Synthesis and antiproliferative activity of some variously substituted acridine and azacridine derivatives

Article abstract of DOI:10.1016/S0223-5234(00)00170-7

A group of 9-substituted acridine and azacridine derivatives (m-AMSA analogues) were synthesised following classical procedures as potential antitumour agents with inhibitory effects on DNA topoisomerase II. Some were found to have noticeable cytotoxicity against human HL-60 and HeLa cells grown in culture. Their non-covalent interactions with calf thymus DNA have been studied using fluorescence quenching. We evaluated DNA damage produced by the tested compounds by means of DNA filter elution and protein precipitation techniques. Catalytic studies carried out with purified topoisomerase confirmed these agents as antitopoisomerase inhibitors. Chemotherapy of solid-tumour-bearing mice with tested compounds allowed an aza-analogue (compound IIIb), as potent as m-AMSA but less toxic towards the host, to be recognised. (C) 2000 Editions scientifiques et medicales Elsevier SAS.

Full text of DOI:10.1016/S0223-5234(00)00170-7

Eur. J. Med. Chem. 35 (2000) 827−837
© 2000 Éditions scientifiques et médicales Elsevier SAS. All rights reserved
S0223523400001707/FLA
Original article
Synthesis and antiproliferative activity of some variously substituted acridine
and azacridine derivatives
Maria Grazia Ferlin, Cristina Marzano, Gianfranco Chiarelotto, Francarosa Baccichetti, Franco
Bordin*
Department of Pharmaceutical Sciences of Padua University, Centro di Studio sulla Chimica del Farmaco e dei Prodotti Biologicamente
Attivi del CNR, (associated with the National Institute for the Chemistry of Biological Systems), Padova, Italy
Received 1 July 1999; accepted 28 March 2000
Abstract – A group of 9-substituted acridine and azacridine derivatives (m-AMSA analogues) were synthesised following classical
procedures as potential antitumour agents with inhibitory effects on DNA topoisomerase II. Some were found to have noticeable cytotoxicity
against human HL-60 and HeLa cells grown in culture. Their non-covalent interactions with calf thymus DNA have been studied using
fluorescence quenching. We evaluated DNA damage produced by the tested compounds by means of DNA filter elution and protein
precipitation techniques. Catalytic studies carried out with purified topoisomerase confirmed these agents as antitopoisomerase inhibitors.
Chemotherapy of solid-tumour-bearing mice with tested compounds allowed an aza-analogue (compound IIIb), as potent as m-AMSA but less
toxic towards the host, to be recognised. © 2000 Éditions scientifiques et médicales Elsevier SAS
amsacrine / topoisomerase II inhibitors / DNA damage
1. Introduction
benzo-naphthyridine derivatives showed antileukaemic
activities comparable to those of the parent compound,
but with lower potency [18, 19].
During the last 20–30 years a large number of deriva-
tives belonging to the general class of anilinoacridines
have been prepared and evaluated extensively as antima-
larial [1, 2], antileishmanial [3], antitrypanosomal [4, 5]
and anticancer agents [6–8].
As part of our program on polyetherocyclic com-
pounds showing anti-topoisomerase activity, we were
interested in preparing new aza-analogues with the aim of
evaluating the influence of a substituted tricyclic ring on
their biological activity. As a first approach, we planned
the compound having the benzo[b] [1, 5]naphthyridine
bi-substituted nucleus, as in the antimalarial azacridine
[20], and the same side chain (o-methoxy-p-
methansolfonamido-aniline) as m-AMSA.
Among these compounds, amsacrine (m-AMSA) has
shown high antileukaemic activity [9]. This was attrib-
uted later to topoisomerase II-mediated cleavage of
double-stranded DNA by a mechanism which appears to
be common to DNA-intercalating agents [10, 11], and
which has now become a clinically useful drug [12, 13].
In the field of chemotherapeutics, the further develop-
ment of m-AMSA as lead compound has yielded several
subsets of many analogues, which permit consistent SAR
and QSAR findings [14–17]. Aza analogues were pre-
pared as non-classical bioisosters of tumour inhibiting
electro-deficient nitro-substituted 9-(2′-methoxy-4′-
methansolfonamido-anilino)-acridines, and some isomeric
In the present work we describe the synthesis and
results of some preliminary in vitro and in vivo biological
tests of both the novel N-isosteric compound and the
known corresponding m-AMSA analogous derivative
taken as a comparison [16]. At the same time we were
also interested in testing the available corresponding
9-chloro- and 9-phenoxy- intermediates with the aim of
gaining a greater knowledge of their possible biological
activity than has been found in the literature [21].
* Correspondence and reprints: bordin@dsfarm.unipd.it

828
M.G. Ferlin et al. / Eur. J. Med. Chem. 35 (2000) 827–837
Figure 2. Reagents and conditions: i: ClCOOCH2CH3, THF,
reflux; ii: H2, Pd/C; ii: CH3SO2Cl, pyridine; iv: NaOH 20%,
reflux; v: HCl, absolute ethanol.
Figure 1. Reagents and conditions: A) Direct method, i:
absolute alcohol, NH2Ar(o-OCH3)(p-NHSO2CH3), reflux; B)
Indirect method, ii: phenol, NaOH, 120–130 °C; iii: phenol,
NH2Ar(o-OCH3)(p-NHSO2CH3). HCl, 130–140 °C.
ethyl-carbamate protecting group finally produced the
target compound in good yields.
2. Chemistry
3. Biology
Compounds IIIa,b were prepared starting from com-
mercial 2-methoxy-6,9-dichloro-derivatives Ia,b follow-
ing classical synthetic procedures: routes A and B shown
in figure 1. Acridine derivatives have already been
described [16]. Direct nucleophylic substitution of the
chlorine in the 9 position of both acridine and aza-
acridine rings by aniline derivative afforded desired
chlorides IIIa,b in good yields (route A).
9-Substituted acridine and azacridine (m-AMSA ana-
logues) were investigated for their affinity to DNA in
vitro. They were submitted to screening for antiprolifera-
tive activity and DNA damage to mammalian cells.
Antitumour activity was also tested in solid-tumour-
bearing mice. In all biological studies, m-AMSA was
used as reference compound.
Indirectly, (route B) compounds IIIa,b were also
obtained, but in higher yields than the above by the
reaction between phenoxy-derivatives IIa,b and
o-methoxy-p-methansolfonamido-aniline chloride in phe-
nol.
4. Results
4.1. Interaction with DNA in vitro
Intermediate phenoxy-derivatives were prepared from
Ia,b by heating with phenate, isolated and purified [22].
Figure 2 describes the known pathway to o-methoxy-
p-methansolfonamido-aniline in which the original amine
protecting acetyl group [8] is replaced by an ethyl-
formate group. This replacement was found to be particu-
larly efficient in the cases of both protection and depro-
tection because of the mild reaction conditions applied,
which produced higher yields. The starting p-nitro-m-
methoxy-aniline was easily transformed in urethanic
derivative and after carrying out catalytic reduction and
mesylation of the amine group, basic hydrolysis of the
Spectophotometric analysis of the interactions between
drug and DNA revealed several isosbestic points, thus
suggesting only one mode of binding (data not shown).
Unfortunately the changes in absorption spectra of the
free and DNA-bound compounds were too weak to
permit correct determinations of DNA binding by UV
spectroscopy. As m-AMSA-analogues showed high emis-
sion intensities in the characteristic acridine region
(400–500 nm) of the UV spectra, we assessed their ability
to bind to nucleic acids by fluorescence titrations, as
described by Gimenez-Arnau et al. [23]. Figure 3 reports
the quenching of fluorescence of tested compounds at

M.G. Ferlin et al. / Eur. J. Med. Chem. 35 (2000) 827–837
829
Table I. Detection of cell lethality.
Compound Dye exclusion Colony formation [3H]TdR uptake
HL-60 cellsa HeLa cellsb Ehrlich cellsc
IC50 ± SE (µM) IC50 ± SE (µM) IC50 ± SE (µM)
Ia
Ib
IIa
IIb
IIIa
IIIb
m-AMSA
34.9 ± 0.09
22.7 ± 0.10
18.5 ± 0.12
5.1 ± 0.17
1.43 ± 0.14
1.14 ± 0.08
1.15 ± 0.01
3.8 ± 0.11
2.44 ± 0.04
2.7 ± 0.17
0.9 ± 0.07
0.52 ± 0.01
0.49 ± 0.03
0.75 ± 0.06
37.66 ± 0.54
21.80 ± 0.33
33.49 ± 0.29
13.38 ± 0.21
12.4 ± 1.67
8.77 ± 0.10
10.5 ± 0.22
IC50 values were determined by probit analysis. Linearity of the
thus computed regressions was checked by the ꢀ2 test. (P = 0.05).
SE = standard error. a Cell viability was determined by Trypan blue
dye exclusion in HL-60 cells (1 × 106.mL–1) after 12 h incubation
in the presence of increasing concentrations of m-AMSA ana-
logues. b Clonal growth capacity of HeLa cells cultivated in vitro
was evaluated after 3 h incubation in the presence of different
concentrations of m-AMSA analogues. c Ehrlich cells were incu-
bated with tritiated thymidine and exposed to increasing concen-
trations of m-AMSA analogues for 1 h. Acid insoluble radioactivity
was then determined. Each experimental condition was checked in
at least triplicate samples.
Figure 3. Fluorescence of m-AMSA analogues as a function of
calf thymus DNA concentration. F/F0 is the relative fluores-
cence, where F is the fluorescence intensity of the ligand
quenched with DNA and F0 is the fluorescence intensity of free
ligand (see materials and methods). The symbols are: Ia (▲), Ib
(▲), IIa ( ), IIb (™), IIIa (▲), IIIb (▲), m-AMSA ([) was
used as reference.
human leukaemia HL-60 cells and the results from this
set of experiments are summarised in terms of IC50 values
in table I. For the purposes of comparison, the cytotox-
icity of m-AMSA was evaluated under the same condi-
tions. Both derivatives Ia and Ib, lacking the 9-aromatic
substituent, were poorly effective. Of the 9-phenoxy
derivatives, only IIb exhibited a significant effect while
IIa was much less cytotoxic.
Methanesulfonamidoaniline derivatives IIIa and IIIb
appeared to be the most potent derivatives with IC50
values ranging from 0.1–0.4 µM very similar to those of
m-AMSA.
Similar findings were observed measuring the inhibi-
tion of [3H]TdR incorporation in mouse ascite cells
treated with increasing concentrations of tested m-AMSA
analogues. The IC50 values are reported in table I.
Compounds Ib, Ia and IIa induced a moderate inhibi-
tion of DNA synthesis. An evident increase in activity
was obtained with derivatives IIb, IIIa and IIIb which
are more, or at least as effective, as amsacrine.
The clonogenic assay was used to determine human
adenocarcinoma HeLa cell survival following exposure
to various concentrations of tested compounds. The drug
inhibitory effect on cell proliferation was concentration
dependent (data not shown). Table I shows the results
obtained in terms of IC50 values. Again, compounds IIb,
IIIa and IIIb were the most potent derivatives, provoking
a strong inhibition of colony forming ability.
constant concentrations with increasing concentrations of
DNA. The observed fluorescence intensities were af-
fected differently by the presence of calf thymus DNA.
However, the fluorescence of each derivative promptly
decreases on addition of DNA up to a certain ratio
between drug and DNA, and is followed by stabilisation
of fluorescence at higher DNA/drug ratios. The evident
quenching of fluorescence in the case of compounds Ib,
IIb and IIIb suggests that aza-acridines are strong
DNA-binding ligands showing DNA affinities signifi-
cantly greater than those of m-AMSA. The 9-anilino-
acridine derivative IIIa, with the same side chain of
m-AMSA shows an affinity for binding with DNA very
similar to that of m-AMSA. In the same order of
magnitude were the affinities between DNA and acridine
Ia or phenoxyacridine IIa, but their fluorescences were
also quenched less in the presence of high amounts of
DNA.
4.2. Cytotoxic activity
m-AMSA analogues were examined for their cytotoxic
properties. The dye exclusion test was performed on

830
M.G. Ferlin et al. / Eur. J. Med. Chem. 35 (2000) 827–837
Table II. Detection of DNA damage in HeLa cells.
correlated to cytotoxic potency. Moreover, most disap-
peared within 30 min after drug removal (data not shown).
This phenomenon has been previously described with
topoisomerase inhibitors [26].
Treatment
Cross-linking
Elution rate
Coefficient ± SEa
(× 102 ± SEb)
Control
IIb
IIIa
IIIb
m-AMSA
1
1.55 ± 0.15
1.89 ± 0.16
2.53 ± 0.18
3.26 ± 0.20
2.88 ± 0.23
1.59 ± 0.26
2.09 ± 0.22
2.83 ± 0.19
2.42 ± 0.21
4.4. Inhibition of topoisomerase II activity
The three most potent cytotoxic m-AMSA analogues,
i.e. derivatives IIb, IIIa and IIIb were assayed for their
ability to inhibit the activity of topoisomerase II. We
performed the relaxation test of supercoiled DNA of PM2
phage and electrophoresis on agarose gel. In this cell-free
assay, 0.125 µg of supercoiled DNA was incubated with
increasing concentrations (from 10–160 µM) of drug plus
purified topoisomerase II from Drosophila melanogaster
embryos. Figure 4 reports results from an electrophoresis
gel where each of the three m-AMSA analogues was
tested at a lower concentration (10 µM).
a Hela cells were incubated for 3 h in the presence of m-AMSA
analogues (10 µM) and then processed for potassium–SDS precipi-
tation assay [24]. The cross-linking coefficient is the ratio between
the fraction of DNA precipitated together with proteins for treated
and control samples. b HeLa cells, treated for 3 h with m-AMSA
analogues were then submitted to neutral elution [25]. Elution rate
values were calculated as described in Experimental protocols.
Statistical evaluations were carried out by the F test of variance
(P < 0.01 compared with control).
Noticeably, the IC50 values of both compounds IIIa
and IIIb were significantly lower than those of the
reference compound, m-AMSA.
Lane 1 shows the electrophoretic pattern of PM2 DNA,
where the two bands correspond to the relaxed (slower)
and supercoiled (faster) forms, respectively. Lane 2
shows the effect of topoisomerase II activity, where the
supercoiled form completely disappeared. Lanes 3–5
show the inhibition of topoisomerase II induced by
compounds IIb, IIIa and IIIb, respectively. There was a
similar inhibition of topoisomerase II activity by each of
the three derivatives since the bands corresponding to the
supercoiled form of PM2 DNA were comparable to those
of the control sample.
4.3. Detection of DNA damage
4.3.1. DNA–protein cross-links
HeLa cells labelled with 3H-thymidine were incubated
for 3 h with the analogues of amsacrine (10 µM) and then
processed for the protein precipitation assay as already
described [24]. Table II shows the data concerning the
three most potent cytotoxic m-AMSA analogues (IIb,
IIIa and IIIb). m-AMSA was tested in the same experi-
mental conditions as reference. Treatment with all tested
The figure also displays the migration pattern obtained
from experiments with m-AMSA (10 µM), used in the
same experimental conditions as a reference (lane 6).
3
compounds significantly increased the amount of H-
DNA precipitated together with proteins in comparison
with untreated controls. The cross-linking coefficients
were all similar to those of m-AMSA and for compound
IIIb the value was even slightly higher. Dose–response
relationships were observed for the formation of DPCs
after exposure to the tested compounds (data not shown).
4.5. Antitumour activity
The antitumour activity of m-AMSA analogues against
solid tumour was studied in Ehrlich carcinoma-bearing
mice. We only tested the derivatives that exhibited better
antiproliferative activity against cell lines in culture, i.e.
derivatives IIb, IIIa and IIIb, amsacrine was used as
reference compound. The results of these in vivo experi-
ments are reported in table III. Chemotherapy of solid-
tumour-bearing mice with 10 mg.kg–1 of derivatives IIb,
IIIa and IIIb resulted in significant tumour volume
reduction. In particular compound IIIb appeared approxi-
mately as potent as amsacrine, inducing about 40%
inhibition in tumour growth. However, chemotherapy
with amsacrine appeared more toxic toward the host,
killing one third of treated animals, while mice inoculated
with compound IIIb rapidly recovered lost body weight.
4.3.2. Double-strand breaks
The induction of double-strand breaks (DSBs) into
DNA of HeLa cells was assayed by means of neutral
elution. The rate of elution depends on the molecular size
of polynucleotide chains. Undamaged double helix DNA
eluted slowly and when DSBs were introduced into DNA
elution, kinetics increased [25]. Table II summarises the
results when performed with the three most potent
cytotoxic m-AMSA analogues (IIb, IIIa and IIIb). Also
in these experiments, m-AMSA was used as reference.
Treatment of cells for 3 h in complete growth medium in
the presence of each agent (15 µM) yielded a substantial
number of breaks and the frequency of DSB clearly

M.G. Ferlin et al. / Eur. J. Med. Chem. 35 (2000) 827–837
Table III. Antitumour activity against Ehrlich carcinoma*.
831
Compound
Average tumour
weight
Inhibition of tumour
growth
(g ± SE)
(% ± SE)
Control
IIb
IIIa
IIIb
m-AMSA
0.097 ± 0.013
0.067 ± 0.02a
0.064 ± 0.020b
0.059 ± 0.013a
0.043 ± 0.01a
30.48 ± 9.1
34.16 ± 10.6
38.87 ± 8.5
45.52 ± 10.6
* Mice bearing Ehrlich carcinoma were treated intraperitoneally
with 10 mg.kg–1 m-AMSA analogues as described in Experimental
protocols. Average weight in grams was measured on day 7.
Statistical evaluations were performed by F test of variance. SE =
standard error. a P < 0.01. b P < 0.1.
[7, 18, 19]. Because of the similar electronic effects of
both a ring nitrogen and a nitro substituent, the corre-
sponding aza-bioisosters were also synthesised and stud-
ied, proving effective [7], even if the presence of a –N=
function did not always provide an improvement in their
biological activity [18]. As part of our program on
polyetherocyclic compounds showing antitopoisomerase
activity, we planned the synthesis of an m-AMSA aza-
analogue having bisubstituted the tricyclic nucleus (com-
pound IIIb). As synthetic intermediate we obtained the
corresponding phenoxy-derivative (IIb). With the aim of
investigating the effect of this 1-azaisosterism in the
2-methoxy-6-chloro-acridinic nucleus, we studied the
biological activity of aza analogues Ib, IIb and IIIb
compared with the corresponding acridinic derivatives
(compounds Ia, IIa and IIIa). All studies were performed
considering m-AMSA as reference compound.
Studying DNA binding in vitro of the tested com-
pounds by fluorescence titrations we observed that acri-
dine derivatives showed a fluorescence behaviour very
similar to that of m-AMSA; conversely, the 1-aza-
acridines exhibited a strong binding affinity towards
DNA. Their DNA-binding properties are essentially simi-
lar, in spite of the noticeable differences in the nature of
the 9-substituent. The presence of two hydrophobic
groups in the acridine skeleton, like in the 6,9-
dichlorosubstituted aza-analogue Ib, highly enhances the
affinity of the tricyclic ring toward DNA.
Various in vitro methods were used to evaluate cellular
damage following exposure to the tested AMSA-
analogues. The dye exclusion test was carried out as an
indicator of cell membrane integrity, while the incorpo-
ration of radioactive DNA precursors was an index
correlating inhibition of DNA synthesis with cell death
after drug exposure. Finally, as these kinds of biological
tests evaluate metabolic arrest, a clonogenic assay was
carried out to assess reproductive cellular death. The
Figure 4. Inhibition of topoisomerase II activity studied by the
relaxation assay. PM2 DNA was incubated in the presence of
topoisomerase II and different concentrations of m-AMSA
analogues. Agarose gel electrophoresis of the samples is shown.
Lane 1: PM2 DNA alone, which migrates as two bands; lane 2:
PM2 DNA incubated in the presence of topoisomerase II,
showing only the relaxed form (the slower band); lane 3–5:
PM2 DNA incubated in the presence of topoisomerase II and
10 µM compounds IIb, IIIa and IIIb, respectively. Lane 6:
PM2 DNA incubated in the presence of topoisomerase II and
m-AMSA used as reference (10 µM). The arrow indicates gel
mobility.
5. Discussion
A large number of the 9-anilinoacridine class of
antitumour drugs have been prepared and tested for in
vivo activity [7–9]. Among them, some nitro m-AMSA
congeners have shown good tumour inhibitory activity

832
M.G. Ferlin et al. / Eur. J. Med. Chem. 35 (2000) 827–837
Table IV. Analysis of compounds Ia–IIIb.
mammalian cells where DNA breaks re-seal with a
mismatch of the DNA strands after drug removal [26].
We found a strong correlation between the ability of these
m-AMSA analogues to induce topoisomerase II-mediated
strand breaks in intact HeLa cells and cytotoxicity, thus
suggesting that DNA topoisomerase is the intracellular
target for these compounds.
Performing some experiments in vitro with topoi-
somerase II from embryos of Drosophila melanogaster,
we observed that derivatives IIb, IIIa and IIIb inhibit the
activity of this enzyme at very low concentrations.
Keeping in mind the results regarding the formation of
non-covalent complexes with calf thymus DNA, the data
obtained on purified enzyme suggest that, for this class of
compounds, a relationship between the capacity to inhibit
topoisomerase activity and DNA binding properties does
not exist.
All the data regarding DNA damage suggest that the
side chain on the 9 position of the acridine skeleton is
essential for biological activity with little difference
between the anilino and phenoxy rings. In all probability,
these compounds also bind by intercalation of the acri-
dine or aza-acridine chromophore between the base pairs
with the side chain lodging in the minor groove and may
interact with enzymes such as topoisomerase as already
described for the acridinylamino-sulfonanilides [27].
Finally, to detect the antitumour properties of m-AMSA
analogues, we tested them against a solid tumour in mice,
and then compared with those of m-AMSA. Our data
show that compound IIIb was similarly effective against
murine tumour as m-AMSA, but was much less toxic
toward the host, and this feature is regarded as extremely
important among potential antitumour drugs.
b
Compound
Formula
C14H9NOCl2
C13H8N2OCl2
C20H14NO2Cl
C19H13N2O2Cl
Analysisa
pKa
6.02
7.55
5.54
5.88
Ia
Ib
Iia
Iib
IIIa
IIIb
C, H, N, Cl
C, H, N, Cl
C, H, N, Cl
C, H, N, Cl
C22H20N3O4Cl S·HCl C, H, N, Cl, S 5.75
C21H19N4O4Cl S·HCl C, H, N, Cl, S 4.51
a Analyses for elements indicated were within ± 0.4% of the
theoretical figures for the formula quoted. b Ionisation constants
were measured as detailed in ref. [23]. pKa of m-AMSA = 7.19
[23].
results suggested that compounds IIb, IIIa and IIIb were
the most potent derivatives, provoking strong antiprolif-
erative effects. In particular, for compounds IIIa and IIIb
the IC50 values were similar or significantly lower than
those of the reference compound, m-AMSA. These re-
sults, combined with those regarding DNA affinity, show
that a correlation between DNA–drug binding and drug
antiproliferative activity could not be detected. In fact,
compound Ib, which appears as a strong DNA-binding
ligand, was poorly effective in all biological tests. More-
over, it is well-known that m-AMSA forms weak DNA
intercalation complexes [21].
Previous studies have suggested that pKa values for
substituted acridines are an important determinant of
activity [16]. In particular it has been pointed out that
derivatives with high pKa values do not provide antipro-
liferative activity [18]. It has also been seen that the
introduction of a 1-aza function in the acridinic nucleus
provides very little base weakening effect [18]. On the
basis of pKa values detected (table IV), we could suppose
that the neighbouring presence of the 1-aza function and
the 2-methoxylic group in the pyridinic ring, as in
derivative IIIb, lowers the pKa values, presumably due to
the ready formation of a resonant cation stabilising the
charge. It has usually been considered that the cation of
chemotherapeutic acridines provides observed biological
activity, thus more highly ionised acridines normally
prove more effective.
6. Experimental protocols
6.1. Chemistry
Melting points were determined on a Gallenkamp
MFB 595010M/B capillary melting point apparatus and
are not corrected. Infrared spectra were recorded on a
Perkin-Elmer 1760 FTIR spectrometer as potassium bro-
1
To investigate the mechanism of action of the new
derivatives, we studied the damage induced in DNA in
HeLa cells in vivo. Using the potassium–SDS protein
precipitation assay, we observed that derivatives IIb, IIIa
and IIIb are capable of forming covalent DPC, similarly
to m-AMSA. Neutral elution studies have given evidence
that derivatives IIb, IIIa and IIIb induce DSB to an
equivalent extent. However, DSB disappeared quickly
after drug removal. This phenomenon has been previ-
ously described for topoisomerase II inhibitors acting in
mide pressed disks. Values are expressed in cm–1. H-
NMR spectra were recorded on a Varian Gemini appara-
tus (200 MHz) using the indicated solvents: chemical
shifts are reported in δ (ppm) downfield from tetrameth-
ylsilane as internal reference. J values are given in hertz
(Hz). In the case of multiplets, the chemical shift quoted
is measured from the approximate centre. Integrals cor-
responded satisfactorily to those expected on the basis of
compound structure. Elemental analyses were performed
in the Microanalytical Laboratory of the Department of

M.G. Ferlin et al. / Eur. J. Med. Chem. 35 (2000) 827–837
833
Pharmaceutical Sciences of the University of Padova,
6.1.2. General procedure
for the indirect synthesis of anilino-derivatives: route B
using a Perkin-Elmer Elemental Analyser Model 240B.
Results fell within the range ± 0.4% with respect to
calculated values. Column flash chromatography was
carried out on Merck silica gel (250–400 mesh ASTM)
and reactions were monitored by analytical thin-layer
chromatography (TLC) using Merck silica gel 60 F-254
glass plates. Solutions were concentrated in a rotary
evaporator under reduced pressure. Starting compounds
Ia,b and m-methoxy-p-nitro-aniline were purchased from
Aldrich Chimica and Janssen Chimica (now Acros),
respectively.
6.1.2.1. Preparation of phenoxy derivatives
A solution of phenol and NaOH was stirred at
120–130 °C for 1 h. The methoxy-chloro-substituted start-
ing compound was then added and the reaction mixture
was stirred at the above temperature for about 2 h. After
incomplete cooling, the melted mixture was poured into a
vigorously stirred 2 N aqueous solution of NaOH to give
a yellow suspension which was collected and dried.
Purification by crystallisation followed.
6.1.2.2. 2-Methoxy-6-chloro-9-phenoxy-acridine IIa
Prepared from Ia in good yields. R.f. 0.92 (ethylacetate/
n-hexane 1:1). 1H-NMR (CD3OD) δ: 3.61 (s, 3H, OCH3),
6.87 (m, 3H), 7.07 (m, 1H), 7.18 (d, 1H, J = 2.9 Hz,
HC-1), 7.32 (m, 2H), 7.5 (m, 2H), 8.07 (m, 3H), 8.14 (d,
1H, J = 1.6 Hz, HC-5).
6.1.1. General procedure
for the direct synthesis of anilino derivatives: route A
A solution obtained dissolving equimolar amounts of
the starting tricyclic compound and o-methoxy-p-
methansolfonamido-aniline in 80–150 mL of absolute
ethanol, was refluxed for 2.5–3 h until starting materials
disappeared (TLC, ethylacetate/n-hexane 7:3). On stand-
ing and cooling a red–orange precipitate formed which
was filtered and dried (40–50% yields). Purification by
repeated crystallisation afforded pure chlorides.
6.1.2.3. 2-Methoxy-7-chloro-10-
phenoxy-benzo[b] [1, 5]naphthyridine IIb
Prepared from compound Ib as yellow powder. Yield
83%. M.p. 185–187 °C (benzene/petroleum ether
1
40–60 °C). R.f. 0.92 (ethylacetate/ n-hexane 1:1). H-
NMR (CD3OD) δ: 3.52 (s, 3H, OCH3), 6.99 (m, 2H),
7.05 (d, 1H, J = 10 Hz, HC-3), 7.31 (m, 2H), 7.38 (d, 1H,
J = 3.3 Hz), 7.69 (dd, 1H, J = 2.2 and 10 Hz), 8.27 (d,
1H, J = 1.9 Hz), 8.32 (d, 1H, J = 9.2 Hz), 8.39 (d, 1H, J =
9.4 Hz).
6.1.1.1. 2-Methoxy-6-chloro-9-(2′-methoxy-4′-
methansolfonamido-aniline)-acridine chloride IIIa
Prepared from compound Ia following the general
procedure A as red microcrystals. Yield 50%. M.p.
307–309 °C dec. (absolute ethanol) lit. [16] 300–303 °C.
6.1.2.4. Preparation of 9-aniline-derivatives
1
R.f. 0.56. H-NMR (CD3OD) δ: 3.09 (s, 3H, CH3), 3.61
A mixture of phenoxy derivative and phenol was
melted by heating at 80 °C, and then a stoichiometric
amount of 2-methoxy-4-methansolfonamido-aniline chlo-
ride was added. Gradually the temperature of the reaction
mixture was then raised to 130–140 °C and maintained
for about 3 h. After incomplete cooling the melted mix-
ture was poured into vigorously stirred anhydrous ether
giving a flaky dark red suspension. After standing and
cooling at –20 °C for 2 days a microcrystalline product
separated which was collected and dried (55–65%). This
raw material was crystallised from absolute ethanol to
afford pure chlorides IIIa,b.
(s, 3H, 3′-OCH3), 3.72 (s, 3H, 2-OCH3), 7.05 (m, 2H,
HC-3′ and HC-5′), 7.42 (dd, 1H, J = 2.2 and 9.5 Hz,
HC-3), 7.49 (d, 1H, J = 8.9 Hz, HC-6′), 7.56 (d, 1H, J =
2.5 Hz, HC-1), 7.68 (dd, 1H, J = 2.6 and 9.3 Hz, HC-7),
7.83 (d, 1H, J = 9.7 Hz, HC-4), 7.88 (d, 1H, 1.6 Hz,
HC-5), 8.19 (d, 1H, J = 9.6 Hz, HC-8).
6.1.1.2. 2-Methoxy-7-chloro-
10-(2′-methoxy-4′-methansolfonamido-
aniline)-benzo[b] [1, 5]naphthyridine chloride IIIb
Prepared from compound Ib following the general
procedure A as orange crystals. Yield 40%. M.p.
270–273 °C (absolute ethanol). R.f. 0.61. 1H-NMR
(CD3OD), δ: 3.09 (s, 3H, CH3), 3.66 (s, 3H, 2-OCH3),
4.01 (s, 3H, 3′-OCH3), 6.03 (dd, 1H, J = 2.4 and 8.3 Hz,
HC-5′), 7.08 (d, 1H, J = 2.2 Hz, HC-3′), 7.38 (dd, 1H,
J = 2.1 and 9.5 Hz, HC-9), 7.49 (d, 1H, J = 8.5 Hz,
HC-6′), 7.54 (d, 1H, J = 9.2 Hz, HC-3), 7.9 (m, 2H,
HC-2′ and HC-8), 8.19 (d, 1H, J = 9.2 Hz, HC-4).
6.1.3. N-(2-Methoxy-4-nitro-benzen)-ethylcarbamate
1.4 mL (12 mM) of ethylchloroformiate was added to a
solution of commercial 2-methoxy-4-nitro-aniline (2 g,
12 mM) in 50 mL dry THF and the mixture was refluxed
until the starting compound disappeared at TLC analysis
(ethylacetate/n-hexane 7:3). The yellow solution was
evaporated to dryness giving a crystalline yellow product
(95% yield). C9H12O5N2. M.p. 154–156 °C. R.f. 0.45

834
M.G. Ferlin et al. / Eur. J. Med. Chem. 35 (2000) 827–837
1
(n-hexane/ethylacetate 5:5). H-NMR (deutero-acetone)
δ: 1.29 (t, 3H, J = 7.1 Hz, CH3), 4.06 (s, 3H, OCH3), 4.23
(q, 2H, J = 7.1 Hz, CH2), 7.82 (d, 1H, J3,5 = 2.6 Hz,
HC-3), 7.92 (dd, 1H, J5,3 = 2.6 and J5,6 = 9.2 Hz, HC-5),
8.21 (sa, 1H, NH), 8.31 (d, 1H, J6,5 = 9.2 Hz, HC-6).
6.2.2. Interaction with DNA in vitro
6.2.2.1. UV determinations
Absorption spectra were recorded using a Kontron
UVIKON-930 UV-Visible spectrophotometer. Increasing
micro-volumes of a DNA solution (1 mg.mL–1), having
the same above-mentioned drug concentrations, were
added to a fixed volume of an aqueous drug solution
(3.0–5.0 × 10–5 M). The additions were performed up to
a high DNA/drug ratio. An equal amount of DNA was
also added to the reference cell.
6.1.3.1. N-(2-Methoxy-4-amino-benzen)-ethylcarbamate
Easily air oxidable intermediate compound. Yield 90%.
R.f. 0.6 (n-hexane/ethylacetate 5:5).
6.1.3.2. N-(2-Methoxy-4-
methansolfonamido-benzen)-ethylcarbamate
C11H16N2O5. Yield 75%. M.p. 139–141 °C. R.f. 0.75
1
(ethylacetate/n-hexane 7:3). H-NMR (deutero-acetone)
6.2.2.2. Fluorescence determinations
δ: 1.26 (t, 3H, J = 7 Hz, CH3), 2.96 (s, 3H, SO2CH3),
3.88 (s, 3H, OCH3), 4.16 (q, 2H, J = 7 Hz, CH2), 6.9 (dd,
1H, J5,3 = 2.5 Hz and J5,6 = 8.7 Hz, HC-5), 7.04 (d, 1H,
3,5 = 2.5 Hz, HC-3), 7.58 (sa, 1H, uretanic NH), 7.94 (d,
1H, J6,5 = 8.6 Hz, HC-6), 8.4 (sa, 1H, amidic NH).
The formation of a molecular complex with DNA was
investigated by fluorimetric measurements according to
Gimenez-Arnau et al. [23]. The experiments were carried
out by means of a spectrophotofluorimeter, Kontron
Instruments model SFM-25. Molar concentrations of
aqueous solutions of calf thymus DNA were determined
spectrophotometrically using 6600 as extinction coeffi-
cient (M–1.cm–1) at 260 nm and expressed in base pairs.
Fluorescence emission spectra were performed using an
excitation wavelength of 380 nm and fluorescence inten-
sity values were recorded at the emission maximum of
the tested compounds. Spectra were obtained by adding
an aqueous drug solution (3.0–5.0 × 10–6 M), with micro-
volumes of a DNA solution (1 mg.mL–1) at the same drug
concentration. During the titrations we recorded the
values of F, the fluorescence intensity of the ligand
quenched by DNA, and F0, the fluorescence intensity of
free ligand. We then calculated F/F0 as relative fluores-
cence.
J
6.1.3.3. 2-Methoxy-4-methansolfonamido-aniline
A solution of the uretanic derivative in aqueous 20%
NaOH was refluxed for a useful period of time (TLC).
After cooling at room temperature the reaction mixture
was extracted with diethylic ether and the organic layer
was separated, dried over Na2SO4, and evaporated in
vacuo to dryness to give the aniline compound [8] in 94%
yield. Crystallisation from ethanol afforded pure com-
pound.
6.1.3.4. 2-Methoxy-4-
methansolfonamido-aniline chloride
An ethanolic solution of aniline derivative was satu-
rated with HCl dry gas and after a night at –15 °C, a
crystalline yellow product was filtered and dried.
C8H12O3N2S.HCl.
6.2.3. Cell cultures
HeLa cells (kindly provided by Prof. F. Majone, Dept.
of Biology of Padua University, Italy) were grown as
monolayers in Nutrient Mixture F12 Ham medium (Sigma
Chemical Co.) supplemented with 10% foetal calf serum
(Biological Industries, Kibbutz Beth Haemek, Israel).
Trypsin (0.25%, Boehringer Mannheim) was routinely
used for subcultures.
6.2. Biological methods
6.2.1. Chemicals
Calf thymus DNA (Cat. D 1501) and tetrapropylam-
monium hydroxide (1 M aqueous solution) were obtained
from Sigma Chemical Co, St Louis, MO, USA.
Compounds were dissolved in dimethyl sulfoxide
(DMSO) (4.5 mM) and the solutions were stored at
–20 °C. Just before the experiments, a calculated amount
of drug solution was added to phosphate buffer saline
(PBS) or to the growth medium to a final solvent
concentration of 0.5%, which had no discernible effect on
cell killing. 3H-Thymidine (4.77 Tbq.mM–1) was ob-
tained from Amersham International Inc., UK. Proteinase
K was obtained from Boehringer Mannheim GmbH
(Germany).
HL-60 cells were grown in RPMI-1640 medium (Whit-
taker Bioproduct, Walkersville, MD, USA) containing
5% foetal calf serum (Biological Industries, Kibbutz Beth
Haemek, Israel), supplemented with 25 mM HEPES
buffer and l-glutamine. Both HeLa and HL60 media were
supplemented with antibiotics, penicillin (50 units.mL–1)
and streptomycin (50 µg.mL–1), and cell growth was
accomplished at 37 °C in a 5% carbon dioxide atmo-
sphere.

M.G. Ferlin et al. / Eur. J. Med. Chem. 35 (2000) 827–837
835
6.2.3.1. Clonogenic survival
6.2.5. DNA damage
1.5–2 × 105 HeLa cells were seeded in 60 mm Petri
dishes in growth medium (4 mL). After 24 h the medium
was removed and replaced with a fresh one containing the
compound to be studied at the appropriate concentrations.
Cells were then incubated for 3 h. Triplicate cultures were
established for each treatment. The dishes were then
washed with PBS and 200 cells from each treated and
untreated culture were then seeded in complete growth
medium. After 7 days incubation colonies were stained
and counted, discarding colonies with less than 50 cells.
The efficiency of clonal growth, that is the ratio between
the number of colonies formed and the number of cells
seeded, was then calculated and used to normalise the
cytotoxicity induced by the drugs. In each experiment,
triplicate samples of untreated cells were plated as
controls and plating efficiency ranged from 88–92%.
6.2.5.1. Detection of double-strand breaks
DNA double-strand breaks (DSB) were detected by
neutral elution carried out according to Kohn [25].
HeLa cells in exponential growth were labelled by
3
overnight incubation in the presence of H-thymidine
(7.4 KBq.mL–1). The radioactive medium was removed
and replaced by a fresh one containing the compound to
be studied or 0.5% dimethyl sulfoxide for the controls. In
both cases, the cells were incubated for 3 h in the dark.
The cells were then washed with PBS.
3
About 0.5–1.0 × 106 treated H-cells were deposited
on a polycarbonate filter (pores 2 µm in diameter, Nucle-
opore Corp. Pleasanton, CA, USA) in a Swinnex-25 filter
holder (Millipore Corp. Bedford, MA, USA) and imme-
diately lysed with 2% sodium dodecylsulfate (SDS),
0.1 M glycine, 0.025 M EDTA, pH 9.6, (5 mL). Thereaf-
ter, the solution was allowed to flow through by gravity.
2 mL of the same solution containing 0.5 mg.mL–1 of
proteinase K was then gently poured onto the filter,
followed by 40 mL of the eluting solution, (EDTA0.02 M,
SDS 0.1%, and tetrapropylammonium hydroxide pH 9.6).
The elution was carried out by a Gilson Minipuls peri-
staltic pump, at a flow of 0.03–0.04 mL.min–1. The
fractions were collected with a Gilson fraction collector
(approximately 3.5 mL per fraction) and the radioactivity
in each fraction was then determined.
6.2.3.2. Non-clonogenic assay
Cytotoxicity against human leukaemia HL-60 cells
was studied using the trypan blue dye exclusion test [28].
Cells at a concentration of 2 × 105.mL–1 were incubated
for 3 h in the presence of different concentrations of the
compound to be tested. Cells were then incubated for
4 min with 0.25% trypan blue (Sigma Chemical Co, St
Louis, MO, USA) and 5% foetal calf serum. Viable cells
were identified by their ability to exclude dye, whereas
the dye diffuses into non-viable cells. At least 100 cells
were counted for each experimental point recorded. The
controls were triplicate samples of untreated cells and
their viability ranged from 92–98%.
The results obtained with the neutral elution assay
were expressed as K values according to the formula
[30]:
V
K =
− ln r
͑ ͒
6.2.4. DNA synthesis in Ehrlich cells
Ehrlich ascite tumour cells (Lettrè strain, from Heidel-
berg) were routinely transferred by injecting intraperito-
neally 2 × 106 cells per animal into NCL mice. The
tumour cells, collected on the 6–7th day after transplant,
were processed as already described [29].
where K is the average elution rate constant of DNA, r is
the fraction of DNA retained on filter and V is the eluted
volume. The formula reflects the assumption of a first-
order kinetics for DNA elution, as a first approximation
[25].
2 × 107 cells.mL–1 in Hank’s solution containing the
compound to be studied were incubated at 37 °C for
60 min. 3H-thymidine (40 KBq.mL–1), in a small volume
of the same medium, was then added and the cells were
further incubated at 37 °C for 30 min. The acid-insoluble
fraction was precipitated by adding 5% ice-cold trichlo-
roacetic acid and then filtered through Whatman GF/C
filters. After several washings with cold 1% trichloroace-
tic acid, the filters were dried and counted. The results
were calculated as the percentage of radioactivity incor-
porated into the DNA of untreated control cells (about
3–6 MBq; quadruplicate samples). Filtrations were car-
ried out by a Sample Manifold apparatus (Millipore
Corp., Bedford, MA, USA).
6.2.5.2. Potassium–SDS precipitation assay
Precipitation of DNA covalently associated with pro-
teins was performed as described previously [24]. Briefly,
HeLa cells in exponential growth were labelled by
3
overnight incubation in the presence of H-thymidine as
described above for filter elution. Cells (1–2 × 106
cells.mL–1, 0.1 mL aliquots) were incubated for 3 h in the
dark with the tested compounds, washed twice with PBS
and then lysed with 0.1 mL of SDS 2%, 1 mM EDTA,
pH 7.5. The samples were mixed energetically for 10 s,
warmed for 10 min at 60 °C. 0.5 mL of 200 mM KCl and
20 mM Tris-HCl, pH 7.5, were added and the mixtures
were cooled in an ice-bath for 5 min. The precipitates

836
M.G. Ferlin et al. / Eur. J. Med. Chem. 35 (2000) 827–837
formed were collected at 4 °C by an Eppendorf centrifuge
and suspended in 1 mL of 100 mM KCl. These mixtures
were heated again and the above procedure repeated.
Finally, the pellets were dissolved in 1 mL aliquots of
water and the solutions and corresponding supernatants
were counted. The results obtained with protein precipi-
tation were expressed in terms of cross-linking coefficient
(CC), a parameter which is proportional to the number of
DNA–protein cross-links (DPC). It is defined as follows:
Ft
6.2.8. Antitumour activity
Antitumour tests were carried out as described [32]
using groups of 10 Swiss mice (20 ± 3 g body weight)
and Ehrlich carcinoma. The tumour was maintained by
weekly transplantation into recipient mice. For the ex-
periments, about 5 × 108 tumour cells were implanted
subcutaneously in the right axillary region of mice. This
transplant yielded the development of a tumour in the
insertion place. Even 10 days after transplant, these
tumours appeared to be solid and well delimited, necrosis
and ascite free. The day after transplant, the drugs were
administered i.p. every day (10 mg.kg–1), for 5 days. At
day 7 after transplant the animals were sacrificed, the
tumours removed and their weight determined. The
control received the same amount of physiological solu-
tion.
CC =
Fc
where Ft and Fc represent the fractions of radioactivity
precipitated together with proteins in the treated and
control samples, respectively. Generally, in the control
cells the total radioactivity of a sample was about
5–6 MBq and only 0.4–0.7% of this amount was recov-
ered together with proteins in the precipitated fraction.
6.2.9. Statistical evaluations
Cell lethality (dye exclusion, clonal growth and thy-
midine uptake) was assayed in the presence of increasing
drug concentrations (triplicate samples). IC50 values were
calculated by probit analysis using the ꢀ2 test to check the
linearity of the thus obtained regressions.
Experimental data related to DNA damage in HeLa
cells and antitumour activity against Ehrlich carcinoma
were submitted to statistical analysis by the F test of
variance.
6.2.6. Inhibition of topoisomerase II activity
The inhibition of topoisomerase II activity was studied
using a purified enzyme from Drosophila melanogaster
embryos (USB, from Amersham Italia S.r.l.) [31]. Ali-
quots of 0.125 µg of PM2 DNA (from Boehringer Man-
nheim) were incubated for 15 min at 30 °C in the pres-
ence of two units of topoisomerase II (one unit is defined
as the activity capable of relaxing 0.3 µg of super-coiled
DNA) in the presence of reaction buffer (10 mM Tris-
HCl, pH 7.9; 50 mM NaCl; 50 mM KCl; 5 mM MgCl2;
0.1 mM EDTA; 15 µg.mL–1 bovine serum albumin (BSA);
1 mM ATP). Aliquots of 2 µL of a DMSO solution
containing tested compound (4.5 mM) were added in
order to reach final concentrations of 10, 20, 40 and
160 µM. A suitable amount of reaction buffer was then
added to each sample to reach the final volume of 20 µL.
The reaction was blocked by adding EDTA 7 mM (3 µL)
containing 0.77% of SDS. 2 µL of bromophenol blue
containing 15% glycerol were added to the samples and
electrophoresis was carried out on 0.7% agarose gel
containing TAE (40 mM Tris-sodium acetate, pH 8.2,
1 mM EDTA) for 90 min. The gel was stained for 1 h in
aqueous ethidium bromide 0.5 µg.mL–1 and then photo-
graphed using a Polaroid camera.
Acknowledgements
Part of this work has been supported by MURST.
References
[1] Chavalitshewinkoo P., Wilairat P., Gamage S.A., Denny W.A.,
Figgitt D.P., Ralph R., Antimicrob. Agents Chemother. 37 (1993)
403–406.
[2] Gamage S.A., Tepsiri N., Wilairat P., Wojcik S.J., Figgitt D.P.,
Ralph R., Denny W.A., J. Med. Chem. 37 (1994) 1486–1494.
[3] Mauel J., Denny W.A., Gamage S.A., Ransijn A., Wojcik S.J.,
Figgitt D.P., Ralph R., Antimicrob. Agents Chemother. 37 (1993)
991–996.
[4] Gamage S.A., Figgitt D.P., Wojcik S.J., Ralph R., Ransijn A.,
6.2.7. Radiochemical determinations
Mauel J., et al., J. Med. Chem. 40 (1997) 2634–2642.
Radioactivity measurements were performed using
Ultima Gold (Packard Instruments, Meriden, CT, USA)
as scintillation fluid. All determinations were carried out
by a Packard Tri-Carb 1900 TR spectrometer. Isotope
counting was accomplished automatically on the basis of
[5] Figgitt D.P., Denny W.A., Chavalitshewinkoo P., Wilairat P.,
Ralph R., Antimicrob. Agents Chemother. 36 (1992) 1644–1647.
[6] Atwell G.J., Cain B.F., Seelye R.N., J. Med. Chem. 15 (1972)
611–615.
[7] Cain B.F., Seelye R.N., Atwell G.J., J. Med. Chem. 17 (1974)
922–930.
3
quenching curves obtained using H-radioactivity stan-
[8] Cain B.F., Atwell G.J., Denny W.A., J. Med. Chem. 18 (1975)
dards.
1110–1117.

M.G. Ferlin et al. / Eur. J. Med. Chem. 35 (2000) 827–837
837
[9] Cain B.F., Atwell G.J., Denny W.A., J. Med. Chem. 19 (1976)
[21] Su T.-L., Chou T.-C., Kim J.Y., Huang J.-T., Ciszewska G.,
Ren W.-Y., Otter G.M., Sirotnak F.M., Watanabe K.A., J. Med
Chem. 38 (1995) 3226–3235.
772–777.
[10] Nelson E.M., Tewey K.M., Leu L.F., Proc. Natl. Acad. Sci. USA 81
(1984) 1361–1365.
[22] Duprè D.J., Robinson F.A., J. Chem. Soc. (1945) 549–551.
[11] Pommier Y., Zwelling L.A., Kao-Shan C.S., Wanng-Peng J., Bra-
[23] Gimenez-Arnau E., Missalidis S., Stevens M.F., Anti-Cancer Drug
dly M.O., Cancer Res. 45 (1985) 3143–3149.
Des. 13 (1998) 431–451.
[12] Cain B.F., Atwell G.L., Eur. J. Cancer 10 (1974) 539–549.
[13] Arlin Z., Cancer Treat. Rep. 67 (1983) 967–970.
[24] Zhitkovich A., Costa M., Carcinogenesis 13 (1992) 1485–1489.
[25] Kohn K.W., Pharmaco. Ther. 49 (1991) 55–77.
[14] Figgitt D.P., Denny W.A., Gamage S.A., Ralph R.K., Anti-Cancer
[26] Holm C., Covey J.M., Kerrigan D., Pommier Y., Cancer Res. 49
Drug Des. 9 (1994) 199–208.
(1989) 6365–6366.
[15] Baguley B.C., Denny W.A., Atwell G.L., Cain B.F., J. Med. Chem.
[27] Chen K.-X., Gresh N., Pullman B., Nucleic Acids Res. 16 (1988)
24 (1981) 170–177.
3061–3073.
[16] Denny W.A., Cain B.F., Atwell G.L., Hansch C., Panthananickal A.,
[28] Durkin W.J., Ghanta V.K., Balch C.M., Davis D.W., Hiramoto R.N.,
Leo A., J. Med. Chem. 25 (1982) 276–315.
Cancer Res. 39 (1979) 402–407.
[17] Gao H., Denny W.A., Garg R., Hansch C., Chem. Biol. Interact. 116
[29] Bordin F., Carlassare F., Conconi M.T., Capozzi A., Majone F.,
(1998) 157–180.
GuiottoA., Baccichetti F., Photochem. Photobiol. 55 (1992) 221–226.
[18] Denny W.A., Atwell G.J., Cain B.F., J. Med. Chem. 20 (1977)
[30] Noviello E., Aluigi M.G., Cimoli G., Rovini E., Mazzoni A.,
1242–1246.
Parodi S., De Sessa F., Russo P., Mutation Res. 311 (1994) 21–29.
[19] Denny W.A., Atwell G.J., Cain B.F., J. Med. Chem. 22 (1979)
[31] Tao-Shih H., Methods Enzymol. 199 (1983) 161–170.
1453–1460.
[32] E.O.R.T.C screening group, Eur. J. Cancer 8 (1972) 185–196.
[20] Besly D.M., Goldberg A.A., J. Am. Med. Assoc. 151 (1954)
2448–2455.