Photo-inducible cytotoxic and clastogenic activities of 3,6-di-substituted acridines obtained by acylation of proflavine

Article abstract of DOI:10.1016/j.ejmech.2009.01.010

The cytotoxicity and photo-enhanced cytotoxicity of a series of 18 3,6-di-substituted acridines were evaluated on both tumour CHO cells and human normal keratinocytes, and compared to their corresponding clastogenicity as assessed by the micronucleus assa

Full text of DOI:10.1016/j.ejmech.2009.01.010

European Journal of Medicinal Chemistry 44 (2009) 2459–2467
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Photo-inducible cytotoxic and clastogenic activities of 3,6-di-substituted acridines
obtained by acylation of proflavine
,
b
a
c
Yohann Benchabane b, Carole Di Giorgio a , Gerard Boyer , Anne-Sophie Sabatier , Diane Allegro ,
*
´
c
a
´
Vincent Peyrot , Michel De Meo
a Laboratoire de Bioge´notoxicologie et Mutagene`se Environnementale (EA 1784, FR 3098 – ECCOREV), Universite´ Aix-Marseille, Faculte´ de Pharmacie, 27 Bd Jean Moulin, 13385
Marseille, Cedex 05, France
b Laboratoire HIT, case 552, (iSm2 – UMR 6263) – Universite´ Paul Ce´zanne, Faculte´ St Je´roˆme, 13397 Marseille, Cedex 20, France
c CRO2 INSERM-U91 – Universite´ Aix-Marseille, Faculte´ de Pharmacie, 27 Bd Jean Moulin, 13385 Marseille, Cedex 05, France
a r t i c l e i n f o
a b s t r a c t
Article history:
The cytotoxicity and photo-enhanced cytotoxicity of a series of 18 3,6-di-substituted acridines were
evaluated on both tumour CHO cells and human normal keratinocytes, and compared to their corre-
sponding clastogenicity as assessed by the micronucleus assay.
Received 7 July 2008
Received in revised form
9 December 2008
Compounds 2f tert-butyl N-[(6-tert-butoxycarbonylamino)acridin-3-yl]carbamate and 2d N-[6-(piv-
alamino)acridin-3-yl]pivalamide displayed a specific cytotoxicity on CHO cells. These results suggested
that the two derivatives could be considered as interesting candidates for anticancer chemotherapy and
hypothesized that the presence of 1,1-dimethylethyl substituents was responsible for a strong non-
clastogenic cytotoxicity. Compounds 2b and 2c, on the contrary, displayed a strong clastogenicity. They
indicated that the presence of nonbranched aliphatic chains on positions 3 and 6 of the acridine rings
tended to induce a significant clastogenic effect. Finally, they established that most of the acridine
compounds could be photo-activated by UVA-visible rays and focussed on the significant role of light
irradiation on their biological properties.
Accepted 9 January 2009
Available online 20 January 2009
Keywords:
Proflavine derivatives
Phototoxicity
Photogenotoxicity
Micronucleus assay
Ó 2009 Elsevier Masson SAS. All rights reserved.
1. Introduction
leading to mutagenicity, reproductive toxicity and carcinogenicity,
and has been shown to result in an increased risk of secondary
Since their first discovery in the 1880s, acridines have
demonstrated a broad spectrum of pharmacological properties
[1]. First employed as antibacterial agents during the beginning
of the twentieth century [2], they have rapidly revealed inter-
esting antiproliferative activities against both protozoa [3] and
tumour cells [4]. As a consequence, they have been extensively
used in antiparasitic [3] and antitumoral [4] chemotherapy and
a wide range of new acridine-derived analogues have been
continuously synthesized and assessed for their pharmacological
properties.
tumour in patients overcoming acridine-based therapies [8,9].
The other important physicochemical property of acridines is
their capacity to absorb light energy. This property allows the use
of acridine chromophores in sensitive spectroscopic techniques;
however it also results in a different biological behaviour of the
molecules when they are exposed to light radiations as compared
to their activity in the dark. As a consequence, several elements of
the acridine series have exhibited photo-inducible pharmacolog-
ical properties, mainly photo-toxicity, photo-bactericidal activity
[10] and photo-enhanced mutagenicity [11,12].
The antitumoral and anti-infectious activities of acridines are
mainly related to their capacity to reversibly bind with DNA [5]. Due
to their planar polycyclic structure, they have been shown to
intercalate between DNA double-strands, to interfere with DNA
regulatory enzymes such as topoisomerase I and II and to disrupt
DNA functions in cells [6]. This DNA-intercalating activity, first
demonstrated with proflavine [7], has revealed genotoxic properties
In a preceding study, we have assessed the antileishmanial
activity of eight 3,6-di-substituted analogues of proflavine and we
have demonstrated that some of these compounds exerted an
interesting antiparasitic activity [13]. However, we have also
observed that several compounds exhibited a concomitant signif-
icant DNA-binding activity. In the present work, new acridine
derivatives were produced by a new synthesis protocol, and
assessed for their photo-inducible cytotoxicity against both trans-
formed and normal cells as compared to their capacity to induce
chromosomal mutations and rearrangements.
* Corresponding author.
E-mail address: Carole.Digiorgio@Pharmacie.univ-mrs.fr (C. Di Giorgio).
0223-5234/$ – see front matter Ó 2009 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2009.01.010

2460
Y. Benchabane et al. / European Journal of Medicinal Chemistry 44 (2009) 2459–2467
9
8
5
1
4
O
2
3
7
i) or ii)
+
R1
R2
HN
N
NH
H2N
N
NH2
6
1
R2 = Cl, OCOR1
R1
O
O
R1
2a-f
i) Pyridine, Et3N, 1h, 80°C
ii) Acetone, K2CO3, 15h, rt
R
Comp.
2a
Yields
60%
65%
32%
74%
70%
75%
MW
293
321
349
378
325
410
Molecular formula
C17H15N3O2
Mp °C
> 260
> 260
216
158
> 260
165
1
Me
Et
Pr
tBu
OMe
OtBu
C19H19N3O2
C21H23N3O2
2b
2c
C23 27
C17 15
C23 27
H
H
H
N3O2
N3O4
N3O4
2d
2e
2f
Compounds 2a-c: see ref. 13
Fig. 1. Synthesis of compounds 2a–f.
2. Chemistry
Last, naphthoyl and furoyl chlorides were improved as reagents
to prepare polyheteroaromatic receptors with proflavine leading to
compounds 4a and 4b with good yields (86 and 70% respectively),
as reported in Fig. 3.
The study of the spectral absorption properties showed that the
maximal absorption wavelength observed for all the compounds
averaged 330 nm. At this absorption wavelength, all the
compounds emitted fluorescence at 350 and 400 nm. An example
of absorption and fluorescence graphs obtained with compound 2c
is reported in Fig. 4.
The preparation of symmetrical acetamides was performed
using neutral proflavine 1, which was acylated alternatively with
different anhydrides and benzoyl chlorides to yield the corre-
sponding amides as reported in Figs. 1–3. Acetone was sometimes
preferred to the more toxic pyridine and allowed more favourable
conditions of reaction at room temperature. However, the purifi-
cation of these products was very difficult owing to their low
solubility in organic solvents, and inorganic base K2CO3 preferred to
NEt3 to improve the technique.
First, proflavine
1 was allowed to react with different
3. Pharmacology
anhydrides and alkoyl chlorides to lead to the corresponding
acetamides (2a–f), low to good yields were obtained (32–91%)
as reported in Fig. 1. As shown in Fig. 2, different benzoyl
chlorides were employed and led to diamido acridines 3a–i with
yields reaching 47–96%; and the 4-nitro-N-[6-(4-nitro-
benzoylamino)acridin-3-yl]benzamide 3e was reduced according
to a modified procedure, using palladium under hydrogen atmo-
sphere (3 bar) in EtOH [14]. The symmetrical bis amine 3j was
recovered with 26% yield after purification on column
chromatography.
The toxicity and phototoxicity of acridine derivatives were
evaluated against both transformed Chinese Hamster Ovary Cells
(CHO, ATCC) and primary cultures of human keratinocytes. Their
corresponding photo-inducible clastogenicity was studied by
measuring their capacity to induce chromosome mutations and/
or misaggregations, as assessed by the micronucleus assay [15].
Structure–activity relationships were performed in order to
better analyze the mechanisms by which acridine derivatives
exerted their biological activities. Predictive models were used to
O
9
8
5
1
4
R3
R2
2
3
NH
7
Cl
i) or ii)
R3
+
NH2
N
6
HN
2'
H2N
N
1
6'
5'
R3
R2
R1
O
O
3a-i
R2
4'
i) Pyridine, Et3N, 1h, 80°C
ii) Acetone, K2CO3, 15h, rt
3'
R1
H2, Pd/C
R1
3j
3e
EtOH, 30h
Comp
R
R
R
3
Yields
MW
Molecular formula
Mp°C
1
2
a,b
C27H19N3O2
3a a
H
H
H
OMe
Cl
H
H
79%
85%
84%
47%
57%
417
477
553
689
507
C29H23N3O4
3b
3c
3d
3e
> 260
> 260
190
C27H15N3O2Cl4
H
Cl
CF3
H
H
CF3
H
C31H15N3O2F12
NO2
> 260
C27H17N O
C29H23N53O6
c
3f a
H
H
H
H
Me
Cl
H
H
H
H
82%
90%
96%
23%
445
486
453
467
> 260
> 260
> 260
> 260
2
C27H17N3O2Cl2
C27H17N3O2F2
C29H17N5O2
3g
3ha
3i
F
CN
d
H
NH2
H
26%
447
C27H21N5O2
238
3j
a See reference 13, b 3a-e obtained with acetone, c 3f-i obtained with pyridine, d 3j obtained with EtOH, H2, Pd/C
Fig. 2. Synthesis of compounds 3a–j.

Y. Benchabane et al. / European Journal of Medicinal Chemistry 44 (2009) 2459–2467
2461
Acetone, K2CO3
O
15h, rt
+
HN
N
NH
Ar
Mp °C
Cl
Ar
H2N
NH2
N
4a-b
1
Ar
O
O
Comp
4a
Ar
Yields
MW
Molecular formula
35H23N3O2
86%
517
397
C
>260
174
4b
70%
C23H15N3O4
O
Fig. 3. Synthesis of compounds 4a and 4b.
calculate the lipohilicity (log P), the solubility (log S) and the
electrophilicity of the compounds [16]. Then, the experimental
data were compared with the antineoplastic, mutagenic and
photosensitizer activities predicted by the PASS database [17],
according to their chemical similarity with known active
molecules.
Finally, in order to better understand the mechanisms by
which acridine derivatives exerted their clastogenicity and
photo-clastogenicity against CHO cells, the role of oxidative
events on the clastogenic effects of compound 2c, the most
photo-inducible clastogenic compound of the series, was inves-
tigated in the presence of the protective antioxidant mannitol
[18].
0.5 < Pa < 0.7 suggested that the compound was likely to reveal
activity in experiments, while Pa < 0.5 implied that the compound
was unlikely to reveal activity in experiments. All the predictive
probabilities calculated for the antineoplastic activity were higher
than 0.5: they implied that the 3,6-di-substituted proflavine
derivatives were likely to exert cytotoxic activity against tumour
cells. Among these compounds, four derivatives (3h, 3b, 3a and
3e) showed Pa higher than 0.7. Concerning the mutagenic or the
photosensitising activities on the contrary, almost all the esti-
mated Pa were lower than 0.3. They indicated that, according to
the predictive model, the 3,6-di-substitued acridines were
unlikely to exert mutagenicity or photo-inducible biologic
activities.
The cytotoxicity of proflavine derivatives was primarily
assessed on transformed Chinese Hamster Ovary cells. In the
dark, ten derivatives exerted a significant cytotoxic activity
against CHO cells. However, only 2d displayed a strong cyto-
4. Results and discussion
Complete data concerning the physicochemical properties and
the photo-inducible biological activities of 3,6-di-substituted acri-
dines are summarized in Table 1.
toxicity with IC50 averaging 0.11
m
M, while 4b and 2f appeared
M respectively). These
less active (IC50 ¼ 2.5 M and IC50 ¼ 6.1
m
m
results suggested the important cytotoxic role of the 1,1-dime-
thylethyl group on the border of the molecules. After UV irra-
diation on the contrary, almost all the compounds exhibited
a significant cytotoxic activity. This photo-enhanced toxicity
could be explained by the capacity of the compounds to mainly
absorb light rays at 330 nm. Compounds 3a and 4a displayed the
highest photo-inducible properties, since UV light induced
respectively a 90 and 130 fold increase of their cytotoxicity.
Nevertheless, 2d and 2f remained the most cytotoxic derivatives
The lipophilicity of each chemical compound was estimated by
the prediction of the n-octanol/water partition coefficient log P,
defined as the ratio of concentration in an immiscible solvent such
as n-octanol to the concentration in the aqueous phase. Results
showed that all partition coefficients reported in Table 1 were
greater than 1, indicating that proflavine derivatives exhibited
lipophilic properties. Nevertheless, the lipophilicity varied accord-
ing to the nature of the 3,6-substituted groups: compounds bearing
aliphatic substituents displayed lower lipophilicity than those
bearing aryl rings. As expected, all the acridine derivatives exhibi-
ted weak solubility in water. For this reason, the highest tested
(IC50 ¼ 0.1
m
M), with cytotoxic activities that did not depend on
concentration was 50
mM.
Electrophilicity was expressed as the probability Pe for each
compound to initiate electrophilic reactions: acridine
analogues were considered electrophilic when Pe > 0.5. Results
reported in Table 1 showed that Pe greatly varied according to
the chemical structure of the compounds. 3d and 3j displayed
the highest predictive values (0.941 and 0.966 respectively), 2a,
2e, 3f and 3i were characterized by a moderated electrophi-
licity, while the other compounds displayed low electrophilic
properties.
Predictive values for biological activities were obtained by
comparing the chemical structure of each compound with struc-
tures or sub-structures of more than 30,000 well-known biolog-
ically active drugs. They were expressed as the probability Pa of
each compound to be active and illustrated its degree of similarity
with well-known antineoplastic compounds, mutagenic mole-
cules, or photosentitizers. Pa > 0.7 indicated that the correspond-
ing compound was very likely to reveal activity in experiments,
3,5e+7
0,8
0,6
0,4
0,2
0,0
Absorption spectrum
3,0e+7
2,5e+7
2,0e+7
1,5e+7
1,0e+7
5,0e+6
0,0
Fluorescence spectrum
300
350
400
450
Wavelength (nm)
500
550
600
Fig. 4. Absorption and fluorescence spectra of compound 2c.

2462
Y. Benchabane et al. / European Journal of Medicinal Chemistry 44 (2009) 2459–2467
Table 1
Photo-inducible cytotoxicity and clastogenicity of proflavine derivatives.
Compounds
Physicochemical properties
Predictive values
Experimental data
Cytotoxicity
Biological effects, Pa
Clastogenic activity
MCC ( M)
m
CHO cells IC50-CHO
Keratinocytes IC50-
M)
(
mM)
(m
kera
log P
log S
ꢀ3.5
Pe
AntiN
0.789
Mut
Photo
0.305
ꢀ light
1.1
þ light
0.52
ꢀ light
>50
þ light
>50
ꢀ light
0.005
þ light
0.004
1
1.7
0.958
0.481
2a
2b
2c
2d
2e
2f
1.9
2.9
3.9
4.5
2.5
5.1
ꢀ4.3
4.7
ꢀ4.9
ꢀ5.3
ꢀ4.7
ꢀ5.8
0.809
0.409
0.208
0.137
0.843
0.162
0.588
0.600
0.598
0.543
0.642
0.578
0.280
0.160
0.160
0.152
0.218
0.194
0.266
0.260
0.260
0.254
0.273
0.280
>50
>50
38.9
0.11
15.8
6.1
37.2
2.9
1.34
0.1
3.8
0.12
>50
>50
>50
3.28
>50
>50
41.2
2.4
0.57
4.69
8.1
11.01
1.15
1.01
NS
0.47
NS
0.24
0.03
0.007
NS
0.21
NS
0.14
3a
3b
3c
3d
3e
3f
3g
3h
3i
4.9
5.1
7.3
8.1
4.8
5.7
6.1
5.1
4.9
3.2
ꢀ6.1
ꢀ5.9
ꢀ7.9
ꢀ7.9
ꢀ6.3
ꢀ6.4
ꢀ6.7
ꢀ6.3
ꢀ6.3
ꢀ5.8
0.567
0.393
0.218
0.941
0.281
0.624
0.396
0.384
0.743
0.966
0.716
0.744
0.635
0.608
0.709
0.613
0.658
0.783
0.587
0.649
0.189
0.179
0.196
0.147
0.400
0.211
0.196
0.140
0.146
0.320
0.281
0.260
0.242
0.270
0.236
0.277
0.256
0.240
0.257
0.262
23.8
45.2
>50
>50
40.2
43.1
>50
>50
>50
>50
0.26
43.41
>50
>50
>50
1.34
0.94
>50
3.46
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
>50
0.33
>50
>50
>50
>50
4.04
5.4
>50
33.5
>50
NS
NS
NS
NS
NS
2.11
2.38
NS
NS
NS
NS
NS
NS
NS
NS
0.02
0.12
1.81
0.06
0.13
3j
4a
4b
6.8
3.6
ꢀ7.6
ꢀ5.2
0.474
0.423
0.669
0.538
0.205
0.140
0.276
0.295
42.3
2.5
0.32
16.4
>50
34.1
24.1
19.3
NS
NS
0.13
NS
log P: Lipophilicity; log S: Solubility; Pe: Electrophilic properties; AntiN: Antineoplastic activity; Mut: Mutagenicity; Photo: Photosensitizer; MCC: Minimal Clastogenic
Concentration.
UV irradiation.
A
significant correlation could be observed
on the 3- and 6-substituted amino groups displayed higher
clastogenicity than those bearing aryl rings. This result demon-
strated that planar molecules with aliphatic substituents (e.g. 2a–
f) could better intercalate between base-pairs than diphenyl
substituted molecules 3a–j and focused on the critical role of the
steric cluster on DNA–acridine interactions. They also established
that the presence of a methyl radical on the border of the
substituted groups was determinant for the clastogenic activity.
However, all the 3,6-di-substituted acridines were far less clas-
togenic than the parent molecule proflavine: this indicated that
the addition of aliphatic or aromatic substituents on the 3- and 6-
amino groups of proflavine greatly decreased the capacity of the
molecule to interact with DNA. After UV irradiation, all the
derivatives found to be clastogenic in the dark, displayed a higher
activity. Moreover, four molecules which were deprived from
genotoxicity in the dark revealed a photo-inducible clastoge-
nicity: 3h, 3i, 3j and 4a. All these compounds, except 4a con-
tained a single substituent (F, Cl or CN) on position 4 of the aryl
rings. No correlation could be established between the cytotoxic
and clastogenic activities, suggesting that the interaction of
acridines with DNA or with DNA-replication enzymes could not
be considered as the single mechanism leading to cell growth
arrest. No significant correlation could be found between the
calculated electrophilic or mutagenic properties of the
compounds and their experimental clastogenicity. Moreover, no
significant correlation could be obtained between their calculated
photosensitising activity and their photo-inducible clastogenicity.
between the cytotoxic and the photo-cytotoxic activities against
CHO cells (P ¼ 0.01). This indicated that, although UV irradiation
greatly increased the toxicity of the compounds, the mechanism
by which proflavine derivatives may exert a photo-inducible
inhibition of cell proliferation was quite similar to those occur-
ring in the dark. However, no correlation could be established
between the calculated physicochemical properties of the
compounds (lipophilicity, solubility, electrophilicity) and their
cytotoxic or photo-cytotoxic activities. Finally, no correlation
could be observed between the calculated antineoplastic and the
measured cytotoxic properties.
In order to evaluate the specificity of acridine derivatives for
rapid growing tumour cells, the cytotoxicity of each compound
was also measured on normal human keratinocytes. In the dark,
almost all the 3,6-di-substituted acridines, except 2d and 4b, were
deprived from activity against keratinocytes. After UV irradiation
on the contrary, almost all the compounds exerted a significant
photo-cytotoxicity. 2c, 2f and 3a displayed the most important
photo-inducible properties, with IC50 averaging 0.14–0.57 mM and
corresponding to a 100 fold increase of their cytotoxicity as
compared to experiments performed in the dark. A significant
correlation could be observed between the cytotoxicity and the
photo-cytotoxicity of 3,6 di-substituted acridines against both
CHO cells and human keratinocytes (P ¼ 0.02 and P ¼ 0.001
respectively). This result implied that a similar mechanism of
action occurred in CHO cells and in human keratinocytes. They
also focussed on the weak specificity of the compounds for
tumour cells, as compared to their parent therapeutic molecule
proflavine.
The clastogenic properties of newly synthesized acridines
were assessed on Chinese Hamster Ovary cells. In the dark, six
derivatives exhibited a significant clastogenic activity. Among
them, 2c, 2b and 2e appeared the most clastogenic compounds,
with MCC reaching 1.01, 1.15 and 0.47
3f, 3g and 2a were less active. Acridines that bore aliphatic chains
Nevertheless,
a significant correlation could be established
between the clastogenic and photo-inducible clastogenic effects.
This result implied that UV irradiation could enhance the inter-
actions of the compounds with DNA without greatly modifying
their mechanism of action.
In order to better visualize the biological effects of acridines
on both tumour cells and keratinocytes, a specific cytotoxic
power (IC50-CHO/IC50-Kera), a clastogenic power (1/MCC) and
a specific photo-induced cytotoxic power were calculated for
mM respectively, whereas

Y. Benchabane et al. / European Journal of Medicinal Chemistry 44 (2009) 2459–2467
2463
5
4
r² = 0.62
P<0.001
3
1
2
1
2e
0
2c
3f
2b
3g
-1
-2
2a
Compounds with specific cytotoxicity
3c 3h
3i 3j 3e
3b 3d 4a
3a
2f
4b
2d
1,5
0,0
0,5
Log(Specific Cytotoxic Power)
1,0
5
4
r² = 0.76
P<0.001
Compounds with strong
photo-induced clastogenicity
3
1
3i
2c
2
3f
2b
3j
1
3g
4a
2e
2a
Fig. 6. Combined clastogenic effects of the antioxidant mannitol and compound 2c.
0
3h
5. Conclusion
Compound with specific
photo-induced cytotoxicty
-1
-2
2f
Results observed in this study clearly demonstrated
that compound 2d (N-[6-(pivalamino)acridin-3-yl]pivalamide)
displayed a specific cytotoxicity on CHO cells and was deprived
from significant clastogenicity. They also established that most
of the acridine compounds, which have been shown to
mainly absorb light rays at a wavelength of 330 nm, could be
photo-activated by UVA-visible irradiation. Compound 2c (N-[6-
(butyrylamino)acridin-3-yl]butanamide) appeared the strongest
photosensitizer, however its genotoxicity and photo-inducible
genotoxicity did not depend on oxidative events. This result
confirmed our previously published data [13], which showed that
compound 2c directly interacted with DNA through intercalation
between base-pairs.
Structure–activity relationships indicated that the presence of
a 1,1-dimethylethyl substituent on the positions 3 and 6 of the
acridine compounds was responsible for an interesting non-
clastogenic and specific cytotoxicity, whereas nonbranched
aliphatic chains tended to enhance the ability of the molecules to
interfere with DNA. They also demonstrated that compounds
bearing nonbranched aliphatic chains could better intercalate
between DNA base-pairs than those containing aromatic substitu-
ents, and underlined the important role of the steric cluster on
DNA–acridine interactions.
3c 3d
32decf34bb 3a
3e 4b
2d
0,0
0,5
1,0
1,5
2,0
Log(Specific Photo-induced Cytotoxic Power)
Fig. 5. Schematic representations of the cytotoxic and clastogenic properties in the
dark (5.1.) and after irradiation (5.2.).
each acridine derivative and reported in Fig. 5. A significant
linear correlation could be observed between the clastogenic
power and the specific cytotoxic power for all the compounds
except compound 2d. Similarly, a significant non-linear corre-
lation could be obtained between the photo-induced cytotoxic
power and the photo-induced clastogenic power, for all the
derivatives except compound 2d. This result indicated that all
the derivatives presented the same mechanisms of action,
except compound 2d, that interfered with different cellular
targets.
The clastogenic and photo-enhanced clastogenic effects of 2c on
CHO cells in the presence of mannitol are displayed in Fig. 6. Results
showed that mannitol, which significantly decreased by 42%
the micronucleated cell rates induced by a high UV irradiation
(20 J/cm2, data not shown), did not prevent cells from the clasto-
genic and photo-enhanced clastogenic activities of compound 2c.
This result indicated that the clastogenic and photo-enhanced
clastogenic activities of compound 2c did not significantly depend
on oxidative events.
Comparisons between the experimental results and the
physico-chemical or biological properties calculated by predictive
models showed that the cytotoxic and clastogenic activities
of acridine compounds resulted from various complex mecha-
nisms. They focussed on the difficulty to properly analyze the

2464
Y. Benchabane et al. / European Journal of Medicinal Chemistry 44 (2009) 2459–2467
pharmacological properties of acridine derivatives and to predict
their biological activities. They also underlined the necessity of
extended studies to better define their numerous cellular targets.
(DMSO-d6): 10.35 (NH), 8.86 (H9), 8.49 (H4 and H5), 7.59 (H2 and
H7), 8.03 (H1 and H8), 2.15 (Me). ESI-MS: 294 (M þ H)þ.
6.3.2. N-[6-(Propionylamino)acridin-3-yl]propanamide (2b) [13]
Proflavine 1 (210 mg, 1 mmol, 1 equ.) was dissolved into pyri-
dine (6 mL) under N2 and NEt3 (0.5 mL) was added. Then, propionic
anhydride (2.1 mmol, 2.1 equ., 0.27 mL) was added dropwise at
50 ꢁC and stirring was continued for 50 min more at 80 ꢁC. The
resulting mixture was poured into water (80 mL) and the obtained
precipitate was filtered, washed with water and dried. The resulting
solid was recrystallized from ethanol to give an orange powder 2b.
Yield 65%. M.p. > 260 ꢁC. NMR 1H (DMSO-d6): 10.28 (NH), 8.81 (H9),
8.48 (H4 and H5), 8.01 (H1 and H8), 7.60 (H2 and H7), 2.43 (CH2-10),
1.13 (Me-20). ESI-MS: 322 (M þ H)þ.
6. Experimental
6.1. General procedures
All reagents were of analytical grade, dried and purified when
necessary. Proflavine derivatives were dissolved in sterile dimethyl
sulfoxide (analytical grade, Sigma) and stored frozen at ꢀ70 ꢁC until
used. Proflavine (Pro) was used as positive control.
6.2. Mathematical models for prediction of physicochemical and
biological properties
6.3.3. N-[6-(Butyrylamino)acridin-3-yl]butanamide (2c) [13]
As reported for 2b but with butyric anhydride (0.35 mL). The
resulting solid was recrystallized from ethanol to give an orange
powder 2c. Yield 32%. M.p. 216 ꢁC. NMR 1H (DMSO-d6): 10.27 (NH),
8.80 (H9), 8.50 (H4 and H5), 7.98 (H1 and H8), 7.58 (H2 and H7), 2.38
(CH2-10), 1.67 (CH2-20), 1.09 (Me). ESI-MS: 350 (M þ H)þ.
Predictive values calculated for lipophilicity (log P representing
n-octanol/water partition coefficient) and solubility in water (log S),
as well as prediction of acid–base ionization constant pKa were
estimated by mathematical methods using ALOGPS v.2.0 software
according to the methodology described by Tetko et al. [19].
Predictive values concerning the electrophilic properties in relation
with a possible mutagenic activity were calculated according to the
mathematical model developed by Zheng et al. [16], available at the
web-server address http://dddc.ac.cn/adme/run.php.
Predictive values for antineoplasic, mutagenic and photo-
sensizer activities were investigated using the chemistry software
server PASS for prediction of biological activities spectrum (http://
195.178.207.233/PASS/), according to the mathematical model and
the database developed by Lagunin et al. [17] and Poroikov et al.
[20].
6.3.4. N-[6-(Pivalamino)acridin-3-yl]pivalamide (2d)
As report for 2a but with pivaloyl chloride (0.53 mL). A beige
powder 2d was recovered. Yield 74%. M.p. 158 ꢁC. NMR 1H (DMSO-
d6): 9.55 (NH), 8.83 (H9), 8.52 (H4 and H5), 8.02 (H1 and H8), 7.79 (H2
and H7), 1.30 (Me). ESI-MS: 379 (M þ H)þ.
6.3.5. Methyl N-[6-(methoxycarbonylamino)acridin-3-yl]-
carbamate (2e) [21]
Proflavine 1 (300 mg, 1.43 mmol, 1 equ.) and NaHCO3 (602 mg,
7.17 mmol, 5 equ.) was dissolved into acetone (15 mL) and methyl
chloroformate (1.11 mL, 14.34 mmol, 10 equ.) was added dropwise
at room temperature. The solution was stirred for 3 days at 60 ꢁC.
Then, the mixture was poured into a solution of aq. NaHCO3 (40%,
80 mL). After cooling at 5 ꢁC, the obtained precipitate was filtered,
washed in water and dried. The solid product was finally washed
with diethyl oxide (15 mL) and dried to give a brown powder 2e.
Yield 70%. M.p. > 260 ꢁC, (litt. >360 ꢁC) [21]. NMR 1H (CD3OD): 8.77
(H9), 8.27 (H4 and H5), 7.96 (H1 and H8), 7.57 (H2 and H7), 3.82
(OMe). ESI-MS: 326 (M þ H)þ.
6.3. Chemistry
Melting points were performed on an Electrothermal 9200
melting point apparatus and are uncorrected. The 1H and 13C NMR
spectra were measured on
a BRUKER AC 300 spectrometer
(300.13 MHz for 1H). The mass spectra were performed with a QStar
Elite mass Spectrometer (Applied Biosystems SCIEX, Concord, ON,
Canada) equipped with an electrospray ionization (ESI) source
operated in the positive ion mode. The absorption spectra were
performed at a fixed concentration of 1 ꢂ10ꢀ5 mg/mL in DMSO
with a Perkin–Elmer Lambda 800 spectrophotometer maintained
at 25 ꢁC. Fluorescence measurements were achieved with
a Horiba–Jobin–Yvon fluoromax-3 spectrofluorimeter equipped
with slit widths of 2/2 nm. Fluorescence spectra were obtained by
using 1 ꢂ0.2 cm cells (Hellma), at an excitation wavelength of
330 nm, corresponding to the maximum absorption wavelength for
all the compounds.
6.3.6. Tert-butyl N-[(6-tert-butoxycarbonylamino)acridin-3-
yl]carbamate (2f) [22]
Proflavine 1 (500 mg, 2.39 mmol, 1 equ.) was dissolved into
acetone (63 mL) and di-tert-butyl dicarbonate (6.25 g, 28.57 mmol,
12 equ.) was added dropwise at room temperature under N2. The
solution was stirred for 3 days under reflux. The solvent was
evaporated in vacuo and the crude product was purified by column
chromatography on silica gel (EtOAc/Hexane: 1/1) to yield a yellow
powder 2f. Yield 75%. M.p. 165 ꢁC, (Litt. 163–165 ꢁC) [22]. NMR 1H
(DMSO-d6): 9.80 (NH), 8.76 (H9), 8.20 (H4 and H5), 7.96 (H1 and H8),
7.58 (H2 and H7), 1.54 (Me). ESI-MS: 411(M þ H)þ.
The proflavine used neutral. Hemisulfate of 3,6-diaminoacridine
(proflavine), commercially available, was dissolved into water and
NH4OH 10% was added until pH ¼ 8 under stirring. Then, the
solution was filtered, washed with water and the resulting product
dried to yield neutral proflavine.
6.3.7. N-[6-(Benzoylamino)acridin-3-yl]benzamide (3a) [13]
6.3.1. N-[6-(Acetylamino)acridin-3-yl]acetamide (2a) [13]
Proflavine 1 (300 mg, 1.43 mmol, 1 equ.) and K2CO3 (1.98 g,
14.34 mmol, 10 equ.) was dissolved into acetone (40 mL) under N2.
Then, a solution of benzoyl chloride (4.30 mmol, 3 equ.), (0.5 mL)
into acetone (10 mL) was added dropwise at 0 ꢁC under vigorous
stirring. Stirring was continued 15 h more at room temperature;
then the mixture was poured into water (30 mL) and followed by
20 mL of satd. aq. NaHCO3 solution. After cooling at 5 ꢁC, the
obtained precipitate was filtered, washed with water, followed by
ether (15 mL) and dried to yield an orange powder 3a. Yield 79%.
M.p. 248 ꢁC. NMR 1H (DMSO-d6): 10.69 (NH), 8.95 (H9), 8.73 (H4 and
Proflavine 1 (300 mg, 1.43 mmol, 1 equ.) and K2CO3 (1.98 g,
14.34 mmol, 10 equ.) was dissolved into acetone (40 mL) under N2.
Then, a solution of acetyl chloride (0.31 mL, 4.30 mmol, 3 equ.) into
acetone (10 mL) was added dropwise at 0 ꢁC under vigorously
stirring. Stirring was continued 15 h more at room temperature.
Then, the mixture was poured into a solution of aq. NaHCO3 (40%,
80 mL). After cooling at 5 ꢁC, the obtained precipitate was filtered,
washed in water and dried. Recrystallization from ethanol yielded
pure 2a as brown powder. Yield 60%. M.p. > 260 ꢁC. NMR 1H

Y. Benchabane et al. / European Journal of Medicinal Chemistry 44 (2009) 2459–2467
2465
0
0
0
H5), 8.13 (H1 and H8), 8.05 (H2 and H6 ),þ7.92 (H2 and H7), 7.65 (H4 ),
6.3.15. 4-Cyano-N-[6-(4-cyanobenzoylamino)-acridin-3-yl]-
benzamide (3i)
As reported for 3f but with 4-cyanobenzoyl chloride (350 mg).
The resulting product was washed with hot ethanol (15 mL) and
dried to yield a brown solid 3i. Yield 23%. M.p. > 260 ꢁC. NMR 1H
0
0
7.58 (H3 and H5 ). ESI-MS: 418 (M þ H) .
6.3.8. 4-Methoxy-N-[6-(4-methoxybenzoylamino)acridin-3-
yl]benzamide (3b) [13]
0
As reported for 3a but with 4-methoxybenzoyl chloride
(DMSO-d6): 10.87 (NH), 8.95 (H9), 8.71 (H4 and H5), 8.20 (H2 and
0
0
0
(0.58 mL). 3b was obtained as
a
beige powder. Yield 85%.
H6 ), 8.14 (H1 and H8), 8.07 (H3 and H5 ), 7.89 (H2 and H7). ESI-MS:
M.p. > 260 ꢁC. NMR 1H (DMSO-d6): 10.50 (NH), 8.91 (H9), 8.65 (H4
468 (M þ H)þ.
0
0
and H5), 8.10 (H1 and H8), 8.03 (H2 and H6 ), 7.8þ9 (H2 and H7), 7.09
0
0
(H3 and H5 ), 3.85 (OCH3). ESI-MS: 478 (M þ H) .
6.3.16. 4-Amino-N-[6-(4-aminobenzoylamino)acridin-3-
yl]benzamide (3j)
6.3.9. 3,4-Dichloro-N-[6-(3,4-dichlorobenzoylamino)acridin-3-
yl]benzamide (3c)
As reported for 3a but with 3,4-dichlorobenzoyl chloride
(901 mg). 3c was obtained as an orange powder. Yield 84%.
M.p. > 260 ꢁC. NMR 1H (DMSO-d6): 10.78 (NH), 8.94 (H9), 8.69 (H4
A solution of 3e (400 mg, 0.79 mmol, 1 equ.) and 10% palladium
on carbon (60 mg, 15% wt) in ethanol (40 mL) was prepared. H2 was
introduced into the flask (3 bar) and the reaction mixture was
stirred vigorously for 20 h. Then, more ethanol (30 mL) was added
and the solution was filtered by passing through Celite. The alco-
holic filtrate was concentrated and the resulting crude product was
chromatographied on silica gel (DCM/MeOH: 9/1 to 8/2) to yield
a red-brown powder 3j. Yield 26%. M.p. 238 ꢁC. NMR 1H (DMSO-d6):
10.22 (NH), 8.90 (H9), 8.67 (H4 and H5), 8.07 (H1 and H8), 7.93 (H2
0
0
0
and H5), 8.31 (H2 ), 8.14 (H1 and H8), 8.03 (H6 ), 7.89 (H5 ), 7.87 (H2
and H7). ESI-MS: 554/556/558 (M þ H)þ.
6.3.10. 3,5-Di-(trifluoromethyl)-N-[6-(3,5-di-
0
0
0
0
(trifluoromethyl)benzoylamino)acridin-3-yl]benzamide (3d)
As reported for 3a but with 3,5-di-(trifluoromethyl)benzoyl
chloride (0.78 mL). The resulting crude product was dissolved in
methanol (30 mL), filtered and concentrated to yield a brown
product 3d. Yield 47%. M.p. 190 ꢁC. NMR 1H (DMSO-d6): 11.03 (NH),
and H7), 7.83 (H2 and H6 ), 6.65 (H3 and H5 ), 5.87 (NH2). ESI-MS:
468 (M þ H)þ.
6.3.17. N-[6-(1-Naphtamino)acridin-3-yl]-1-naphthamide (4a)
As reported for 3a but with 1-naphthoyl chloride (0.65 mL). The
obtained product was washed with hot water (20 mL) followed by
diethyl oxide (20 mL). After drying an orange powder 4a was
recovered. Yield 86%. M.p. > 260 ꢁC. NMR 1H (DMSO-d6): 11.02
0
0
0
8.98 (H9), 8.71 (H2 and H6 ), 7.69 (H4 ), 8.42 (H4 and H5), 8.18 (H1
and H8), 7.89 (H2 and H7). ESI-MS: 690 (M þ H)þ.
0
6.3.11. 4-Nitro-N-[6-(4-nitrobenzoylamino)acridin-3-
yl]benzamide (3e)
As reported for 3a but with 4-nitrobenzoyl chloride (800 mg).
The resulting product was washed with diethyl oxide (10 mL) and
dried to yield a brown solid 3e. Yield 57%. M.p. > 260 ꢁC. NMR 1H
(NH), 8.96 (H9), 8.79 (H4 and H5), 8.28 (H8 ), 8.15 (H1 and H8), 8.14
0
0
0
0
0
(H4 ), 8.06 (H5 ), 7.89 (H2 and H7), 7.89 (H2 ), 7.67 (H3 ), 7.64 (H7 ),
þ
0
7.62 (H6 ). ESI-MS: 518 (M þ H) .
6.3.18. N-[6-(Furan-2-carboxamido)acridin-3-yl]furan-2-
carboxamide (4b)
0
(DMSO-d6): 10.97 (NH), 8.97 (H9), 8.72 (H4 and H5), 8.43 (H2 and
0
0
0
H6 ), 8.28 (H3 and H5 ), 8.15 (H1 and H8), 7.91 (H2 and H7). ESI-MS:
As reported for 3a but with 2-furoyl chloride (0.43 mL). 4b was
obtained as an orange powder. Yield 70%. M.p. 174 ꢁC. NMR 1H
(DMSO-d6): 10.58 (NH), 8.89 (H9), 8.66 (H4 and H5), 8.10 (H1 and
508 (M þ H)þ
0
0
0
6.3.12. 4-Methyl-N-[6-(4-methylbenzoylamino)acridin-3-
yl]benzamide (3f) [23]
H8), 8.00 (H4 ), 7.91 (H2 and H7), 7.46 (H2 ), 6.76 (H3 ). ESI-MS: 398
(M þ H)þ.
Proflavine 1 (210 mg, 1 mmol, 1 equ.) was dissolved into pyri-
dine (6 mL) under N2 and NEt3 (0.5 mL) was added. Then, 4-
methylbenzoyl chloride (2.1 mmol, 2.1 equ., 0.40 mL) was added
dropwise at 50 ꢁC and stirring was continued for 50 min. more at
80 ꢁC. The resulting mixture was poured into water (80 mL) and the
obtained precipitate was filtered, washed with water and dried.
Recrystallization from ethanol yielded a brown powder 3f. Yield
82%. M.p. > 260 ꢁC [23]. NMR 1H (DMSO-d6): 10.55 (NH), 8.90 (H9),
6.4. Biology
6.4.1. Irradiation procedure
Irradiation of cell cultures was carried out with a solar simulator
Suntest CPSþ (Atlas Material Testing Technology BV, Moussy le
Neuf, France) apparatus equipped with a xenon arc lamp (1100 W),
special glass filters restricting transmission of light below 290 nm
and near IR-blocking filter. The irradiance for the photoactivation of
acridine derivatives was fixed at 750 W mꢀ2 throughout the
experiments and the combined light dose was 4.5 J/cm2 for one
minute irradiation, corresponding to 0.36 J/cm2 for UVA and 4.14 J/
cm2 for visible light. The temperature of the samples was kept at
4 ꢁC using a water cooling circuit in the irradiation chamber. UVA–
visible light (320–800 nm) was obtained using the solar ID65 filter
plus a window glass filter.
0
0
8.68 (H4 and H5), 8.09 (H1 and H8), 8.00 (H2 and H6 ), 7.91 (H2 and
þ
0
0
H7), 7.43 (H3 and H5 ), 2.42 (Me). ESI-MS: 446 (M þ H) .
6.3.13. 4-Chloro-N-[6-(4-chlorobenzoylamino)acridin-3-
yl]benzamide (3g) [13]
As reported for 3f but with 4-chlorobenzoyl chloride (0.24 mL).
The resulting solid was recrystallized from ethanol to give an orange
powder 3g. Yield 90%. M.p. > 260 ꢁC. NMR 1H (DMSO-d6): 10.71
0
0
(NH), 8.92 (H9), 8.68 (H4 and H5), 8.11 (H1 and H8), 8.06 (H2 and H6 ),
þ
0
0
7.89 (H2 and H7), 7.66 (H3 and H5 ). ESI-MS: 486/488 (M þ H) .
6.4.2. Photo-inducible antiproliferative activity on Chinese Hamster
ovary cells and human keratinocytes
6.3.14. 4-Fluoro-N-[6-(4-fluorobenzoylamino)acridin-3-
yl]benzamide (3h) [13]
As reported for 3f but with 4-fluorobenzoyl chloride (0.25 mL).
After recrystallization in ethanol, 3h was obtained as an orange
powder. Yield 96%. M.p. > 260 ꢁC. NMR 1H (DMSO-d6): 10.77 (NH),
Chinese Hamster Ovary cells (CHO) were maintained in Mc Coys’
5A medium (Sigma, St Quentin-Fallavier, France) containing 10%
foetal calf serum (Eurobio, Paris, France), 2 mM glutamine and
penicillin–streptomycin (100 U mLꢀ1–10 g mLꢀ1). Human normal
m
keratinocytes were isolated from infant foreskin obtained after
circumcisions according to the technique described by Decome
et al. [24]. Cells were suspended in K-SFM medium containing
0
0
9.07 (H9), 8.76 (H4 and H5), 8.16 (H1 and H8), 8.13 (H2þand H6 ), 7.91
0
0
(H2 and H7), 7.41 (H3 and H5 ). ESI-MS: 454 (M þ H) .

2466
Y. Benchabane et al. / European Journal of Medicinal Chemistry 44 (2009) 2459–2467
25
m
g/mL BPE and 0.1–0.2 ng/mL rEGF. Under these conditions,
were identified according to the morphological criteria previously
defined by Kirsch-Volders et al. [25]. Statistical differences between
controls and treated samples were performed by the
pure keratinocyte cultures were obtained within 2 weeks. CHO and
keratinocyte cultures were incubated at 37 ꢁC in humidified
atmosphere containing 5% CO2 and the culture mediums were
changed every 3 days.
c
2-test:
a significant clastogenic activity was detected when a dose-
dependent increase of micronucleated cells was observed and
when the induced micronucleated cell rates were significant for
one concentration at least. For each acridine, a dose–effect rela-
tionship was calculated by non-linear regression analysis using the
TableCurve software (Jandel Scientific Software, San Rafael, CA,
USA). The minimal clastogenic concentration (MCC) was defined as
the lowest acridine concentration capable to induce a significant
increase of the number of micronucleated cells.
The effects of acridines on the growth of CHO cells (ATCC,
Manassas VA, USA) were assessed by the colorimetric determi-
nation of cell viability using the oxidation–reduction indicator
WST1Ò. Non-confluent CHO cells (50,000 by well) were diluted
in Mc Coys’ 5A medium (Sigma, St Quentin-Fallavier, France)
containing 10% foetal calf serum (Eurobio, Paris, France), 2 mM
glutamine and penicillin–streptomycin (100 U mLꢀ1–10 g mLꢀ1
m )
and seeded in 96-well plates. After a 24 h incubation period at
37 ꢁC with 5% CO2, a range of acridine concentrations was incor-
porated in duplicate cultures (final DMSO concentration less than
0.5%). Cells were immediately irradiated and incubated for 24 h at
37 ꢁC. At the end of the incubation period, cultures were rinced by
three successive washes in Phosphate Saline Buffer (PBS)
and incubated for 20 min in RPMI medium without phenol red
containing 10% of the oxidation–reduction indicator WST1Ò at
37 ꢁC. Cell viability was evaluated by the assessment of WST1
absorbance at 450 nm in a microplate spectrophotometer MRXÒ II
(Dynex technologies, VA, USA). IC50-CHO was defined as the
concentration of acridine required to induce a 50% decrease of cell
growth, corresponding to a 50% reduction of WST1Ò indicator as
compared to the control culture.
6.4.4. Clastogenic effects of compound 2c on CHO cells pre-treated
with mannitol
A total of 50,000 Chinese Hamster Ovary cells (CHO-K1, ATCC)
were plated in chamber slides containing Mc Coy’s 5A medium
(Sigma, St Quentin-Fallavier, France) supplemented with 10% foetal
calf serum, 1 mM glutamine and penicillin–streptomycin
(100 U mLꢀ1–10 g mLꢀ1) and were incubated at 37 ꢁC in humidi-
m
fied atmosphere containing 5% CO2. After a 24 h incubation period,
2 mM mannitol was added in cell cultures and cells were incubated
at 37 ꢁC for 2 h. Then, three sets of experiments were performed:
- control cultures and cultures containing 2 mM mannitol were
submitted to a 20 J/cm2 irradiation, in order to verify the
protective effect of mannitol against the genotoxic effects of
UVA light,
6.4.3. Photo-inducible clastogenic activity on CHO cells
The micronucleus assay was performed according to the meth-
odology recommended by Kirsch-Volders et al. [25]. A total of
50,000 Chinese Hamster Ovary cells (CHO-K1, ATCC) were plated in
chamber slides containing Mc Coy’s 5A medium (Sigma, St Quentin-
Fallavier, France) supplemented with 10% foetal calf serum, 1 mM
- control cultures and cultures containing 2 mM mannitol were
treated with 15
the dark,
mM of compound 2c and were maintained in
- control cultures and cultures containing 2 mM mannitol were
treated with 0.15 mM of compound 2c and were submitted to
glutamine and penicillin–streptomycin (100 U mLꢀ1–10
m
g mLꢀ1
)
a 4.5 J/cm2 irradiation.
and were incubated at 37 ꢁC in humidified atmosphere containing
5% CO2. After a 24 h incubation period, two sets of experiments were
performed: a range of acridine concentrations was incorporated in
duplicate cultures maintained in the dark, an other range of acridine
concentrations was incorporated in duplicate cultures and imme-
diately irradiated. Cells were incubated at 37 ꢁC during 3 h, then
culture medium was removed and cells were incubated in fresh
Cells were incubated at 37 ꢁC during 3 h, then culture medium
was removed and cells were incubated in fresh medium containing
5 m
g mLꢀ1 cytochalasin B to arrest cytokinesis. After an additional
24 h incubation period at 37 ꢁC, the culture medium was removed;
cells were submitted to two successive washes with PBS and fixed
with methanol (HPLC purity grade solvent). Staining of air-dried
slides was performed with 5% Giemsa stain in Milli-Q water for
20 min. Slide analysis was performed microscopically at 1000ꢂ
magnification.
medium containing 5 m
g mLꢀ1 cytochalasin B to arrest cytokinesis.
After an additional 24 h incubation period at 37 ꢁC, the culture
medium was removed; cells were submitted to two successive
washes with PBS and fixed with methanol (HPLC purity grade
solvent). Staining of air-dried slides was performed with 5% Giemsa
stain in Milli-Q water for 20 min. Slide analysis was performed
microscopically at 1000ꢂ magnification.
6.5. Statistical analysis
The rank correlation test of Spearman was used for studying
coupled variables. Analysis was performed by Statgraphics plus
software (Statistical graphics corporation, USA).
The Proliferative Index (PI) was considered as a measure of
cytotoxicity. It was determined by scoring the number of multi-
nucleated cells among 500 Giemsa-stained cells with well-
preserved cytoplasm and was calculated as follows:
References
½MNC þ 2 ꢂ ðBNCÞ þ 3 ꢂ ðTNCÞꢃ
PI ¼
[1] W.A. Denny, Curr. Med. Chem. 9 (2002) 1655–1665.
[2] M. Wainwright, J. Antimicrob. Chemother. 47 (2001) 1–13.
[3] I. Antonin, Curr. Med. Chem. 9 (2002) 1701–1716.
[4] M. Demeunynck, A. Charmantray, A. Martelli, Curr. Pharm. Des 7 (2001) 1703–
1724.
500
With:
[5] L.S. Lerman, J. Mol. Biol. 3 (1961) 18–30.
[6] Z. Topcu, J. Clin. Pharm. Ther 26 (2001) 405–416.
MNC: number of cells containing one nucleus
BNC: number of cells containing two nuclei
TNC: number of cells containing three nuclei at least.
[7] L.S. Ripley, A. Clark, Proc. Natl. Acad. Sci. U.S.A. 83 (1986) 6954–6958.
[8] A. Nasim, T. Brychcy, Mutat. Res. 65 (1979) 261–288.
[9] L.R. Ferguson, W.A. Denny, Mutat. Res. 258 (1991) 123–160.
[10] M. Wainwright, D.A. Phoenix, J. Marland, D.R. Wareing, F.J. Bolton, J. Anti-
microb. Chemother. 40 (1997) 587–589.
The micronucleated cell (MNC) rates were determined for
concentrations inducing less than 50% toxicity: for each concen-
tration, 2000 binucleated cells were examined and micronuclei
[11] W.J. Firth, S.G. Rock, B.R. Brown, L.W. Yielding, Mutat. Res. 81 (1981) 295–309.
[12] Y. Iwamoto, L.R. Ferguson, A. Pearson, B.C. Baguley, Mutat. Res. 268 (1992) 35–41.

Y. Benchabane et al. / European Journal of Medicinal Chemistry 44 (2009) 2459–2467
2467
[13] C. Di Giorgio, K. Shimi, G. Boyer, F. Delmas, J.P. Galy, Eur. J. Med. Chem. 42
(2007) 1277–1284.
[20] V.V. Poroikov, D. Filimonov, Y. Borodina, A. Lagunin, A. Kos, J. Chem. Inf.
Comput. Sci. 40 (2001) 1349–1355.
[14] A. Miloudi, D. El Abed, G. Boyer, J.P. Galy, Heterocycles 68 (2006) 2595–2605.
[15] R. Fenech, Mutat. Res. 285 (1993) 35–44.
[21] A. Lorente, M. Fernandez-Saiz, J.-F. Espinosa, C. Jaime, J.-M. Lehn, J.-
P. Vigneron, Tetrahedron Lett. 36 (1995) 5261–5264.
[16] M. Zheng, Z. Liu, C. Xue, W. Zhu, K. Chen, X. Luo, H. Jiang, Bioinformatics 22
(2006) 2099–2106.
[22] A. Lorente, Y.-G. Vazquez, M.-J. Fernandez, A. Ferrandez, Biorg. Med. Chem. 12
(2004) 4307–4312.
[17] A. Lagunin, A. Stepanchikova, D. Filimonov, V. Poroikov, Bioinformatics 16
(2000) 747–748.
[23] K. Shimi, G. Boyer, R. Faure, J.-P. Galy, Magn. Reson. Chem. 43 (2005) 1077–1079.
[24] L. Decome, M.P. De Meo, A. Geffard, O. Doucet, G. Dumenil, A. Botta, J. Pho-
tochem. Photobiol. B 79 (2005) 101–108.
´
[18] C. Botta, C. Di Giorgio, A.S. Sabatier, M. De Meo, J. Photochem. Photobiol., B 25
(2008) 24–34.
[25] M. Kirsch-Volders, T. Sofuni, M. Aardema, S. Albertini, D. Eastmond Fenech,
M. Ishidate, S. Kirchner, E. Lorge, T. Morita, H. Norppa, J. Surrales,
A. Vanhauwaert, A. Wakata, Mutat. Res. 540 (2003) 153–163.
[19] I.V. Tekto, Y. Tanchuk, A. Villa, J. Chem. Inf. Comput. Sci. 41 (2001) 1407–
1421.