Synthesis and biological activity of 5-aza-ellipticine derivatives

Article abstract of DOI:10.1016/j.bmcl.2006.09.093

Novel 5-aza-ellipticine derivatives were synthesized and tested as antitumor agents. The new compounds were prepared more readily than the analogous ellipticine derivatives, which are known to be potent anti-tumor agents Although the novel 5-aza-elliptici

Full text of DOI:10.1016/j.bmcl.2006.09.093

Bioorganic & Medicinal Chemistry Letters 17 (2007) 2380–2384
Synthesis and biological activity of 5-aza-ellipticine derivatives
Deborah L. Moody, Marcin Dyba, Teresa Kosakowska-Cholody,
Nadya I. Tarasova* and Christopher J. Michejda
Structural Biophysics Laboratory, Molecular Aspects of Drug Design Section, Structural Biophysics Laboratory,
NCI-Frederick, PO Box B Frederick, MD 21702, USA
Received 24 July 2006; revised 28 September 2006; accepted 29 September 2006
Available online 4 October 2006
Abstract—Novel 5-aza-ellipticine derivatives were synthesized and tested as antitumor agents. The new compounds were prepared
more readily than the analogous ellipticine derivatives, which are known to be potent anti-tumor agents Although the novel 5-aza-
ellipticine derivatives are not as biologically active as their corresponding ellipticine analogues, the new compounds represent a new,
readily accessible class of heteroaromatic catalytic inhibitors of topoisomerase II and possible anti-tumor agents.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Ellipticine (1) was first identified in 1959 as a compound
in the leaves of an Australian evergreen, Ochrosia ellip-
tica labil (Apocynaceae family).1 Ellipticine is one of
many carbazole alkaloids that exhibit biological, espe-
cially anti-tumor, activities.2 Many synthetic variants
of the natural product were made and tested for anti-tu-
mor activity. The most likely target of ellipticine is topo-
isomerase II, whose inhibition results in the observed
cytotoxicity. However, the low water solubility of ellip-
ticine at physiological pH, as well as its systemic toxici-
ty, prevents its use as a therapeutic agent. Several simple
structural modifications to ellipticine derivatives gave
compounds with increased toxicity (such as a methoxy
substituent at the C9 position and a methyl substituent
at the C11 position).3,4
at low nanomolar concentrations but is relatively nonse-
lective, while the heptagastrin conjugate, while slightly
less toxic to sensitive cells, was completely selective to
only those cells that expressed the CCK-B receptor.
Compound 2 was synthesized in thirteen steps with an
overall 7% yield. Recently, Zhang et al.7 reported the
synthesis of 5-aza analogues of ellipticines (3) via an
interesting radical cyclization reaction. We wished to
examine the anti-tumor properties of these novel com-
pounds because their synthesis was more readily accom-
plished than the original ellipticines and because they
offered the possibility of additional elaboration of the
original structures.
In this study, novel 5-aza-ellipticine derivatives were
synthesized, which contain a C9 methoxy substituent
and a C11 methyl substituent. Consistent with known
Previously, the novel ellipticine derivative (2) was cou-
pled to heptagastrin, which binds with high affinity to
the gastrin/cholecystokinin type
B
receptor.5 The
hypothesis was that since CCK-B receptors are over-ex-
pressed in some gastrointestinal cancers,6 the heptagas-
trin conjugate could be delivered very selectively to
CCK-B positive cancer cells by receptor-mediated endo-
cytosis. We demonstrated that heptagastrin conjugate
was delivered to lysozomes in CCK-B receptor positive
cells where it was processed to release the cytotoxic ellip-
ticine derivative. Compound 2 is cytotoxic to tumor cells
Keywords: 5-Aza-ellipticine; Anti-tumor agents; Topoisomerase II;
Ellipticine.
*
Corresponding author. Tel.: +1 301 846 5225; fax: +1 301 846
6231; e-mail: tarasova@ncifcrf.gov
Figure 1. Ellipticines.
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.09.093

D. L. Moody et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2380–2384
2381
Scheme 1. Reagents and conditions: (i) Pd(PPh3)2Cl2, i-Pr2NEt, CuI, propyne, DMF, 86%; (ii) triphosgene, toluene, 96%; (iii) benzene, p-xylene,
17%.19
Scheme 2. Reagents and conditions: (i) for 8, 3,30-diamino-N-methyldipropylamine; for 9, N-(3-aminopropyl)-1,3-propanediamine; for 10, 1,4-bis(3-
aminopropyl)piperazine.20
structure–activity relationships of ellipticine, many of
these novel 5-aza-ellipticine derivatives inhibited the
growth of HCT116 cells, with approximately the same
potency as ellipticine.
oped by Zhang et al.7,14,15 The precursor 416 was con-
verted to the isocyanate 5 using triphosgene.17 The
radical cycloaddition of 5 with phosphorane 6 occurs
at elevated temperature and is accompanied by precipi-
tation of triphenylphosphine oxide.18 The yield of this
reaction is only 17%, but since the precursors are readily
available, the final product can be prepared in quantity,
in spite of the low overall yield.
The synthesis of the 1-chloro-5-aza-ellipticine (7) was
accomplished in six steps with a 6.7% overall yield
(Scheme 1) using a modification of the procedure devel-
Scheme 3. Reagents: (i) methyl trifluoromethane sulfonate, DCM; (ii) for 11 and 13, N-(3-aminopropyl)-1,3-propanediamine; for 12, 1,4-bis(3-
aminopropyl)piperazine.
Scheme 4. Reagents: (i) ethyl trifluoromethane sulfonate, DCM; (ii) N-(3-aminopropyl)-1,3-propanediamine.

2382
D. L. Moody et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2380–2384
Several modifications of 7 were carried out. The chlorine
in position 1 was readily replaced by three amine side
chains that could be utilized in conjugation reactions
with peptide ligands (Scheme 2). Data from our labora-
tory and others suggest that the basic side chains fre-
quently enhance the activity of intercalating drugs,
probably by binding to the minor groove of DNA.8–11
Ethylation of 7 also resulted in very unstable products,
but replacement of the chloro group by a polyamine side
produced a stable product 14, along with the hydrolysis
product 15 and the dimer 16.22
In vitro toxicity of the novel 5-aza-ellipticine derivatives
was determined by the MTT assay12 using human colon
adenocarcinoma cells HCT116 and murine NIH 3T3
cells. The latter was of interest because we had previous-
ly transfected the CCK-B receptor into this cell line and
showed that this line was sensitive to ellipticine deriva-
tives.5 The results (Table 1) indicate that some of the
novel compounds were moderately active cytotoxic
agents (IC50 values in the sub-micromolar range). The
1-chloro substituted 5-aza-ellipticine derivative, 7, did
not inhibit cell growth. Compounds 8 and 9 while more
potent than 7, were an order of magnitude less active
than the corresponding ellipticine derivative 2. Interest-
ingly, compound 12, which has a piperizine-containing
side chain, was inactive, in contrast to other agents
where that structural feature enhanced activity. The
5-aza ellipticine derivative with the 5-ethyl substituent,
14, had an order of magnitude greater potency than
the derivative with the 5-methyl substituent 11. In
contrast, the 5-methyl derivative, 13, was an order of
magnitude more toxic than the 5-ethyl derivative, 15.
The one dimer tested, 16, has approximately the same
potency as the monomer 8. These data suggest that
appropriately substituted 5-aza-ellipticine derivatives
may produce molecules with interesting anti-tumor
properties (Fig. 1).
We also prepared derivatives where the endocyclic nitro-
gen at position 5 was quaternized with a methyl group
(Scheme 3). The introduction of a positive charge in
the molecule improved water solubility but we also
expected profound difference in the biological activity.
The synthesis was accomplished by quaternization of
the 1-chloro derivative. The side chains were attached
directly without purification because the quaternary salt
was very unstable. However, during the purification of
both 11 and 12, some of the chloro compound was
hydrolyzed to 13, which was also isolated.21 It is inter-
esting to note that hydrolysis of the chloro substituent
was especially facile in compounds where the 5-position
was quaternized, perhaps by enhancement of Meisenhei-
mer complexes (Scheme 4).
Table 1. Cell growth inhibition by 5-aza-ellipticine derivatives
Compound
IC50 (lM) on
HCT116 cells
IC50 (lM)
on 3T3 cells
Ellipticine (1)
0.3
0.07
>1
0.3
N/A
2
0.003
0.02
>1
(2)
7
8
0.6
1
Ellipticine 1 is known to be an inhibitor of topoiso-
merase II (topoII).13 We therefore tested the ability
of 8, an active derivative of the 5-aza-ellipticine series,
to inhibit the de-catenation of catenated kinetoplast
DNA (kDNA) by topoisomerase II. Topoisomerase
II-mediated kDNA de-catenation assays were per-
formed using the the manufacturer’s recommended
procedure (TopoGen, Inc., Port Orange, FL). The
9
11
12
13
14
15
16
N/A
N/A
N/A
N/A
N/A
N/A
>10
0.2
0.4
3
0.3
N/A, not tested.
Figure 2. Effect of ellipticines on catalytic activity of topoisomerase II in vitro. Under these conditions, the catenated kinetoplast DNA does not
move from the gel origin. Two bands appear in the gel when the topoisomerase catalyzes the double strand break and re-ligation of the DNA.
Concentration-dependent delay in the appearance of the two bands correlates with inhibition of topoisomerase II.

D. L. Moody et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2380–2384
2383
Hollingshead, M. G.; Hariprakasha, H. K.; Michejda, C.
J. J. Med. Chem. 2005, 48, 4474.
13. Monnot, M.; Mauffret, O.; Simon, V.; Lescot, E.; Psaume,
B.; Saucier, J. M.; Charra, M.; Belehradek, J., Jr.;
Fermandjian, S. J. Biol. Chem. 1991, 266, 1820.
de-catenation reaction requires a double strand break
and re-ligation, a hallmark of topoisomerase II. Ellip-
ticine 1 was chosen as a positive control. The data in
Figure 2 indicate that ellipticine 1 as well as the 5-aza-
ellipticine derivative 8 are potent catalytic inhibitors of
the enzyme.
14. Unless indicated, reagents were commercially available
and used without further purification. Flash chromatog-
raphy was performed on silica under argon pressure or on
an ISCO CombiFlash Companion instrument and Silicy-
cle 230–400 mesh cartridges. HPLC purification was
performed using preparative XTerraꢁ Prep RP18 column
from Waters (19 mm id · 300 mm, 10 lm pore size).
Acetonitrile/water gradient with 0.1% TFA was used as
the eluent for all purifications. Mass spectroscopic analysis
was preformed using a Agilent Series 1100 MSD HPLC–
MS with atmospheric pressure ionization electrospray
(API-ES). The colums used were the Agilent Poroshell 300
SB-C18 (2.1 mm id · 75 mm, 5 lm pore size) and the
Zorbax 300 SB-C18 (2.1 mm id · 150 mm, 5 lm pore size).
NMR’s were recorded on Varian 200 or 400 MHz
instruments. The de-catenation assay was preformed using
a kit from TopoGen, Inc. (Port Orange, FL).
15. Compounds 8–10 were the products of nucleophilic
aromatic substitution on 7 under standard conditions.
Because the C1 position on 7 was the only site with an
activated leaving group for nucleophilic substitution, the
structures were verified using HPLC–ESI-MS. The com-
pounds identified were either the starting material 7 or the
product 8, 9 or 10. Compounds 11–16 were the products
of nucleophilic aromatic substitution on either the
5-methylated or 5-ethylated derivative of 7. Byproducts
that were formed were identified with the products by
using HPLC–ESI-MS.
16. Synthetic procedure for 4. m-Iodophenol, 1, (5.5 g,
25 mmol) was dissolved in water containing KOH (1.4 g,
25 mmol). A cold solution of diazotized sulfanilic acid
(5.09, 27.5 mmol) was added to the mixture. The mixture
stood for a 30 min and then sodium hydrosulfite (12 g,
69 mmol) was added. The solution changed color imme-
diately upon addition of the sodium hydrosulfite and the
solution was stirred at 45 ꢂC for 15 min. Diethyl ether was
added to the mixture, the mixture was filtered and the
filtrate was concentrated. The solid was recrystallized from
hot water to yield needles. Methyl iodide was added
(0.98 ml, 15.7 mmol) to 4-amino-3-iodophenol (3.905 g,
16.6 mmol) in 109 ml DMF in the presence of CsCO3
(13.45 g, 41.3 mmol). The reaction was left at room
temperature for 48 h. The mixture was then diluted with
water, extracted with diethyl ether, washed with water,
then dried over MgSO4 and filtered. The product was
isolated after flash column chromatography on silica gel
(diethyl ether/h exane 3:7) to yield a light brown oil. NMR
(CDCl3) 1H d7.21 (1H, d, J = 2.8 Hz), 6.77 (1H, dd,
J = 8.7, 2.8 Hz), 6.70 (1H, d, J = 8.7 Hz), 3.72 (3H, s). An
outgassed (argon) solution of Pd(PPh3)2Cl2 (0.017 g,
0.024 mmol) and CuI (0.009 g, 0.047 mmol), N,N-diiso-
propylethylamine (0.35 ml, 2.0 mmol) and 4-amino-3-
iodo-1-methoxybenzene (0.200 g, 0.80 mmol) in 5 ml
DMF was prepared. Gaseous propyne (290 ml,
12.9 mmol) was introduced with a syringe and the reaction
was stirred at room temperature for 12 h. The reaction
mixture was diluted with 20 ml dichloromethane and
washed with 20 ml saturated NH4Cl. The organic layer
was washed with water and dried over Na2SO4. The
organic layer was concentrated and subjected to flash
chromatography on silica gel with diethylether/hexanes
3:7. NMR (CDCl3) 1H d 6.81 (1H, d, J = 2.9 Hz), 6.72
(1H, dd, J = 8.7, 2.9 Hz), 6.63 (1H, D, J = 8.7 Hz), 3.72
(3H, s), 2.11 (3H, s).
It is interesting to note that while 1 and 8 were about
equally potent against HCT116 colon cancer cells in vi-
tro, compound 8 is actually a more potent catalytic
inhibitor of topoisomerase II.
In conclusion, 5-aza-ellipticine derivatives are readily
available compounds that have some interesting anti-tu-
mor properties. Although the currently available deriva-
tives are not as active as the corresponding ellipticines,
this new heteroaromatic class of compounds should con-
tinue to be examined because of the strong indication of
anti-tumor activity and some clear advantages (ease of
synthesis and possibly better PK/PD properties) over
other polycyclic heteroaromatic anti-tumor agents. It
is likely that different substitution pattern on the basic
5-aza-ellipticine scaffold may yield compounds with bet-
ter anti-cancer activity.
Acknowledgments
This research was supported by the Intramural Research
Program of the NIH, National Cancer Institute, Center
for Cancer Research. We also thank the Biophysics
Resource of the Structural Biophysics Laboratory,
NCI-Frederick, for assistance with LC–MS studies.
References and notes
1. Garbett, N. C.; Graves, D. E. Curr. Med. Chem. Antican-
cer Agents 2004, 4, 149.
2. Knolker, H. J.; Reddy, K. R. Chem. Rev. 2002, 102,
4303.
3. Peczynska-Czoch, W.; Pognan, F.; Kaczmarek, L.; Bora-
tynski, J. J. Med. Chem. 1994, 37, 3503.
4. Ducrocq, C.; Wendling, F.; Tourbez-Perrin, M.; Rivalle,
C.; Tambourin, P.; Pochon, F.; Bisagni, E.; Chermann, J.
C. J. Med. Chem. 1980, 23, 1212.
5. Czerwinski, G.; Tarasova, N. I.; Michejda, C. J. Proc.
Natl. Acad. Sci. U.S.A. 1998, 95, 11520.
6. Kusyk, C. J.; McNiel, N. O.; Johnson, L. R. Am. J.
Physiol. 1986, 251, G597.
7. Zhang, Q.; Shi, C.; Zhang, H. R.; Wang, K. K. J. Org.
Chem. 2000, 65, 7977.
8. Moron, J.; Rautureau, M.; Huel, C.; Pierre, A.; Berthier,
L. K.; Atassi, G.; Bisagni, E. Anticancer Drug Des. 1993,
8, 399.
9. Escude, C.; Nguyen, C. H.; Mergny, J. L.; Sun, J. S.;
Bisagni, E.; Garestier, T.; Helene, C. J. Am. Chem. Soc.
1995, 117, 10212.
10. Cholody, W. M.; Hernandez, L.; Hassner, L.; Scudiero, D.
A.; Djurickovic, D. B.; Michejda, C. J. J. Med. Chem.
1995, 38, 3043.
11. Cholody, W. M.; Kosakowska-Cholody, T.; Michejda, C.
J. Cancer Chemother. Pharmacol. 2001, 47, 241.
12. MTT assay procedure was preformed as previously
described Cholody, W. M.; Kosakowska-Cholody, T.;

2384
D. L. Moody et al. / Bioorg. Med. Chem. Lett. 17 (2007) 2380–2384
17. Synthesis of 5. Triphosgene (0.3182 g, 1.1 mmol) was
dissolved in 6 ml freshly dried anhydrous toluene.
Compound 4 (0.471 g, 1.6 mmol) was added to anhy-
drous TEA (0.81 ml, 5.8 mmol) in 9 ml anhydrous
toluene and this solution was added dropwise to the
triphosgene solution. The mixture was stirred at room
temperature for 30 min and concentrated. The residual
white solid was dissolved in 10% diethyl ether in hexanes
and filtered through a short silica gel column. NMR
(CDCl3) 1H d 6.91 (1H, d, J = 8.9 Hz), 6.89 (1H, d,
J = 2.9 Hz), 6.75 (1H, dd, J = 8.9, 3.0 Hz), 3.76 (3H, s),
2.12 (3H, s).
18. Synthesis of 6. Fresh, anhydrous TEA (1.72 ml,
12.3 mmol) was added dropwise to a solution of Ph3PBr2
(2.6 g, 6.2 mmol), 4-amino-2-chloropyridine (0.7192 g,
5.6 mmol), in 42 ml freshly dried anhydrous p-xylene.
The solution was heated at 105 ꢂC for 5 h and concen-
trated. The reaction mixture was purified by chromatog-
raphy on a silica gel column using 20% diethyl ether in
chloroform as the eluant. NMR (CDCl3) 1H d 7.79 (1H,
dd, J = 5.7, 1.0 Hz), 7.47–7.74 (15H, m), 6.59 (1H, d,
J = 2.0 Hz), 6.46, (1H, dd, J = 5.7, 1.5 Hz). 13C 161.2,
151.5, 148.6, 151.5, 151.4, 132.6, 132.5, 132.4, 129.6, 129.1,
128.9, 128.6, 117.8, 117.7, 117.6, 117.5.
19. Synthesis of 7. A solution of 6 (0.339 g, 0.87 mmol) and 5
(0.187 g, 1.0 mmol) in 10 ml fresh anhydrous benzene was
heated at 60 ꢂC for 1 h. The solution was filtered through a
short silica gel column using argon pressure and the
column was additionally washed with fresh anhydrous p-
xylene. The benzene was evaporated and the mixture in
residual xylene was heated at 180 ꢂC for 12 h. The xylene
was removed by evaporation and the residue was frac-
tionated by low pressure chromatography on silica gel
with ethyl acetate in hexane using a gradient from 0% to
100% in 30 min. The crude product was purified by second
low pressure chromatography on silica gel using diethyl
ether in chloroform with a gradient from 0% to 100% in
60 min as the eluant. NMR (CDCl3) 1H d 8.19 (1H, d,
J = 6.0 Hz), 8.01 (1H, s), 7.85 (2H, dd,) 7.44–7.70 (60H,
m), 7.31 (1H, dd, J = 9.1, 2.8 Hz), 7.01 (1H, d, J = 2.7 Hz),
3.92 (3H, s), 2.46 (3H, s).
20. Synthesis of 8. Compound 7 (0.0093 g, 0.03 mmol) was
stirred with 3,30-diamino-N-methyldipropylamine (1 ml,
6.2 mmol) under argon at 150 ꢂC for 46 h. The mixture
was concentrated and purified by flash chromatography
on basic alumina with methanol in chloroform with a 0–
5% gradient in 80 min. The mixture was dissolved in 1:1
acetonitrile/water and purified by preparative HPLC on a
C18 column using a water/acetonitrile solvent system. The
purity of the compound was established by HPLC–ESI-
MS. Compounds 9 and 10 were prepared in an analogous
manner.
21. Synthesis of 11 and 13. A solution of 7 (0.0062 g,
0.02 mmol) in anhydrous dichloromethane was treated
with methyl trifluoromethane sulfonate (2.8 ll,
0.025 mmol) and stirred at room temperature for 24 h,
concentrated, and dissolved directly in an excess of N-(3-
aminopropyl)-1,3-propanediamine and stirred for 1 h at
room temperature. The product was purified by reverse
phase preparative HPLC using a C18 column and water/
acetonitrile solvent system. The side product, 13, was
isolated from the HPLC separation. The purity of the
compounds were established by HPLC–ESI-MS. A similar
procedure was used for the preparation of compound 12.
22. Synthesis of 14, 15, and 16. A solution of 7 (0.0055 g,
0.019 mmol) in anhydrous dichloromethane was treated
with ethyl trifluoromethanesulfonate (5 ll, 0.039 mmol)
and stirred for 160 h at 30 ꢂC. Additional 5 ll portions of
ethyl trifluoromethanesulfonate and 1 ml portions anhy-
drous dichloromethane were added every 48 h. The product
was concentrated and used directly in a reaction with N-(3-
aminopropyl)-1,3-propanediamine (3.2 ll, 0.022 mmol).
The mixture dissolved in 1 ml THF was stirred for 1 h at
room temperature. The products were separated and
purified by reverse phase preparative HPLC using a C18
column and water/acetonitrile solvent system. Compounds
16 and 15 were side products of the reaction. The purity of
the compounds was determined by HPLC–ESI-MS.