Synthesis, biological activity and comparative analysis of DNA binding affinities and human DNA topoisomerase I inhibitory activities of novel 12-alkoxy-benzo[c]phenanthridinium salts

Article abstract of DOI:10.1016/S0960-894X(01)00520-0

New antitumor 12-alkoxy-benzo[c]phenanthridinium derivatives were obtained in high yields through multistep syntheses. Analysis of DNA binding and human DNA topoisomerase I inhibitory activities demonstrates that new compounds, combining 2, 6, and 12 substitutions, interact strongly with DNA and exhibit important topoisomerase I inhibition. The cytotoxicities against solid tumor cell lines are also determined and compared with those for fagaronine and ethoxidine.

Full text of DOI:10.1016/S0960-894X(01)00520-0

Bioorganic & Medicinal Chemistry Letters 11 (2001) 2643–2646
Synthesis, Biological Activity and Comparative Analysis of DNA
Binding Affinities and Human DNA Topoisomerase I Inhibitory
Activities of Novel 12-Alkoxy-benzo[c]phenanthridinium Salts
Michael A. Lynch,a Olivier Duval,a,* Alyona Sukhanova,c Jerome Devy,c
Simon P. MacKay,b Roger D. Waighb and Igor Nabievc
aLaboratoire SONAS, UFR de Pharmacie, 16 boulevard Daviers, F-49100 Angers, France
bDepartment of Pharmaceutical Sciences, Strathclyde Institute for Biomedical Sciences, 27 Taylor Street, Glasgow G4 ONR, UK
c
´ ´ ´
EA 3306 ‘Interactions Moleculaires et Cellulaires en Cancerologie’, IFR no. 53 ‘Biomolecules’, UFR de Pharmacie,
´
51 rue Cognacq Jay, 51100 Reims, France
Received 28 May 2001; revised 4 July 2001; accepted 25 July 2001
Abstract—New antitumor 12-alkoxy-benzo[c]phenanthridinium derivatives were obtained in high yields through multistep synth-
eses. Analysis of DNA binding and human DNA topoisomerase Iinhibitory activities demonstrates that new compounds, com-
bining 2, 6, and 12 substitutions, interact strongly with DNA and exhibit important topoisomerase Iinhibition. The cytotoxicities
against solid tumor cell lines are also determined and compared with those for fagaronine and ethoxidine. # 2001 Elsevier Science
Ltd. All rights reserved.
Benzo[c]phenanthridines (BZPs) are naturally occurring
alkaloids which have demonstrated numerous biological
activities.1 Among these structures, nitidine 1 and
fagaronine 2 have solicited much interest as anticancer
drugs. Preclinical studies were conducted at the US
National Cancer Institute, but these compounds were
not considered to be practical anticancer drugs.2 Many
synthetic approaches to benzo[c]phenanthridinic alka-
loids leading to natural products have been reported,
but relatively few studies of synthetic analogues have
been described.3 BZPs have also demonstrated binding
affinities to DNA, have been presumed to be inter-
biological results in different in vitro antileukemic
activity assays. Of these deÀrivatives, ethoxidine 3
(R12=CH2CH3, X=CH3SO3
) demonstrated high
cytotoxic activity on the P388 and K562 cancer cell lines
and also had the ability to induce differentiation in
K562 human cells.
Compounds 2 and 3 (R12=CH2CH3, X=CH3SO3
)
À
were proven to be DNA intercalators4c,6 with, for the
latter, the 12-ethoxy appendage penetrating into the
minor groove.6 Top1 inhibitors are grouped into two
classes: top1 poisons and top1 suppressors. The best
characterized top1 poisons are camptothecins (CPTs)
which kill cells by trapping cleavage complexes between
top1 and DNA.7 The typical DNA binders, including
DNA intercalators and minor groove binders can sup-
press top1-linked DNA breaks due to prevention of
enzyme’s access to DNA and/or distortion of the DNA
structure. The selectivity of the existing top1 sup-
pressors remains questionable.7 Subsequent investiga-
tions showed that 2 induces DNA cleavage by top1 in a
typical CPT-dependent manner4c and that the 2-OH
group greatly contributed to the biological activity as a
calators and have been shown to be inhibitors of DNA
4
topoisomerases I(top1) and I(top2),
the well known
targets for clinically important and emerging antitumor
drugs.
A recent study reported the impact of 12-alkoxy modifica-
tion on the in vitro antileukemic activity of N-methyl-12-
alkoxy-benzo[c]phenanthridinium salts.5 Among the iso-
lated molecules, the ones which had shorter aliphatic
and non-functionalized 12-alkoxy-chains gave the best
hydrogen bond acceptor.6b On another hand,
3
(R12=CH2CH3, X=CH3SO3À) suppresses both top1-
specific and CPT-dependent cleavage, with AT-sequence
*Corresponding author. Tel.: +33-241-22-66-73; fax: +33-241-22-67-33;
e-mail: olivier.duval@univ-angers.fr
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00520-0

2644
M. A. Lynch et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2643–2646
specificity of effect.6b Thus,
3
(R12=CH2CH3,
DNA binding affinities and top1 inhibitory activities of
these BZPs. The contribution of 6-position substitution
of new 2,3,7,8-tetraoxygenated-BZPs was recently eval-
uated on cytotoxicities and reduction-resistance proper-
ties.8 We have thus synthesized 6-methyl-ethoxidine to
measure the impact of the 6-substitution on DNA
binding and top1 inhibition.
X=CH3SO3À) and 2 demonstrate different mechanisms
of action towards top1, respectively, suppression and
poisoning of the enzyme. Moreover, 3 (R12=CH2CH3,
X=CH3SO3À) inhibites DNA relaxation at a concen-
tration 10-fold lower than fagaronine.6a These results
make 3 (R12=CH2CH3, X=CH3SO3À) potentially
attractive for further development of top1 inhibitors.
One aim of the present work is to take into account the
12-O-ethyl and 2-hydroxy substitutions which play
major roles in the biological activities. We thus report
the synthesis of new 12-ethoxy-fagaronine derivatives.
We also report their cytotoxicities against human solid
tumor cell lines and make a comparative analysis of
Chemistry
Harsh acidic conditions were required for the formation
of the B and C rings during the synthesis of 12-alkoxy
BZPs.5 In a previous study, 2 was prepared through de-
ethylation of a phenolic ether protecting group using
concentrated sulfuric acid.9 The major drawbacks in
this type of deprotection are controlling the reaction
time and the need for laborious purification steps to
separate related polar by-products. To avoid this, we
chose the methanesulfonate phenolic protection of
vanillin in the early steps of our synthesis, which was
easily carried out using methanesulfonyl chloride in
pyridine to give 510 (Scheme 1). The amino nitrile 6 was
obtained under Strecker reaction conditions in 86%
yield. A nearly quantitative hydrogen fluoride cycliza-
tion of 6 at room temperature, led to the ketone 7. The
Reformatski reaction, which introduced the two-carbon
synthon required for the ensuing cyclisation, yielded the
amine 8. Sulfuric acidic cyclisation led to the BZP 9,
which was thereafter O-ethylated in the 12-position
using diethylsulfate and sodium carbonate in dried ace-
tone giving 10, and then N-methylated in the presence
of methyl methanesulfonate at 180 ꢀC to yield the
12-ethoxy-2-methanesulfonoxy-3,8,9-trimethoxybenzo[c]-
phenanthridinium salt 11. The target compound, 12-
ethoxyfagaronine 14, was obtained following the reduc-
tion of the iminium bond in 11 with sodium borohy-
dride in methanol to give the tertiary amine 12, sulfonyl
ester hydrolysis of 12 using 5 N potassium hydroxide
to yield the 2-hydroxyBZP 13 and oxidation of 13
with iodine in ethanol to 14. The overall yield of this
multistep synthesis was 20%. In a similar second syn-
thesis, the 6-methyl group was introduced into 3
(R12=CH2CH3, X=CH3SO3À) using methyl magne-
sium bromide in THF. Subsequent oxidation gave 6-
methyl-ethoxidine iodide in 75% yield (Scheme 2).
Scheme 1. Synthesis of the novel 12-ethoxy-benzo[c]phenanthidinium
salts. Reagents: (a) KCN, H2O/C2H5OH, rt; (b) HF, brine 15 ꢀC; (c)
BrCH2CO2Et, Zn, dioxane/DMM; (d) H2SO4; (e) K2CO3, acetone,
(C2H5O)2SO2; (f) CH3SO3CH3, 180 ꢀC; (g) NaBH4, CH3OH; (h) KOH
(5 N), CH3OH; (i) I2, C2H5OH.
Scheme 2. Synthesis of the 6-methyl-12-ethoxybenzo[c]phenanthidinium
salt. Reagents: (a) CH3MgBr, THF; (b) I2, C2H5OH.

M. A. Lynch et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2643–2646
2645
.
Affinities of DNA Binding
of intercalation: binding with poly(dA-dT) poly(dA-dT)
is found to be 25-fold higher than with poly(dG-
.
dC) poly(dG-dC). Compounds 11 and 14 showed
Stock solutions (10 mM) of 2 and its derivatives were
prepared in DMSO and diluted to the desired con-
centration immediately before use. Concentrations of
calf thymus (CT) DNA and the double-stranded
poly(dA-dT)poly(dA-dT) and poly(dG-dC)poly(dG-
dC) polymers were determined by using molar extinc-
tion coefficients of 13,200, 13,900 and 13,200 MÀ1 cmÀ1
respectively. UV–vis spectra were recorded with a
JASCO V-530 UV–vis scanning spectrophotometer.
DNA binding constants were calculated as described.6b
slightly less affinity for CT DNA compared to 2 and
were similar to 3 (R12=CH2CH3, X=CH3SO3À). As far
as sequence specificity is concerned, 11 is found to be
similar to 2 (no sequence specificity), whereas 14 is more
similar to 3 (R12=CH2CH3, X=CH3SO3À), demon-
strating pronounced AT specificity. Both compounds
,
.
demonstrate lower affinity to poly(dA-dT) poly(dA-dT)
relative to 3 (R12=CH2CH3, X=CH3SO3À) (50- and
8.3-fold less, respectively). On the other hand, the
sequence-specificity of 6-methylethoxidine 16 was found
to be comparable to that for 3 (R12=CH2CH3,
X=CH3SO3À), but with 2- and 3-fold less affinities
Activities of top1-DNA Relaxation Inhibition
.
toward poly(dG-dC) poly(dG-dC) and poly(dA-
dT)Âpoly(dA-dT), respectively.
Recombinant human top1 was purified from insect cells
using a two-step procedure as described.11,12 Specific
activity of top1 used was 1.8Â106 units/mg, where one
unit of activity is an amount of enzyme yielding 100%
of relaxation of 300 ng of supercoiled pGEM7Z(f+)
plasmid DNA in 30 min at 37 ꢀC. The DNA relaxation
assay was carried out as described.12 Inhibition of the
top1-mediated plasmid DNA relaxation was tested
when the aliquots of drug solutions were added to the
reaction mixture.
The inhibition of top1-mediated plasmid DNA relaxa-
tion by the same derivatives was also determined. Each
compound was found to be active at nearly the same
minimal inhibitory concentration (MIC: 0.75 mM), cor-
responding to the lowest concentration of drug able to
inhibit top1-mediated DNA relaxation (Fig. 1).
BZPs are potent inhibitors of the growth of leukemia
tumor cells14 and fagaronine and ethoxidine have both
demonstrated this type of activity.6 In this work, we
have evaluated the cytotoxicity of our molecules against
various solid cancer cell lines (Table 2). Compounds 3
(R12=CH2CH3, X=CH3SO3À) and 2 were found to be
cytotoxic towards all cell lines with better activities for
the former except in the case of the CaOv3 ovarian cell
line. Compounds 11 and 14 exhibited weaker activities
in all in vitro tests. Finally, poor activities were found
with 16, in which 6-position substitution is known to
reduce cytotoxicity.
Cytotoxicities
HT-29, A-549, MCF-7 and CaOv3 human solid tumor
cell lines were examined in cell growth inhibition assays
using the microtiter tetrazolium test.13 Plating densities
were chosen such that exponential growth was main-
tened throughout the subsequent 4 days cultured per-
iod. Drugs were added 24 h after plating in nine
concentrations. After cells had grown for 72 h in the
presence of drug, 20 mL of 2.5 mg/mL solution of
dimethylthiazodiphenyltetrazolium bromide (Sigma) in
PBS was added. After 3 h of incubation, the liquid
content of each well was aspirated and the purple for-
mazan precipitate dissolved in 200 mL dimethylsulf-
oxide. Absorbances were measured at 570 nm (BioRad-
750 microplate reader). IC50 values were determined
from dose/inhibition curves.
In conclusion, the synthetic approach developed here to
prepare new 2,12-functionalized benzo[c]phenanthridine
derivatives in association with their biochemical and
biological evaluation is providing interesting clues in the
step-by-step comprehension of the mechanism of action
of this class of cytotoxic agents.
In terms of affinity for DNA, our results demonstrate
the importance of the 2-substitution in this series.
Modifications in 11 and 14 slightly decrease affinities to
DNA: the introduction of a bulky polar substituent
in the 2-position in 11 resulted in no selectivity effect
related to 2 behavior towards complexes. On the other
We initially evaluated the binding constants of our
.
BZPs within calf thymus DNA, poly(dA-dT) poly(dA-
.
dT) and poly(dG-dC) poly(dG-dC) complexes (Table 1).
Data show that 2 has the highest CT DNA binding
constants but no base preference of intercalation. On
the other hand, 3 (R12=CH2CH3, X=CH3SO3À) has a
lower CT DNA affinity, but exhibits clear AT-specificity
Table 2. Comparative cytotoxicities of BZP derivatives
IC50, mM
Table 1. DNA binding affinities of fagaronine and BZP derivatives
Compound
A-549
CaOv3
HT-29
MCF-7
Compound
K, CT DNA,
MÀ1
K, poly(GC),
MÀ1
K, poly(AT),
M À1
2
3
14
11
16
0.1
0.08
0.5
0.5
10a
0.5
0.5a
na
na
10a
9
2
5a
1a
9a
4
2
na
na
1a
2
3
14
11
16
106
1.1Â106
106
2.5Â105
2.5Â105
2.5Â105
7Â105
105
2.5Â106
3Â105
5Â104
7Â105
6Â104
5.3Â104
5Â104
na, not active.
aOnly IC30 attainable.

2646
M. A. Lynch et al. / Bioorg. Med. Chem. Lett. 11 (2001) 2643–2646
being synthesized to establish the exact nature of sub-
stituents required in this structure for optimal activity.
Site-specific DNA cleavage studies using top1 will also
help to determine the mode of action of each molecule
and will further explore the antitumor potential of this
class of compound.
Acknowledgements
This research was supported by grants from ARC, from
Region Champagne-Ardenne (I.N., A.S.) and from
departement du Maine et Loire (M.A.L., O.D.).
References and Notes
1. Simanek, V. In The Alkaloids; Brossi, A., Ed.; Academic:
New York, 1985; p 26.
2. (a) Phillips, S. D.; Castle, R. N. J. Heterocycl. Chem. 1981,
18, 223. (b) Suffness, M.; Cordell, G. A. In The Alkaloids;
Brossi, A.; Academic: New York, 1985; Vol. 25, p 178.
3. MacKay, S. P.; Meth-Cohn, O.; Waigh, R. D. Adv.
Heterocycl. Chem. 1997, 67, 345.
4. (a) Fang, S. D.; Wang, L.-K.; Hecht, S. M. J. Org. Chem.
1993, 58, 5025. (b) Wang, L. K.; Johnson, R. K.; Hecht, S. M.
Chem. Res. Toxicol. 1993, 6, 813. (c) Larsen, A. K.; Grondard,
L.; Couprie, J.; Desoize, B.; Jardillier, J. C.; Riou, J. F. Bio-
chem. Pharm. 1993, 46, 1403.
5. MacKay, S. P.; Comoe, L.; Desoize, B.; Duval, O.; Jardil-
lier, J. C.; Waigh, R. D. Anti-Cancer Drug Des. 1998, 13, 797.
6. (a) Ianoul, A.; Fleury, F.; Duval, O.; Waigh, R.; Jardillier,
J. C.; Alix, A. J. P.; Nabiev, I. J. Phys. Chem. 1999, 103, 2008.
(b) Fleury, F.; Sukhanova, A.; Ianoul, A.; Devy, J.; Kudelina,
I.; Duval, O.; Alix, A. J. P.; Jardillier, J.-C.; Nabiev, I. J. Biol.
Chem. 2000, 275, 3501.
7. Bailly, C. Curr. Med. Chem. 2000, 7, 39 and references
herein.
8. Nakanishi, T.; Masuda, A.; Suwa, M.; Akiyama, Y.; Hos-
hino-Abe, N.; Suzuki, M. Bioorg. Med. Chem. Lett. 2000, 10,
2321.
Figure 1. Inhibition of the topoisomerase I-mediated plasmid DNA
relaxation by new BZP derivatives. DNA plasmid pGEM-7Zf(+)
(lane 1) incubated with DNA topoisomerase I(lane 2). Topoisomerase
I-mediated plasmid DNA relaxation observed in the presence of
fagaronine 1 (panel A), derivative 14 (panel B), derivative 11 (panel
C), derivative 16 (panel D) at 12 mM (lane 3), 3 mM (lane 4), 0.75 mM
(lane 5), 0.1875 mM (lane 6), 0.0469 mM (lane 7).
hand, selectivity with AT is observed with 14. Hence, a
smaller group in this position induces a different effect,
more similar to that for
3
(R12=CH2CH3,
X=CH3SO3À). The results from both types of mod-
ifications indicate that the nature of the 2-position sub-
stituent is of great significance for DNA interaction.
Moreover, a structural combination of the stronger
DNA binder 3 (R12=CH2CH3, X=CH3SO3À) with the
weaker DNA binder 2 leads to decrease of DNA-bind-
ing potency, giving additional proof that each molecule
interacts with nucleic acids in a particular manner.
9. Lynch, M. A.; Duval, O.; Pochet, P.; Waigh, R. D. Bull.
Soc. Chim. Fr. 1994, 131, 718.
10. Stermitz, F. R.; Gillespie, J. P.; Amoros, L. G.; Romero,
R.; Stermitz, T. A. J. Med. Chem. 1975, 18, 708.
11. Fleury, F.; Ianoul, A.; Kryukov, E.; Sukhanova, A.;
Kudelina, I.; Wynne-Jones, A.; Bronstein, I. B.; Maizieres, M.;
Berjot, M.; Dodson, G. G.; Wilkinson, A. J.; Holden, J. A.;
Feofanov, A. V.; Alix, A. J. P.; Jardillier, J. C.; Nabiev, I.
Biochemistry 1998, 37, 14630.
12. Bronstein, I.; Wynne-Jones, A.; Sukhanova, A.; Fleury,
F.; Ianoul, A.; Holden, J.; Alix, A. J. P.; Dodson, G.; Jardil-
lier, J. C.; Nabiev, I.; Wilkinson, A. J. Anticancer Res. 1999,
19, 317.
The good DNA binding and top1 inhibition effects, in
conjunction with the poor in vitro cytotoxic activities,
may indicate that 11, 14 and 16 experience major pro-
blems in reaching their cellular targets. Further studies
are in progress aimed at understanding the mechanisms
involved in penetration of the cytotoxic agents of this
class through cellular membranes. New BZPs are also
13. Mosmann, T. J. Immunol. Meth. 1983, 65, 55.
14. Olugbade, T. A.; Waigh, R. D. Pharm. Sci. 1996, 2, 259.