.0, 6.1 Hz, H5colch), 3.65 (3H, s,
OMe), 3.89 (3H, s, OMe), 3.93 (3H, s, OMe), 3.99 (3H, s, OMe),
4.60–4.69 (1H, m, J = 6.8, 6.4 Hz, H7colch), 6.52 (1H, s, H4colch), 6.84
(1H, d, J = 10.8 Hz, H11colch), 7.31 (1H, d, J = 10.8 Hz, H12colch),
7.29–7.34 (1H, br s, NH), 7.48 (1H, s, H8colch); 13C NMR: 25.05,
25.31, 28.83, 28.94, 29.13, 29.94, 30.77, 35.66, 36.13, 36.2, 36.75,
41.33, 52.15 (C7colch), 56.09 (OMe), 56.38 (OMe), 61.38 (OMe),
61.62 (OMe), 79.97 (C1adam), 107.32, 112.5, 125.67, 130.64,
134.23, 135.24, 136.56, 141.63, 151.23, 151.96, 153.45, 163.99,
172.84 (C@O), 173.03 (C@O), 179.46 (C@O).
4.2.2.4.
etylcolchicine (26).
N-(8-Adamant-2-yloxy-8-oxooctanoyl)-N-deac-
It was prepared by method B from 15
(0.060 g, 0.195 mmol), EEDQ (0.052 g, 0.213 mmol) and N-deac-
etylcolchicine (0.063 g, 0.177 mmol). Chromatographic purifica-
tion (ethyl acetate/ petroleum ether 1:2, then CH2Cl2/methanol
15:1) gave 26 as yellowish oil (0.036 g, yield 32%). C38H49NO8 re-
quires: C, 70.46; H, 7.62; N, 2.16. Found: C, 70.35; H, 7.72; N,
2.04.1H NMR: 1.33 (4H, m), 1.55–1.66 (6H, m), 1.74-1.84 (9H, m),
1.99–2.02 (4H, m), 2.21–2.45 (6H, m), 2.55 (1H, dd, J = 12.8,
6.6 Hz, H5colch), 3.67 (3H, s, OMe), 3.92 (3H, s, OMe), 3.96 (3H, s,
OMe), 4.02 (3H, s, OMe), 4.68 (1H, m, J = 6.8, 5.8 Hz, H7colch), 4.93
(1H, m, H2adam), 6.22 (1H, br s, NH), 6.55 (1H, s, H4colch), 6.85 (1H,
d, J = 10.4 Hz, H11colch), 7.34 (1H, d, J = 10.4 Hz, H12colch), 7.45 (1H,
s, H8colch); 13C NMR: 24.94; 25.22; 26.96; 27.17; 28.77; 28.92;
29.92; 31.73; 31.83; 34.74; 35.93; 36.29; 36.64; 37.35; 52.20
(C7colch); 56.08 (OMe); 56.36 (OMe); 61.34 (OMe); 61.57 (OMe);
76.66 (C1adam); 107.33; 112.68; 125.62; 130.64; 134.26; 135.33;
136.74; 141.60; 151.18; 152.30; 153.45; 163.97; 172.86 (C@O);
173.14 (C@O); 179.45 (C@O).
4.2.2.8. N-(7-Cyclohexyloxy-7-oxoheptanoyl)-N-deacetylcolchi-
cine (30).
It was prepared by method B from 19 (0.040 g,
0.165 mmol), EEDQ (0.044 g, 0.178 mmol) and N-deacetylcolchi-
cine (0.030 g, 0.084 mmol). Chromatographic purification (ethyl
acetate/petroleum ether 1:2, then CH2Cl2/methanol 20:1) gave 30
as yellowish oil (0.043 mmol, yield 51%). C33H43NO8 requires: C,
68.14; H, 7.45; N, 2.41. Found: C, 67.99; H, 7.38; N, 2.55. 1H
NMR: 1.22–1.43 (7H, m), 1.50–1.53 (5H, m), 1.67–1.74 (2H, m),
1.76–1.89 (3H, m), 2.19–2.31 (5H, m), 2.40 (1H, m, J = 13.2, 6.8,
Hz, H5colch), 2.52 (1H, dd, J = 13.2, 6.1 Hz, H5colch), 3.66 (3H, s,
OMe), 3.91 (3H, s, OMe), 3.94 (3H, s, OMe), 4.00 (3H, s, OMe),
4.65 (1H, m, J = 6.6, 6.1 Hz, H7colch), 4.72 (1H, m, H1cyclohex), 6.53
(1H, s, H4colch), 6.83 (1H, d, J = 10.8 Hz, H11colch), 6.89 (1H, br s,
NH), 7.32 (1H, d, J = 10.8 Hz, H12colch), 7.44 (1H, s, H8colch). 13C
NMR: 23.75, 24.83, 25.14, 25.37, 28.77, 31.34, 31.65, 32.57,
34.64, 36.62, 51.47 (C7colch), 55.96 (OMe), 56.87 (OMe), 60.84
(OMe), 61.39 (OMe), 72.36 (C1cyclohex), 109.22, 112.61, 128.86,
132.94, 135.11, 136.12, 137.25, 140.67, 152.04, 152.39, 153.48,
163.55, 173.13 (C@O), 175.07 (C@O), 179.13 (C@O).
4.2.2.5.
etylcolchicine (27).
N-(9-Adamant-2-yloxy-9-oxononanoyl)-N-deac-
It was prepared by method B from 16
(0.120 g, 0.372 mmol), EEDQ (0.096 g, 0.388 mmol) and N-deac-
etylcolchicine (0.076 g, 0.212 mmol). Chromatographic purifica-
tion (ethyl acetate/petroleum ether 1:5, then CH2Cl2/methanol
20:1) gave 27 as yellowish oil (0.165 g, yield 67%). C39H51NO8 re-
quires: C, 70.78; H, 7.77; N, 2.12. Found: C, 70.69; H, 7.78; N,
2.10. 1H NMR: 1.25 (6H, m), 1.52–1.58 (6H, m), 1.71–1.76 (4H,
m), 1.81–1.89 (5H, m), 1.96–2.0 (4H, m), 2.18–2.31 (5H, m), 2.37
(1H, m, J = 13.2, 6.6 Hz, H5colch), 2.49 (1H, dd, J = 13.2, 6.1 Hz,
H5colch), 3.65 (3H, s, OMe), 3.90 (3H, s, OMe), 3.94 (3H, s, OMe),
3.99 (3H, s, OMe), 4.65 (1H, m, J = 6.3, 6.0 Hz, H7colch), 4.89 (1H,
m, H2adam), 6.53 (1H, s, H4colch), 6.85 (1H, d, J = 10.8 Hz, H11colch),
7.32 (1H, d, J = 10.8 Hz, H12colch), 7.34 (1H, br s, NH), 7.48 (1H, s,
H8colch).13C NMR: 25.09, 25.29, 26.98, 27.20, 28.90, 29.12, 29.93,
31.76, 31.85, 34.8, 36.11, 36.31, 36.75, 37.37, 52.14 (C7colch),
56.09 (OMe), 56.36 (OMe), 61.37 (OMe), 61.6 (OMe), 76.66
(C2adam), 107.32, 112.54, 125.66, 130.64, 134.2, 135.25, 136.59,
141.63, 151.21, 151.98, 153.45, 163.99, 172.83 (C@O), 173.25
(C@O), 179.47 (C@O).
4.2.2.9.
(31).
N-(7-Methoxy-7-oxoheptanoyl)-N-deacetylcolchicine
It was prepared by general method B from 20 (0.050 g,
0.287 mmol), EEDQ (0.077 g, 0.311 mmol) and N-deacetylcolchi-
cine (0.085 g, 0.239 mmol). Chromatographic purification (ethyl
acetate/petroleum ether 1:2, then CH2Cl2/methanol 10:1) gave 31
as yellowish oil (0.05 g, yield 41%). C52H35NO8 requires: C, 65.48;
H, 6.87; N, 2.73. Found: C, 65.45; H, 7.83; N, 2.74. 1H NMR: 1.30
(2H, m, J = 1.9 Hz), 1.53–1.62 (4H, m, J = 7.6, 1.9 Hz), 1.85 (1H, m,
J = 11.8, 6.7 Hz, H6colch), 2.21–2.27 (5H, m), 2.39 (1H, m, J = 13.2,
7.0 Hz, H5colch), 2.47–2.55 (1H, m, J = 13.2, 6.7 Hz, H5colch), 3.63
(3H, s, OMe), 3.65 (3H, s, OMe), 3.89 (3H, s, OMe), 3.93 (3H, s,
OMe), 3.99 (3H, s, OMe), 4.59–4.68 (1H, m, J = 6.5, 5.8 Hz, H7colch),
6.52 (1H, s, H4colch), 6.84 (1H, d, J = 10.7 Hz,, H11colch), 7.16 (1H, br
s, NH), 7.31 (1H, d, J = 10.7 Hz, H12colch), 7.46 (1H, s, H8colch); 13C
NMR: 24.51, 24.98, 28.70, 29.92, 33.78, 35.96, 36.89, 51.45
(OMe), 52.11 (C7colch), 56.11 (OMe), 56.35 (OMe), 61.38 (OMe),
61.59 (OMe), 107.37, 112.48, 125.66, 130.67, 134.16, 135.25,
136.48, 141.70, 151.24, 151.61, 153.13, 153.48, 164.02, 172.47
(C@O), 179.44 (C@O).
4.2.2.6.
etylcolchicine (28).
N-(7-Adamant-1-yloxy-7-oxoheptanoyl)-N-deac-
It was prepared by method B from 17
(0.045 g, 0.153 mmol), EEDQ (0.047 g, 0.190 mmol) and N-deac-
etylcolchicine (0.068 g, 0.190 mmol). Chromatographic purifica-
tion (ethyl acetate/petroleum ether 2:3, then CH2Cl2/methanol
20:1) gave 28 as yellowish oil (0.090 g, yield 93%). C37H47NO8 re-
quires: C, 70.12; H, 7.47; N, 2.21. Found: C, 70.21; H, 7.48; N,
2.18. 1H NMR: 1.38 (2H, m), 1.49–1.61 (10H, m), 1.86 (1H, m,
J = 11.8, 6.6, 6.4 Hz, H6colch), 2.02–2.29 (14H, m) 2.37 (1H, m,
J = 12.9, 6.6, Hz, H5colch), 2.49 (1H, dd, J = 12.9, 6.1 Hz, H5colch),
3.64 (3H, s, OMe), 3.88 (3H, s, OMe), 3.92 (3H, s, OMe), 3.96 (3H,
s, OMe), 4.63 (1H, m, J = 6.4, 5.8 Hz, H7colch), 6.51 (1H, s, H4colch),
6.83 (1H, d, J = 10.8 Hz, H11colch), 7.30 (1H, d, J = 10.8 Hz, H12colch),
7.38 (1H, br s, NH), 7.47 (1H, s, H8colch); 13C NMR: 24.73, 25.06,
28.69, 29.93, 30.75, 35.44, 35.88, 36.18, 36.70, 41.31, 52.17
(C7colch), 56.09 (OMe), 56.34 (OMe), 61.35 (OMe), 61.58 (OMe),
80.01 (C1adam), 107.32, 112.47, 125.68, 130.68, 134.21, 135.2,
136.55, 141.63, 151.21, 151.93, 153.43, 163.99, 172.64 (C@O),
172.85 (C@O), 179.43 (C@O).
4.2.2.10. N-(7-Isopropyloxy-7-oxoheptanoyl)-N-deacetylcolchi-
cine (32).
(0.017 g, 0.084 mmol), EEDQ (0.033 g, 0.134 mmol) and N-deac-
etylcolchicine (0.02 g, 0.056 mmol). Chromatographic purification
(ethyl acetate/petroleum ether 1:2, then CH2Cl2/methanol 10:1)
It was prepared by general method B from 21

5538
O. N. Zefirova et al. / Bioorg. Med. Chem. 19 (2011) 5529–5538
gave 32 as yellowish oil (0.032 g, yield 70%). C30H39NO8 requires:
of compound exposure. Then the supernatant was removed, and
100 L of DMSO containing 10% SDS and 0.6% acetic acid was
C, 66.52; H, 7.26; N, 2.59. Found: C, 66.47; H, 7.13; N, 2.57. MS (EI)
l
1
m/z, (%): 542 (M+H+, 3), 540 (M-H+, 11), 312 (40), 43 (100). H NMR:
added to each well. Resulting formazan crystals were solubilised
by thorough mixing on a plate shaker. Optical density was mea-
sured at 590 nm with 690 nm reference filter using a EL808 Ultra
Microplate Reader (BioTek Instruments, Winooski, USA). Experi-
ments for all compounds were repeated at least three times and
EC50 values were determined by sigmoidal curve fitting using
Excel-based software.
1.19 (3H, d, J = 6.3 Hz, CHMe2), 1.20 (3H, d, J = 6.3 Hz, CHMe2),
1.30 (2H, m, J = 1.9 Hz), 1.53–1.62 (4H, m, J = 7.6, 1.9 Hz), 1.85
(1H, m, J = 11.8, 6.8 Hz, H6colch), 2.21–2.27 (5H, m), 2.38 (1H, m,
J = 13.1, 7.0 Hz, H5colch), 2.47–2.54 (1H, m, J = 13.1, 6.5 Hz, H5colch),
3.66 (3H, s, OMe), 3.90 (3H, s, OMe), 3.94 (3H, s, OMe), 4.00 (3H,
s, OMe), 4.62–4.68 (1H, m, J = 6.8, 5.3 Hz, H7colch), 4.97 (1H,
J = 6.3 Hz, CHMe2), 6.53 (1H, s, H4colch), 6.84 (1H, d, J = 10.8 Hz,
H11colch), 6.99 (1H, m, J = 5.3 Hz, NH), 7.31 (1H, d, J = 10.8 Hz,
H12colch), 7.47 (1H, s, H8colch); 13C NMR: 21.83 CHMe2, 24.62, 25.02,
28.71, 29.92, 34.41, 39.95, 36.85, 52.11 (C7colch), 56.11 (OMe),
56.35 (OMe), 61.38 (OMe), 61.59 (OMe), 67.38 (CHMe2), 107.35,
112.48, 125.67, 130.69, 134.18, 135.24, 136.49, 141.69, 151.24,
151.67, 153.47, 164.01, 172.52, 173.23 (C@O), 179.47 (C@O).
4.3.3. Immunofluorescence staining of cellular microtubules
For microtubule staining A549 cells were cultured in 12-well
plates on small glass coverslips (11 mm diameter) at a density of
2 Â 104 cells per coverslip. Cells were incubated with selected
compounds or colchicine and taxol as positive controls at concen-
trations of 1 and 5 lM at 37 °C and 5% CO2 for 8 h. 0.5% DMSO
served as a negative control. The whole process of cell fixation
4.2.2.11. N-(7-Phenyloxy-7-oxoheptanoyl)-N-deacetylcolchicine
and staining was described previously.8 Fixed cells were labelled
(33).
It was prepared by method
B
from 22 (0.025 g,
with mouse monoclonal antibody against a-tubulin at a dilution
0.106 mmol), EEDQ (0.030 g, 0.121 mmol) and N-deacetylcolchi-
cine (0.034 g, 0.095 mmol). Chromatographic purification (ethyl
acetate/petroleum ether 1:2, then CH2Cl2/methanol 20:1) gave 33
as yellowish oil (0.04 g, yield 73%). C33H37NO8 requires: C, 68.85;
H, 6.48; N, 2.43. Found: C, 68.88; H, 6.50; N, 2.39. 1H NMR: 1.42
(2H, m), 1.70–1.84 (5H, m), 2.26 (1H, m, H6colch), 2.40 (1H, m,
J = 13.2, 6.8, Hz, H5colch), 2.49–2.60 (5H, m, Hb + H5colch), 3.67 (3H,
s, OMe), 3.91 (3H, s, OMe), 3.94 (3H, s, OMe), 3.99 (3H, s, OMe),
4.65 (1H, m, J = 6.6, 6.0 Hz, H7colch), 6.54 (1H, s, H4colch), 6.86 (1H,
d, J = 10.8 Hz, H11colch), 6.96 (1H, br s, NH), 7.07 (2H, m, J = 8.0 Hz,
H2Ph), 7.21 (1H, t, J = 7.7 Hz, H4Ph), 7.34 (1H, d, J = 10.8 Hz, H12colch),
7.36 (2H, m, J = 8.0, 7.7 Hz, H3Ph) 7.54 (1H, s, H8colch); 13C NMR:
24.52, 25.04, 28.69, 29.97, 34.12, 35.93, 36.91, 52.15 (C7colch),
56.15 (OMe), 56.39 (OMe), 61.42 (OMe), 61.64 (OMe), 107.44,
112.72, 121.63, 125.7, 125.8, 129.38, 130.8, 134.24, 135.39,
136.67, 141.74, 150.76, 151.27, 151.85, 153.54, 164.06, 172.6
(C@O), 177.3 (C@O), 179.55 (C@O).
of 1:400 (Sigma, St. Louis. USA), followed by incubation of Alexa
Fluor488 labelled goat anti-mouse IgG at a dilution of 1:200 (Invit-
rogen, Germany). Images of all samples were acquired with a
Nikon Diaphot 300 inverted microscope (Nikon GmbH, Düsseldorf,
Germany) equipped with a cooled charge-couple device camera
system (SenSys; Photometrics, Munich, Germany).
Acknowledgements
This work was done in the frame of a Scientific Cooperation
Agreement between Moscow and Rostock Universities and supported
by the Alexander von Humboldt Foundation (3.4-Fokoop-Rus/
1015567). A part of the work was supported by Russian Foundation
for Basic Research (No. 00879/12088) and Russian Academy of
Sciences.
Supplementary data
4.3. Biology
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2011.07.040.
4.3.1. Cell culture
A549 human lung epithelial carcinoma cells (CCL-185™) were
cultured with Dulbecco’s modified Eagle medium (DMEM) con-
taining 10% fetal bovine serum and 1% antibiotic penicillin/strepto-
mycin at 37 °C under a humidified 5% CO2 atmosphere.
References and notes
1. Kingston, D. G. I. Chem. Commun. 2001, 867.
2. Zefirova, O. N.; Nurieva, E. V.; Ryzhov, A. N.; Zyk, N. V.; Zefirov, N. S. Russ. J. Org.
Chem. 2005, 41, 329.
3. Breen, E. S.; Walsh, J. J. Curr. Med. Chem. 2010, 17, 609.
4. Zefirova, O. N.; Nurieva, E. V.; Lemcke, H.; Ivanov, A. A.; Shishov, D. V.; Weiss, D.
G.; Kuznetsov, S. A.; Zefirov, N. S. Bioorg. Med. Chem. Lett. 2008, 18, 5091.
5. Kuznetsov, S. A.; Rodionov, V. I.; Bershadsky, A. D.; Gelfland, V. I.; Rosenblat, V.
A. Cell Biol. Int. Rep. 1980, 4, 1017.
6. Weiss, D. G. In Cell Biology: A Laboratory Handbook; Celis, J. E., Ed.; Academic
Press: San Diego, 2006; Vol. 3, pp 57–65.
7. Mosmann, T. J. Immunol. Methods 1983, 65, 55.
8. Al-Haddad, A.; Shonn, M. A.; Redlich, B.; Blocker, A.; Burkhardt, J. K.; Yu, H.;
Hammer, J. A., 3rd; Weiss, D. G.; Steffen, W.; Griffiths, G.; Kuznetsov, S. A. Mol.
Biol. Cell 2001, 12, 2742.
9. Lebeau, L.; Ducray, P.; Mioskowski, C. Synth. Commun. 1997, 27, 293.
10. (a) Bourzat, J. D.; Commerçon, A. Tetrahedron Lett. 1993, 34, 6049; (b) Didier, E.;
Fouque, E.; Taillepied, I.; Commerçon, A. Tetrahedron Lett. 1994, 35, 2349.
11. Hill, J. W.; Carothers, W. H. JACS 1933, 55, 5023.
4.3.2. MTT cytotoxicity assay
Cells were seeded in 96-well plates at a density of 3 Â 103 cells
per well. Stock solutions of test compounds were prepared in
dimethylsulfoxide (DMSO). Cells were treated for 24 h with
selected compounds at 1–12000 nM or with colchicine at
5–100 nM or taxol at 1–15 nM as positive controls (8 wells for each
concentration). DMSO (0.4%) served as a negative control. The
number of surviving cells was determined by the colorimetric
MTT assay.7 MTT (3-(4,5-dimethylthiazolyl-2)-2,5-diphenyl-2H-
tetrazoliumbromid, Roth GmbH, Karlsruhe, Germany) was pre-
pared at 5 mg/mL in phosphate buffered saline (PBS) and filtered
12. Zefirova, O. N.; Diikov, A. G.; Zyk, N. V.; Zefirov, N. S. Russ. Chem. Bull., Int. Ed.
2007, 56, 680.
through a 0.22
lm filter. 20 lL of sterile MTT solution (final con-
centration 0.5 mg/mL) was added to each well 2 h before the end
13. Trott, O.; Olson, A. J. J. Comput. Chem. 2010, 31, 455.