Antitumor polycyclic acridines. 7. Synthesis and biological properties of DNA affinic tetra- and pentacyclic acridines

Article abstract of DOI:10.1021/jm9909490

New synthetic routes to a series of tetra- and pentacyclic acridines related in structure to marine natural products are reported. The novel water-soluble agent dihydroindolizino[7,6,5-kl]acridinium chloride 14 has inhibitory activity in a panel of non-small-cell lung and breast tumor cell lines exceeding that of m-AMSA. The salt inhibited the release of minicircle products of kDNA confirming that disorganization of topoisomerase II partly underlies the activity of the compound. COMPARE analysis of the NCI mean graph profile of compound 14 at the GI₅₀ level corroborates this conclusion with Pearson correlation coefficients (> 0.6) to clinical agents of the topoisomerase II class: however, this correlation was not seen at the LC₅₀ level. The inhibitory action of 14 on Saccharomyces cerevisiae transfected with human topoisomerase II isoforms showed a 3-fold selectivity against the IIα isoform over the IIβ isoform. Unlike m-AMSA, 14 is not susceptible to P-glycoprotein-mediated drug efflux and retains activity in lung cells with derived resistance to the topoisomerase II inhibitor etoposide.

Full text of DOI:10.1021/jm9909490

J . Med. Chem. 2000, 43, 1563-1572
1563
An titu m or P olycyclic Acr id in es. 7.1 Syn th esis a n d Biologica l P r op er ties of DNA
Affin ic Tetr a - a n d P en ta cyclic Acr id in es
J ohnson Stanslas,†,§ Damien J . Hagan,† Michael J . Ellis,† Claire Turner,† J ames Carmichael,‡ Wynne Ward,‡
Timothy R. Hammonds,| and Malcolm F. G. Stevens*,†
Cancer Research Laboratories, University of Nottingham, University Park, Nottingham NG7 2RD, U.K.,
CRC Department of Clinical Oncology, University of Nottingham, City Hospital, Hucknall Road,
Nottingham NG5 1PB, U.K., and Department of Biochemistry, University of Leicester, Leicester LE1 7RH, U.K.
Received November 5, 1999
New synthetic routes to a series of tetra- and pentacyclic acridines related in structure to marine
natural products are reported. The novel water-soluble agent dihydroindolizino[7,6,5-kl]-
acridinium chloride 14 has inhibitory activity in a panel of non-small-cell lung and breast tumor
cell lines exceeding that of m-AMSA. The salt inhibited the release of minicircle products of
kDNA confirming that disorganization of topoisomerase II partly underlies the activity of the
compound. COMPARE analysis of the NCI mean graph profile of compound 14 at the GI50
level corroborates this conclusion with Pearson correlation coefficients (>0.6) to clinical agents
of the topoisomerase II class: however, this correlation was not seen at the LC50 level. The
inhibitory action of 14 on Saccharomyces cerevisiae transfected with human topoisomerase II
isoforms showed a 3-fold selectivity against the IIR isoform over the IIâ isoform. Unlike
m-AMSA, 14 is not susceptible to P-glycoprotein-mediated drug efflux and retains activity in
lung cells with derived resistance to the topoisomerase II inhibitor etoposide.
Ch a r t 1
In tr od u ction
The ease of synthesis, attractive coloration, and
crystallinity of acridine derivatives has long attracted
the attention of medicinal chemists.2 The basic tricyclic
framework can be decorated with appropriate substit-
uents to confer specificity against both prokaryotic and
eukaryotic targets which have given acridines a respect-
able reputation in the history of chemotherapy in the
20th century.3 In recent decades the focus of research
into the role of acridines in cancer treatment has
centered in Oceania where New Zealand scientists have
developed a range of derivatives, such as the prototypic
m-AMSA (1) (Chart 1), and variants, which are now
known to target the enzyme topoisomerase II (topo
II).4-6 Some of the most familiar drugs in the cancer
chemotherapeutic armamentarium are thought to act,
in part, by disorganizing the functions of this protein:7
however, common physical properties of basicity and
partial planarity give these agents avid DNA-binding
properties which may interplay at the single-strand,
duplex, or triplex levels. Thus, their true mode of action
as antitumor agents is difficult to define unequivocally.
To be considered for clinical evaluation, newly de-
signed agents of this class would need to demonstrate
unique biological properties, possibly by modulating
specific protein-DNA recognition, stabilizing DNA tri-
plexes in a sequence-directed sense, or selectively
engaging apoptotic pathways without eliciting overall
cytotoxicity. This is a tall order. In following our interest
in polycyclic acridines, we have taken some inspiration
from the natural world, particularly marine organisms
from two phyla, the tunicates and sponges, which
elaborate highly colored alkaloids based on the pyrido-
[2,3,4-kl]acridine nucleus 2. The rich diversity of struc-
tures8,9 presumably evolved as chemical warfare agents
to deter marine predators, and these agents exhibit a
* Corresponding author. Tel: 0115-9513404. Fax: 0115-9513412.
E-mail: malcolm.stevens@nottingham.ac.uk.
† School of Pharmaceutical Sciences, University of Nottingham.
‡ CRC Department of Clinical Oncology, University of Nottingham.
| University of Leicester.
§ Present address: Department of Biomedical Sciences, Faculty of
Medicine and Health Sciences, University Putra Malaysia, 43400
Serdang, Selangor, Malaysia.
10.1021/jm9909490 CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/23/2000

1564 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8
Stanslas et al.
Sch em e 1a
Sch em e 2a
a
Reagents and conditions: (a) NaN3, aq acetone, reflux; (b) for
a
Reagents and conditions: (a) reflux in benzene, 2 h; (b) Me2NH
7 only, R1CtCR2, toluene, 60 °C, 24 h; (c) diphenyl ether, 210-
230 °C; (d) NaH in DMF, 1,2,3-triazole and 6, 25 °C.
in THF, 30-40 °C, 5 days.
cycloaddition reaction with simple alkylalkynes, the
least sterically hindered triazole was the major product
as exemplified in the reaction of 5-cyanopent-1-yne and
7 which gave 9-[4-(3-cyanopropyl)-1,2,3-triazol-1-yl]acri-
dine (8b) (51%) and 9-[5-(3-cyanopropyl)-1,2,3-triazol-
1-yl]acridine (8c) (27%). However, the disadvantage of
the cycloaddition method is that chromatographic sepa-
ration of the mixture of isomeric triazoles is always
required.
In the present work, two new triazoles were prepared
from 7 and 1-dimethylaminoprop-2-yne: the products
were separated laboriously into 9-[4-(dimethylamino-
methyl)-1,2,3-triazol-1-yl]acridine (8j) (58%) and the
isomer 8k (22%). In our previous work we have shown
that 9-(1,2,3-triazol-1-yl)acridines can be prepared in a
regiospecific manner from 9-azidoacridine and reactive
methylenic compounds.14 In a variation of this process,
reaction of 7 and the chloroacetonyltriphenylphos-
phorane ylide 10 gave exclusively 9-[5-(chloromethyl)-
1,2,3-triazol-1-yl]acridine (13) in 67% yield. The mech-
anism of this cycloaddition reaction involves initial
formation of the tetrahedral adduct 11 and cyclization
to the oxaphosphetane intermediate 12, followed by
elimination of triphenylphosphine oxide (Scheme 2) in
the manner of a Wittig olefination reaction. Incubation
of 13 with dimethylamine at 30-40 °C in THF gave
access to pure 8k (72%) which was only the minor
isomer from the azide-alkyne cycloaddition route (see
above).
range of intriguing biological properties, including topo
II inhibition. Most recently, a series of new alkaloids,
styelsamines A-D (3a -d ), have been isolated from the
Indonesian Ascidian, Eusynstyela latericius, which in-
hibit human colon HCT-116 cells in vitro in the 1-100
µM range.10
We have concentrated on developing new synthetic
routes to derivatives of the isomeric pyrido[4,3,2-kl]-
acridine ring system 4. This tetracycle is found in nature
only as the alkaloid necatorone (5) and some analogues,
including dimers, elaborated by the toadstool Lactarius
necator,11 and apart from our own synthetic work (and
references therein)12-14 there is only one further re-
ported example of this ring system in the literature.15
As compounds of this type undoubtedly bind to DNA,16,17
we have set ourselves the challenging task of tracking
the sequence of events following the initial DNA en-
counter1 which eventually leads to the death of cells.
In this paper we survey routes developed to synthe-
size pyrido[4,3,2-kl]acridines and benzo-annealed modi-
fications, together with improvements over published
methods, and give the first account of some of their
biological properties.
Ch em istr y
9-Chloroacridine (6), a convenient starting point for
the synthesis of the target polycyclic acridines, was
converted efficiently to the azidoacridine 7 with sodium
azide in aqueous acetone.18 The azide participated in
cycloaddition reactions with 1-alkylalkynes to yield
regioisomeric 9-(triazol-1-yl)acridines 8a -i (Scheme
1).13 Alternatively, the unsubstituted 9-(1,2,3-triazol-1-
yl)acridine (8a ) can be prepared by reacting 9-chloro-
acridine with the anion of 1,2,3-triazole12 or employing
trimethylsilylacetylene in the cycloaddition reaction
followed by removal of the TMS protecting group.13 The
former route gave the highest yield despite the fact that
the required triazole was accompanied by the unwanted
9-(1,2,3-triazol-2-yl)acridine isomer. Generally, in the
The triazoles 8a -i have been thermolyzed in hot
diphenyl ether to generate reactive diradical or carbene
species which cyclized generally in good yields to 7H-
pyrido[4,3,2-kl]acridines 9a -i (Scheme 1).13 Exceptions
were the thermolyses of 8j and 8k which gave the new
tetracycles 9j and 9k in 60% and 58% yields, respec-
tively. An interesting variant of this synthesis is seen
in the different behavior of the regioisomeric chloropro-
pyltriazolylacridines 8l and 8m : predictably, isomer 8l
cyclized to 3-(3-chloropropyl)pyridoacridine 9l, whereas

Antitumor Polycyclic Acridines
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8 1565
Sch em e 3a
molysis of the nitrobenzotriazolylacridine 16d to the
2-nitroquinoacridine 17d was also conveniently effected
(74%) in refluxing triglyme, but because of the insolubil-
ity of the pentacyclic acridine in diphenyl ether, it could
be recovered efficiently from the latter thermolysis
medium simply by dilution and washing with hexane.
Catalytic hydrogenation of the nitroquinoacridine 17d
over a palladium-charcoal catalyst has given the cor-
responding amine 17e in only moderate yield (<60%).12
In the present work reduction of 17d with Raney
nickel-hydrazine hydrate in ethanol or tin(II) chloride
dihydrate in refluxing ethanol was no more efficient:
the optimum conditions (95% yield) were tin(II) chloride
in 12 M HCl. A more elegant approach to the amine
17e was considered feasible if a nitrenium ion (19) T
π-carbocation (19′) reactive species could be generated
from a suitable precursor such as the azidoaniline 18d :
a sequence nitroanilinoacridine 18a f hydroxylamine
18b f reactive species 19 f intermediate 20 f 17e
(Scheme 4) is typical of intramolecular cyclizations
developed especially by Abramovitch.21 Encouragingly,
reduction of nitroarenes with zinc dust in a mixture of
TFSA and TFA at 0 °C has been utilized by Ohta22 as
a method of generating such reactive intermediates. The
m-nitroanilinoacridine 18a was prepared from 6 and
m-nitroaniline and subjected to zinc dust/TFSA/TFA
degradation. None of the required amine 17e was
detected (TLC), and the only product isolated was the
m-aminoanilinoacridine 18c (70%). In an alternative
approach the azidoanilinoacridine 18d was decomposed
at 0 °C in a mixture of TFSA, TFAA, and TFA. A
vigorous evolution of nitrogen was observed, but only
minor amounts of the pentacyclic amine 17e, together
with the m-aminoanilinoacridine 18c (combined yield
< 30%), were formed. Although this type of reaction
often affords high yields of products,21 the poor outcome
in the present example can be explained by protonation
of the acridine in the highly acidic conditions which
would deactivate the acridinium system for nucleophilic
addition to the π-carbocation reactive intermediate.
a
Reagents and conditions: (a) diphenyl ether, 210-230 °C.
isomer 8m gives the deep maroon pentacyclic indolizino-
[7,6,5-kl]acridinium salt 14 (77%), presumably via in-
tramolecular alkylation by the unstable chloropropyl
intermediate 9m (Scheme 3).13 We have improved this
reaction to furnish 14 in 86% yield. Because the salt 14
was one of the most biologically interesting of the
polycyclic acridines examined (see later), a more con-
venient “one-pot” procedure starting from azidoacridine
7 and 5-chloropent-1-yne was sought which would take
advantage of the polar nature of 14 to facilitate a simple
separation from nonpolar contaminants by solvent
partitioning. After many attempts, a procedure involv-
ing initial incubation of 7 and 5-chloropent-1-yne in
diphenyl ether at 60 °C for 24 h (to effect synthesis of a
mixture of triazoles 8l and 8m ) and brief heating to
210-230 °C to complete cyclization to polycyclic acridines
followed by water extraction of polar materials gave only
a 6% yield of 14, together with much insoluble black
material. The latter was probably 1,2-bis(acridin-9-yl)-
diazene,18 the product of thermal degradation of the
9-azidoacridine unconsumed in the initial cycloaddition
step. Clearly, the “one pot” process is much less efficient
than the two-step procedure (above).
9-Anilinoacridines 15a -d , prepared from 9-chloro-
acridine (6) and o-phenylenediamines, underwent ni-
trosative cyclization to 9-(benzotriazolyl)acridines 16a-d
required as precursors to the pentacyclic acridines 17a -
d .12 An alternative synthesis of 16a has been reported
from 9-azidoacridine (7) and benzyne (47%).19 However,
we now report the most efficient route to 16a (69%), by
reacting 6 directly with the sodio derivative of benzo-
triazole in boiling DMF. Because of the ambident nature
of a substituted benzotriazolyl anion, this approach is
not appropriate for the synthesis of the substituted
benzotriazoles 16b-d . Thermolysis of the benzotriazoles
16a -d in boiling diphenyl ether (259 °C) afforded 8H-
quino[4,3,2-kl]acridines 17a -d in reasonable yields12
(Scheme 4), but entrainment with solvent necessitated
purification of products by vacuum sublimation. The
most convenient process for the preparative-scale de-
composition of 16a utilized boiling triethylene glycol
dimethyl ether (triglyme, bp 216 °C) or diethylene glycol
(bp 245 °C) as the thermolytic medium, despite the fact
that these temperatures are below those at which
decomposition is observed by differential scanning cal-
orimetry:12 the pentacyclic product 17a was collected in
87% and 93% yields, respectively, after diluting the
reaction medium with water. Formation of 17a from 16a
by the thermal nitrogen-extrusion route (above) was
more efficient than the corresponding photochemical
cyclization20 or radical cyclization of 9-(2-iodoanilino)-
acridine.12 Although flash vacuum pyrolysis of 16a at
590 °C gave a 83% yield of 17a on a 100-mg scale, the
process was not amenable to efficient scale-up. Ther-
The weakly basic aminoquinoacridine 17e (pKa
)
6.18)16 formed water-soluble methanesulfonic, ethane-
sulfonic, and tetrafluoroborate acid salts which were
more suitable for biological studies: protonation was
accompanied by a bathochromic shift of the long wave-
length absorption in the visible spectrum from 490 nm
in the base to 518-520 nm, confirming that (mono)-
protonation takes place on the quinoacridine ring
system, presumably at N-13.
Biologica l Resu lts a n d Discu ssion
In Vitr o Cytotoxicities of P olycyclic Acr id in es
in P a n els of Hu m a n Non -Sm a ll-Cell Lu n g a n d
Br ea st Cell Lin es. One of the ways of determining the
possible molecular target of novel compounds with
structural resemblances to agents with a known, or
assumed, mode of action is by testing the activity of new
agents in a panel of cancer cell lines expressing different
levels of the molecular target. Previous studies have
shown that the potency of topo II inhibitors, whether
topo II poisons or topo II catalytic inhibitors,23 is
dependent on the intracellular protein levels.
As a prelude to detailed investigation of the biological
properties of polycyclic acridines, their cytotoxicities

1566 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8
Stanslas et al.
Sch em e 4a
a
Reagents and conditions: (a) substituted aniline, MeOH, reflux; (b) NaNO2, 2 M HCl, 0 °C; (c) diphenyl ether, reflux, 2 h; (d)
benzotriazole, NaH, DMF, reflux; (e) for 17d only, SnCl2‚2H2O, 10 M HCl, 25 °C; (f) NaNO2, 2 M HCl, then NaN3; (g) TFSA/TFAA/TFA,
0 °C.
Ta ble 1. Inhibitory Activity of Polycyclic Acridines and Standard Agents in a Panel of Human NSCL Cancer Cell Linesa
IC50 values (µM)b of cell linesc
compd
NCI-H647d
A549e
NCI-H226d
NCI-H358d
NCI-H460e
NCI-H322d
mean IC50
m-AMSA, 1
9a
9b
9c
14
17a
17b
17c
17d
17e
etoposide
doxorubicin
0.07
2.20
2.92
1.68
0.12
0.54
0.80
5.30
0.95
0.08
0.08
<0.001
0.03
2.14
2.57
1.33
0.16
0.49
0.81
7.10
1.01
2.34
2.72
1.21
0.22
0.15
1.42
2.38
2.01
0.11
0.46
0.60
6.03
9.50
0.40
1.50
0.13
0.06
2.06
2.56
1.17
0.18
0.52
0.71
3.90
>10.0
0.07
0.10
0.01
0.21
1.86
2.27
1.53
0.17
0.26
2.02
2.57
1.49
0.16
0.54
0.83
>7.0
>8.4
0.16
0.61
<0.03
0.75
1.22
0.49
0.84
>10.0
>10.0
0.17
>10.0
>10.0
0.14
>10.0
0.08
0.16
0.004
0.24
0.008
1.59
0.03
a
b
7-day MTT assay. IC50 values are the mean of at least 3 separate determinations. c Cell lines are presented in order of decreasing
topo IIR content from left to right. The ranking of topo IIâ is: NCI-H226 > A549 > NCI-H358 > NCI-H322 > NCI-H460 > NCI-H647
d
(data from ref 24). Other molecular characteristics: mutant p53; e wild-type p53.
against the adenosquamous NCI-H647, a non-small-cell
lung (NSCL) line which is known to exhibit elevated
topo IIR levels,24 were assessed in 4-day MTT assays.
Potency rankings followed the general order: indolizino-
[7,6,5-kl]acridinium salt 14 > m-AMSA (1) > pentacy-
clic acridines 17 > tetracyclic acridines 9 > triazoles 8
(data not shown).
Some of the most active agents, together with refer-
ence compounds, were then evaluated in a larger panel
of NSCL (Table 1) and breast (Table 2) cancer cell lines
with varying levels of topo IIR and â isoforms. The cells
had been characterized for certain other molecular
targets (see footnotes to the tables).24,25 The results
showed that the NSCL cell lines were more resistant
than the breast cell lines, and there were no striking
differences in activity between representative tetracyclic
acridines 9a -c and the pentacyclic series 17a -e.
Results corroborate the rankings in the prescreen,
confirming that the indolizino[7,6,5-kl]acridinium salt
14 and 2-aminoquinoacridine 17e were the most potent

Antitumor Polycyclic Acridines
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8 1567
Ta ble 2. Inhibitory Activity of Polycyclic Acridines and Standard Agents in a Panel of Human Breast Carcinoma Cell Linesa
IC50 values (µM)b of cell linesc
compd
SKBr-3d
MCF-7wte
MDA-231f
T47Dg
ZR-75-1h
MDA-468i
mean IC50
m-AMSA, 1
9a
9b
9c
14
17a
17b
17c
17d
17e
etoposide
doxorubicin
0.06
0.58
0.42
3.45
0.12
0.15
0.45
1.03
0.12
0.02
0.30
0.02
0.075
0.83
0.27
3.79
0.15
0.27
0.18
1.67
0.55
0.05
0.085
0.001
0.14
1.73
0.96
2.05
0.19
0.54
0.55
5.45
>10
0.22
2.10
1.31
3.59
0.16
0.69
2.13
5.55
0.18
1.55
0.72
2.70
0.17
0.85
0.39
5.59
0.03
0.03
0.24
0.009
0.49
1.37
0.32
1.87
0.18
0.09
0.17
0.31
0.04
0.04
0.60
0.005
0.19
1.36
0.66
2.90
0.16
0.43
0.65
3.27
>3.5
0.11
0.26
0.008
>10
0.23
0.15
0.006
0.27
0.16
0.007
a
b
7-day MTT assay. IC50 values are the mean of at least 3 separate determinations. c Cell lines are presented in order of decreasing
topo IIR content from left to right. The ranking of topo IIâ content is: MCF-7wt > SKBr-3 > T47D > MDA468 > ZR-75-1 >MDA-231
d
(data from ref 25). Other molecular characteristics: ER-, amplified erbB2, mutant p53; e ER+, PgR+, wild-type p53; f ER-, EGFR+,
g
h
i
mutant p53;
ER+, mutant p53; ER+, PgR+, mutant p53; ER-, amplified EGFR, mutant p53.
Ta ble 3. Inhibitory Activity of Polycyclic Acridines and
of the polycyclic acridines across the two panels. These
agents were slightly more active than m-AMSA and
etoposide but considerably less cytotoxic than doxoru-
bicin. There were strong correlations between the
antitumor activity of doxorubicin and topo IIR isoform
expression in the NSCL lines (Spearman rank correla-
tion coefficient: r ) 0.829, p ) 0.042) and of m-AMSA
to topo IIR in the breast cell lines (r ) 0.943, p ) 0.005),
but there was no apparent correlation between activity
and topo IIR or â isoform expression for any of the new
polycyclic acridines in the in vitro assays.
The polycyclic acridines were also evaluated in the
NCI in vitro 60-cell line panel.26 An unusual feature of
the dose-response relationships of the most potent
agent 14 against all cell lines (mean GI50 of 0.09 µM)
was the extended concentration range (0.01-10 µM) at
which the compound elicited first growth inhibitory (GI)
and then cytocidal (LC) effects. This was especially
noticeable in the melanoma subpanel where cell lines
LOX IMVI, MALME-3M, SK-MEL-5, and UACC-62
gave LC50 values of 0.7-3 µM against a mean LC50 in
the full 60-cell panel of 25 µM.
In Vitr o Cytotoxicities of P olycyclic Acr id in es
in Resista n t NSCL a n d Br ea st Ca n cer Cell Lin es.
The antitumor activities of polycyclic acridines were
evaluated in cell lines with acquired resistance to known
topo II inhibitors. Results for new agents and reference
compounds against the NCI-H460 lung line, and the
etoposide-resistant counterpart H460pv824 in which topo
IIR is down-regulated, are shown in Table 3. H460pv8
cells not maintained in etoposide showed a resistance
factor (RF) to etoposide of 4, but when these cells were
maintained in 1 µM etoposide (H460pv8/eto) there was
a 2-fold increase in RF, and this variant line showed
increased resistance to the pentacyclic acridine 17e and
doxorubicin, but not to the indolizino[7,6,5-kl]acridinium
salt 14. These results suggest that topo IIR may well
be the intracellular target for the standard agent
doxorubicin, in accordance with published work,27 but
that the lead compound 14 is not operating as a topo II
inhibitor in these cell lines. m-AMSA was approximately
equiactive against both resistant cell lines, possibly
because of its ability to target the isoform topo IIâ which
is not down-regulated in these lines.24
Standard Agents against Etoposide-Resistant NCI-H460 Cellsa
IC50 values (µM)b
resistance factors
compd
NCI-H460 H460pv8 H460pv8/eto
c
d
m-AMSA, 1
9a
9b
9c
14
17a
17b
17c
0.06
2.06
2.56
1.17
0.18
0.52
0.71
3.90
0.07
0.10
0.01
0.20
1.10
2.06
1.12
0.13
0.32
0.49
3.77
0.05
0.39
0.004
0.04
1.07
1.09
1.02
0.10
0.34
0.49
6.47
0.12
0.74
0.02
3.3
0.5
0.8
0.8
0.7
0.7
0.7
0.7
0.8
4.0
0.4
0.6
0.5
0.4
0.8
0.5
0.7
0.7
1.3
1.7
7.6
2.0
17e
etoposide
doxorubicin
a
b
7-day MTT assay. IC50 values are the mean of 3 separate
determinations. c Resistance factor is defined as: (IC50 H460pv8)/
d
(IC50 NCI-H460). Resistance factor is defined as: (IC50 H460pv8/
eto)/(IC50 NCI-H460).
Ta ble 4. Activity of Polycyclic Acridines against MCF-7wt and
MCF-7/adr Cell Linesa
IC50 values (µM)b
compd
MCF-7wt
MCF-7/adr
resistance factorc
m-AMSA, 1
9b
14
17a
17b
17c
17e
0.075
0.27
0.15
0.27
0.18
1.67
0.05
1.165
0.61
0.165
0.63
1.21
6.31
15.5
2.2
1.1
2.4
6.8
3.8
6.6
0.355
a
b
7-day MTT assay. IC50 values are the mean of 3 separate
determinations. c Resistance factor is defined as: (IC50 MCF-7/
adr)/(IC50 MCF-7wt).
modulated resistance which confers the clinically sig-
nificant multidrug resistance (MDR) phenotype to tu-
mors. The antitumor activity of polycyclic acridines
against parental MCF-7wt cells and their MDR coun-
terparts MCF-7/adr, which were derived following pro-
longed exposure of MCF-7wt to adriamycin (doxorubi-
cin),28 is shown in Table 4. Significantly, the MCF-7/
adr cell line was cross-resistant to m-AMSA with a RF
of 15.5 and etoposide (RF ) 74) but not the acridinium
salt 14 (RF ) 1.1), confirming that the polar acridinium
salt is not a substrate for P-glycoprotein drug efflux.
Modest levels of cross-resistance (RF ) 2.2-6.8) were
observed with other neutral polycyclic acridines (Table
4).
Ideally, activity of a new agent considered for clinical
trial should not be subverted by classic P-glycoprotein-

1568 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8
Stanslas et al.
F igu r e 2. Inhibition of the DNA decatenation activity of
human topo IIR by compounds 1, 9b, 14, and 17e. Assays were
performed as described in the Experimental Section, and
results for each drug are plotted as the percentage of decat-
enated products detected when compared with drug-free
controls. The release of minicircle products of kDNA networks
is decreased in the presence of each drug.
F igu r e 1. Inhibition of the DNA decatenation activity of
human topo IIR by compounds 1, 9b, 14, and 17e. Assays were
performed as described in the Experimental Section. The
release of minicircle products of kDNA networks is decreased
in the presence of each drug. Product identities: T ) trimer;
D ) dimer; OC ) open circular DNA; CC ) closed circular
DNA.
In h ibition of Top o II. In a preliminary topo II
assay, compounds were evaluated as inhibitors of the
enzyme from HeLa cells which can relax supercoiled
pBR322. New agents were assayed in the concentration
range 0.3-100 µM, and compound activity was com-
pared to that of m-AMSA by determining the concentra-
tions at which relaxation of pBR322 was inhibited by
100%. The most potent compounds were the indolizino-
[7,6,5-kl]acridinium salt 14 and the methanesulfonic
acid salt of 2-aminoquinoacridine 17e which were equi-
active (IC100 3 µM). The 3-(3-cyanopropyl)pyridoacridine
9b, the most cytotoxic of the tetracyclic compounds, gave
an IC100 of 30 µM (cf. m-AMSA IC100 30 µM).
F igu r e 3. Induction of double-strand DNA cleavage by
compounds 1, 9b, 14, and 17e in the presence of human topo
IIR. Assays were performed as described in the Experimental
Section; each drug was present at 1 µM. DNA cleavage is
identified as an increase in the amount of linear DNA band
in a reaction. Quantitation of these results reveals 2-4 times
more DNA cleavage in drug-treated samples. D ) DNA only;
E ) enzyme only; D + L ) DNA plus a marker of linear
pBR322; OC ) open circular DNA; CC ) closed circular DNA;
L ) linear product.
A detailed investigation of the ability of the polycyclic
acridines 9b, 14, and 17e to inhibit decatenation of
kDNA by human topo IIR was conducted using the
enzyme overexpressed in S. cerevisiae strain J EL-1 from
the plasmid pYEpWob6.29 All three compounds inhibited
the release of minicircle products of kDNA (Figure 1).
Quantification of the results (Figure 2) showed that all
are more active than m-AMSA and that the indoliz-
topo IIR isoform. Comparable published values for
m-AMSA, doxorubicin, and etoposide are 20/26, 3/15,
and 39/131 µM.30
NCI COMP ARE An a lysis. COMPARE is the COM-
puterized PAttern REcognition algorithm used in evalu-
ation of data generated in the NCI in vitro cell panel.
This analysis can determine the degree of similarity,
or lack thereof, of mean graph fingerprints of new
compounds with patterns of activity of standard agents.
Compounds matched with Pearson correlation coef-
ficients (PCCs) > 0.6 frequently share related biochemi-
cal mechanisms of action.31 Three polycyclic acridines
were subjected to COMPARE analysis, and results are
shown in Table 5. At the GI50 level, with the salt 14 as
seed compound, five of the top six standard agents were
in the mechanistic category of topo II inhibitors, the
exception being 3-deazauridine (PCC ) 0.633), an
inhibitor of cytosine triphosphate synthase. A similar
pattern was shown for the pentacyclic acridine free base
17a . However, at the LC50 level (equating to cytocidal
activity), the top six COMPAREs for compound 14 were
headed by homoharringtonine, mitozolomide, and mi-
inoacridinium salt 14 was the most potent with IC50
<
1 µM. Each compound was also capable of inducing DNA
double-strand cleavage in the presence of topo IIR. When
tested at 1 µM, each compound revealed 2-4 times more
DNA cleavage than with enzyme alone (Figure 3). In
this experiment the most potent compound was the 3-(3-
cyanopropyl)pyridoacridine 9b which was approxi-
mately equiactive with m-AMSA.
Compound 14 was also tested for its ability to inhibit
the growth of S. cerevisiae transfected with recombinant
human topo IIR and â isoforms.30 The IC50 value for
killing the topo IIR-expressing yeast (22 µM) was 3-fold
lower than the comparable value (66 µM) of the IIâ-
expressing yeast suggesting a slight selectivity for the

Antitumor Polycyclic Acridines
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8 1569
Ta ble 5. NCI COMPARE Analysis with Polycyclic Acridines
14, 17a , and 17e as Seed Compounds
top six standard agentsa (PCC)b
seed
compd
GI50
LC50
14
N,N-dibenzyldaunomycin (0.701) homoharringtonine (0.694)
amonafide (0.645)
mitozolomide (0.675)
mitomycin C (0.675)
doxorubicin (0.656)
bruceantin (0.643)
amonafide (0.634)
trimetrexate (0.642)
3-deazauridine (0.633)
mitoxantrone (0.625)
daunomycin (0.616)
pyrazoloacridine (0.614)
17a pyrazoloacridine (0.755)
N,N-dibenzyldaunomycin (0.698) phosphotrienin (0.624)
3-deazauridine (0.691)
pyrazofurin (0.662)
dichloroallyllawsone (0.639)
brequinar (0.633)
d
emofolin sodium (0.608)
D-tetrandrine (0.600)
c
17e
tamoxifen (0.810)
o,p′-DDD (0.744)
hycanthone (0.737)
teroxirone (0.718)
penclomedine (0.717)
m-AMSA (0.670)
a
For structures, see the NCI DTP web pages at: http://
b
epnws1.ncifcrf.gov:2345/dis3d/dtp.html. For definitions, see ref
31. c Only four compounds with PCC values > 0.6 at LC50 level.
d
No compounds with PCC > 0.6 at GI50 level.
F igu r e 5. Fluorescence micrographs (magnification ×400;
reproduced at 60% of original for publication) of (A) MCF-7
wt and (B) A549 cells incubated for 30 min in 10 µM indolizino-
[7,6,5-kl]acridine 14 at 37 °C. Yellow fluorescence indicates
localization of drug predominantly in the nuclei of cells.
Con clu sion s
In this paper we have disclosed the first biological
results on a readily accessible series of novel polycyclic
acridines, where the core heterocyclic framework is
structurally related to bioactive marine natural prod-
ucts. Potentially the most interesting compound is the
indolizino[7,6,5-kl]acridinium salt 14 which forms a
binding ‘hot spot’ in DNA with the planar pyridoacridine
moiety intercalating at G-C sequences and the pyrroli-
dinium fragment occupying minor or major grooves.1
Clearly, inhibition of topo II is a component of the
overall activity of this compound, although there is a
distinctive fingerprint distinguishing the compound
from other agents of this type. Thus the sensitivities of
a panel of NSCL (Table 1) and breast (Table 2) cancer
cell lines are not related to their topo IIR isoform
expression; lung cells NCI-H460pv8 and H460pv8/eto
which are resistant to etoposide are not cross-resistant
to 14 (Table 3). Unlike m-AMSA, 14 is not a substrate
for P-glycoprotein-mediated cellular efflux (Table 4), and
its water solubility could be pharmaceutically advanta-
geous.
F igu r e 4. Uptake of indolizino[7,6,5-kl]acridine 14 (0.05 µM)
in MCF-7wt and A549 cells at 4 and 37 °C.
tomycin C, agents not classified as topo II inhibitors.
The amino-substituted pentacyclic acridine 17e did not
reveal any PCC values above 0.6 at the GI50 level but
high PCC values (>0.7) at the LC50 level to a group of
agents of disparate mechanisms.
Dr u g Up ta k e a n d In tr a cellu la r Loca liza tion . The
high fluorescence of compound 14 when excited at 488
nm allowed for the use of fluorescence-activated cell
sorting (FACS) to study the kinetics of drug uptake. At
37 °C compound 14 achieved plateau levels in 2 min
within breast MCF-7 and NSCLC A549 cells (Figure 4);
at 4 °C entry was much slower. The subcellular distri-
bution of the drug is shown in fluorescence images
(Figure 5). After 30 min of incubation of drug (10 µM)
with cells, compound 14 localized predominantly in the
nuclei of both cell types, although in MCF-7 cells there
appeared to be some heterogeneity of fluorescence
within the nuclei. The tetracyclic acridine 9b and
pentacyclic derivative 17e showed similar rapid nuclear
localization (data not shown).
The NCI COMPARE analysis raises the intriguing
prospect that 14 might trigger different inhibitory
molecular mechanisms at the growth inhibition (GI50
)
and cytocidal (LC50) levels. Growth inhibition induced
by 14 involves inhibition of topo II, and the localization
of the drug in the nuclei of sensitive cells supports the
disorganization of a nuclear target by this agent. In
contrast, cytocidal activity appears to be unrelated to

1570 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8
Stanslas et al.
give 8k as yellow crystals (0.61 g, 72%): mp 192-193 °C;
topo II inhibition. We have shown that compound 14 is
a potent inducer of apoptosis in lung and breast cell
lines irrespective of their p53 status (wild type or
mutant).32 The influence of this compound on cell cycle
events, and regulation of genes involved in the apoptotic
cascade, will be the subject of the next paper in this
series.
1
identical (IR, H and 13C NMR) to the sample (above).
9-(1,2,3-Ben zot r ia zol-1-yl)a cr id in e, 16a . 1,2,3-Benzo-
triazole (1.90 g) was added to a mixture of sodium hydride
(1.1 mol equiv as a 60% dispersion in mineral oil) in dry DMF
(60 mL). 9-Chloroacridine (6) (3.43 g) was added and the
mixture was refluxed (3.5 h). Addition of excess water pre-
cipitated the benzotriazolylacridine (3.40 g, 71%), identical
1
(UV, IR, H and 13C NMR) to an authentic sample.12
Syn th esis of Tetr a cyclic Acr id in es. 3-Dim eth yla m in o-
m eth yl-7H-p yr id o[4,3,2-kl]a cr id in e, 9j. A suspension of
9-(4-dimethylaminomethyl-1,2,3-triazol-1-yl)acridine (8j) (0.15
g) in diphenyl ether (5 mL) was maintained at 225 °C for 0.5
h during which effervescence of nitrogen was observed. The
mixture was fractionated on a short silica gel column to remove
diphenyl ether (eluted with hexane) followed by elution of the
product with ethyl acetate-ethanol. The pyridoacridine (0.077
g, 60%) had: mp 200 °C dec; UV 227.5, 262, 318.5, 423 nm;
IR 3389, 2944, 1638, 1555, 1466, 1339, 1154, 766 cm-1; δH
(CDCl3) 8.50 (dd, 1H, J ) 1.5, 8.0, H-11), 8.17 (s, 1H, H-2),
7.52 (t, 1H, J ) 8.2, H-5), 7.35 (m, 1H, NH), 7.20-7.40 (m,
2H, H-4, H-5), 7.11(m, 1H, J ) 1.0, 6.8, H-10), 6.93 (d, 1H, J
) 8.0, H-8), 6.71 (dd, 1H, J ) 1.0, 7.8, H-6), 3.57 (s, 2H, CH2),
2.30 (s, 6H, 2 × CH3); δC (CDCl3) 153.4 (C), 151.6 (C), 140.0
(C), 139.9 (C), 131.8 (CH), 131.5 (CH), 131.5 (CH), 125.8 (CH),
121.8 (CH), 118.9 (C), 115.7 (CH), 115.3 (CH), 114.1 (CH),
105.0 (CH), 66.3 (CH2), 46.0 (CH3); m/z (CI) 276 (MH+). Anal.
(C18H17N3) C, H, N.
Exp er im en ta l Section
Syn th etic Ch em istr y. All melting points were recorded
on a Gallenkamp melting point apparatus and are uncorrected.
UV spectra were measured in 95% EtOH on a Cecil 1020S
spectrometer. IR spectra were recorded on a Mattson 2020
GALAXY series FT-IR spectrometer as KBr disks. 1H and 13C
NMR spectra were recorded on a Bruker ARX250 spectrometer
at 250.1 and 62.9 MHz, respectively, in solvents as specified,
with tetramethylsilane or residual protic solvents as internal
standard, J values being in Hz. Low-resolution mass spectra
were recorded on an AEI MS-902 or VG Micromass 7070E.
High-resolution mass spectra (HRMS) were performed by the
EPSRC Mass Spectrometry Service Centre, Swansea. Silica
gel TLC was performed on 60F-254 precoated sheets (E.
Merck) and column chromatography on silica gel C60 (60-
120 mesh).
Syn th esis of 9-(1,2,3-Tr ia zol-1-yl)- a n d 9-(1,2,3-Ben zo-
tr ia zol-1-yl)a cr id in es. Compounds 8a -i,13 9a -i,13 and 16a -
d 12 were prepared by published methods.
2-D im e t h y la m in o m e t h y l-7H -p y r id o [4,3,2-k l ]a c r i-
d in e, 9k . Prepared (58%) from 9-(5-dimethylaminomethyl-
1,2,3-triazol-1-yl)acridine (8k ) in boiling triglyme (1.5 h), the
pyridoacridine had: mp 200 °C dec; UV 228, 267.5, 318.5, 429
9-(4-Dim e t h yla m in o m e t h yl-1,2,3-t r ia zol-1-y l)a c r i-
d in e, 8j. 9-Azidoacridine (7) (1.10 g) and 1-dimethylamino-
2-propyne (1.24 g) were dissolved in dry toluene (10 mL) and
heated to 60 °C under nitrogen for 24 h. Removal of solvent
(vacuum evaporation) gave a residue which was chromato-
graphically fractionated on silica gel using ethyl acetate-
ethanol (1:1) as eluent. Evaporation of the orange band gave
the dimethylaminotriazolylacridine 8j (0.88 g, 58%), as orange
crystals (from ethanol): mp 121-123 °C (dec at 218 °C); UV
211.6, 249.4, 359 nm; IR 3442, 2773, 1552, 1450, 1384, 1227,
1041, 807 cm-1; δH (DMSO-d6) 8.80 (s, 1H, H-5′), 8.35 (d, 2H,
J ) 8.8, H-4,5), 7.99 (m, 2H, J ) 1.4, H-3,6), 7.74 (m, 2H, J )
1.1, 8.8, H-2,7),), 7.38 (d, 2H, J ) 8.4, H-1,8), 3.77 (s, 2H, CH2),
2.29 (s, 6H, 2 × CH3); δC (DMSO-d6) 149.5 (C), 145.3 (C), 137.8
(C), 132.1 (CH), 130 3 (CH), 128.4 (CH), 128.2 (CH), 123.1
(CH), 122.2 (C), 52.8 (CH2), 45.6 (CH3); m/z (CI) 304 (MH+).
Anal. (C18H17N5) C, H, N.
9-(5-Dim e t h yla m in om e t h y l-1,2,3-t r ia zo l-1-y l)a c r i-
d in e, 8k . The pale yellow band (from ethyl acetate) from the
previous experiment afforded triazolylacridine 8k : mp 192-
193 °C (0.33 g, 22%); UV 211.3, 250.1, 359 nm; IR 3442, 2789,
1514, 1451, 1384, 1237, 1027, 754 cm-1; δH (DMSO-d6) 8.35
(d, 2H, J ) 8.9, H-4.5), 8.18 (s, 1H, H-4′), 7.98 (td, 2H,J ) 1.0,
8.8, H-3,6), 7.71 (t, 2H, J ) 8.6, H-2,7), 7.28 (d, 2H, J ) 8.9,
H-1,8), 3.27 (s, 2H, CH2), 1.83 (s, 6H, 2 × CH3); δC (DMSO-d6)
149.7 (C), 134.8 (C), 132.1 (CH), 130.3 (CH), 129.4 (CH), 128.2
(CH), 123.1 (CH), 122.8 (C), 52.8 (CH2), 45.1 (CH3); m/z (CI)
304 (MH+). Anal. (C18H17N5) C, H, N.
9-(5-Ch lor om eth yl-1,2,3-tr iazol-1-yl)acr idin e, 13. 3-Chlo-
ro-1-triphenylphosphoranylidene-2-propanone (10)33 (5.7 g)
was added to a stirred solution of 9-azidoacridine (7) (3.56 g)
in dry benzene (40 mL) and the mixture was refluxed under
N2 for 2 h. The evaporated mixture was purified by chromato-
graphic fractionation on a silica gel column with ethyl acetate:
hexane (1:1) as eluting solvent. The yellow band gave the
triazolylacridine 13 (3.28 g, 67%): mp 181-182 °C (yellow
crystals, from ethyl acetate); δH (DMSO-d6) 8.39 (s, 1H, H-4′),
8.37 (d, 2H, J ) 8.7, H-4,5), 8.00 (m, 2H, H-3,6), 7.73 (m, 2H,
H-2,7), 7.26 (d, 2H, J ) 8.7, H-1,8), 4.70 (s, 2H, CH2Cl); δC
(CDCl3) 149.2 (C), 137.1 (C), 135.4 (C), 134.0 (CH), 130.9 (CH),
129.9 (CH), 128.5 (CH), 122.7 (C), 122.0 (CH), 31.8 (CH2). Anal.
(C16H11ClN4) C, H, N.
nm; IR 3424, 1638, 1561, 1476, 1449, 1333, 1096, 802 cm-1
;
δH (CDCl3) 8.54 (dd, 1H, J ) 1.5, 8.0, H-11), 7.38 (t, 1H, J )
7.8, H-5), 7.31 (m, 1H, H-9), 7.22 (s, 1H, H-3), 7.04 (td, 1H, J
) 0.8, 8.2, H-10), 7.00 (d, 1H, J ) 8.0, H-4), 6.90 (dd, 1H, J )
1.0, 8.0, H-8), 6.62 (dd, 1H, J ) 0.8, 7.5, H-6), 3.71 (s, 2H, CH2),
2.45 (s, 6H, 2 × CH3); δC (CDCl3) 146 (CH), 139.9 (C), 139.7
(C), 137.0 (C), 132.0 (CH), 131.5 (CH), 125.4 (CH), 124.3 (C),
122.1 (CH), 115.2 (CH), 112.0 (CH), 106.1 (CH), 61.8 (CH2),
46.0 (CH3); m/z (CI) 276 (MH+). Anal. (C18H17N3) C, H, N.
Syn th esis of P en ta cyclic Acr id in es. 1H-2,3-Dih yd r oin -
d olizin o[7,6,5-kl]a cr id in iu m Ch lor id e, 14. A suspension of
9-[5-(3-chloropropyl)-1,2,3-triazol-1-yl]acridine (8m ) (1.80 g) in
diphenyl ether (35 mL) was maintained at reflux temperature
for 0.5 h under nitrogen. The melt was triturated with ethyl
acetate and the red salt 14 (1.42 g, 86%) was collected and
washed free of diphenyl ether with more ethyl acetate. The
product was identical (UV, IR, 1H and 13C NMR) with an
authentic sample.13
8H-Qu in o[4,3,2-kl]a cr id in e, 17a . A suspension of 9-(1,2,3-
benzotriazol-1-yl)acridine (16a ) (1.0 g) in triglyme (5 mL) was
boiled (4 h) and the cooled solution was diluted with excess
water. The product was collected and washed with water to
yield the quinoacridine (87%) which was identical (mp, UV,
IR, 1H and 13C NMR) with an authentic sample.12 Using
diethylene glycol as thermolysis medium afforded the same
quinoacridine (93%).
2-Nitr o-8H-qu in o[4,3,2-kl]a cr id in e, 17d . 9-(6-Nitro-1,2,3-
benzotriazolyl)acridine (16d ) (3.9 g) was stirred with diphenyl
ether (20 mL) under nitrogen to form a mustard-colored
suspension which was refluxed for 6 h. To the cooled melt was
added hexane (100 mL), with stirring, and the quinoacridine
17d (2.96 g, 82%) was collected. The product was identical (UV,
1
IR, H and 13C NMR) with an authentic sample.12
The same quinoacridine (74%) was formed when 9-(6-nitro-
1,2,3-benzotriazolyl)acridine (16d ) was thermolyzed in reflux-
ing triglyme (3.5 h) and the solution diluted with water to
precipitate the product.
2-Am in o-8H-qu in o[4,3,2-kl]a cr id in e, 17e. (i) 2-Nitro-8H-
quino[4,3,2-kl]acridine (17d ) (1.0 g) was added portionwise to
a solution of tin(II) chloride dihydrate (2.88 g, 4 mol equiv) in
10 M hydrochloric acid (15 mL) at 25 °C to give a dark brown
solution which was stirred at 25 °C for 60 h. Basification of
A solution of 13 (0.83 g) and dimethylamine (2.25 g) in THF
(25 mL) was stirred at 30-40 °C for 5 days. Solvent was
evaporated and the residue crystallized from ethyl acetate to

Antitumor Polycyclic Acridines
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8 1571
the mixture to pH >12 with excess ice-cold 10 M sodium
hydroxide solution liberated the aminoquinoacridine base (0.85
g, 95%) as a mustard-colored solid: mp 285-295 °C dec (lit.12
in a panel of 60 cell lines.35 Briefly, cell lines were inoculated
into a series of 96-well microtiter plates, with varied seeding
densities depending on the growth characteristics of specific
cell lines. Following a 24-h drug-free incubation, test agents
were added routinely at five 10-fold dilutions with a maximum
concentration of 10-4 M. After 2 days of drug exposure, the
change in protein stain optical density allowed the inhibition
of cell growth to be analyzed.
In h ibition of Top o II. Topo IIR was overexpressed in S.
cerevisiae strain J EL-1 from the plasmid pYEpWob6.29 Enzyme
was isolated to >95% purity from yeast nuclear extracts by a
four-stage chromatographic process.36 Compounds were as-
sayed for inhibition of topo IIR decatenation using kDNA
(Topogen) as a substrate and for DNA cleavage using pBR322
(gift from A. Howells, University of Leicester, U.K.).
Decatenation mixtures (20 µL) were prepared using topo
assay buffer (50 mM Tris-HCl, pH 7.5; 130 mM NaCl; 10 mM
MgCl2; 2 mM DTT; 0.5 mM EDTA; 0.25 mg kDNA; 1.2 nM
topo IIR) and contained drug at the indicated concentrations.
Reactions were started by addition of ATP to 1.25 mM,
incubated for 15 min at 37 °C and stopped by addition of 10
µL of STEB (100 mM Tris-HCl, pH 8.0; 40% w/v sucrose; 0.1
M EDTA; 0.5 mg/mL bromophenol blue).
1
mp > 300 °C dec). The product was identical (UV, IR, H and
13C NMR) to an authentic sample.12
(ii) 9-(3-Aminoanilino)acridine (18c)34 was converted to 9-(3-
azidoanilino)acridine (18d ) by sequential diazotization/azida-
tion. The azide free base (82%) was crystallized from ethyl
acetate-MeOH: mp 196-200 °C dec; UV 203.5, 225.9, 246.5,
410 nm; IR 3023 (NH), 2116 (N3), 1584, 1485 cm-1; m/z (EI)
311 (M+).
The azide (0.15 g) was added in small portions to a stirred
mixture of TFSA (0.5 mL), TFA (0.6 mL) and TFAA (0.15 mL)
at 0 °C. The mixture was maintained at 0 °C for 0.5 h, then
allowed to warm to 25 °C and stirred for a further 3 h. The
reaction mixture was basified with concentrated aqueous
ammonia-ice and products were extracted in ethyl acetate (3
× 150 mL). The washed, dried (MgSO4) organic fraction was
evaporated to give a brown oil which was chromatographically
fractionated on silica gel (ethyl acetate-hexane-MeOH) to
give 2-amino-8H-quino[4,3,2-kl]acridine (17d ) (0.02 g) and
9-(3-aminoanilino)acridine (18c) (0.022 g), identical to authen-
tic samples.
Syn th esis of Sa lts of 2-Am in o-8H-qu in o[4,3,2-kl]a cr i-
d in e, 17e. The m eth a n esu lfon ic a cid sa lt, prepared (75%)
from 17e free base and methanesulfonic acid (1.1 mol equiv)
in THF, had: mp > 300 °C dec (from aqueous 2-propanol);
UV 207.4, 237.5, 276.3, 324.3, 461.9, 518.3 nm; δH (DMSO-d6)
12.94 (s, 1H, NH), 12.80 s, 1H, NH), 8.82 (d, 1H, J ) 8.5, H-12),
8.22 (d, 1H, J ) 9.0, H-4), 7.96 (m, 3H, H-5,6,10), 7.68 (d, 1H,
J ) 8.5, H-9), 7.54 (t, 1H, J ) H-11), 7.39 (d, 1H, J ) 7.6,
H-7), 7.14 (d, 1H, J ) 1.7, H-1), 6.91 (dd, 1H, J ) 8.8, H-3),
6.31 (brs, 2H, NH2), 2.35 (s, 3H, CH3); δC (DMSO-d6) 151.2
(C), 147.25 (C), 140.3 (C), 140.1 (C), 136.6 (CH), 136.4 (C), 135.4
(CH), 134.1 (C), 125.1 (CH), 124.7 (CH), 123.4 (CH), 118.3
(CH), 116.2 (CH), 112.0 (C), 116.7 (C), 111.5 (C), 111.1 (CH),
110.9 (CH), 40.2 (CH3). Anal. (C20H17N3O3S) C, H, N.
The eth a n esu lfon ic a cid sa lt, similarly prepared from 17e
and ethanesulfonic acid, had: mp > 300 °C dec (from aqueous
ethanol); UV 206.6, 236.7, 275.9, 325.5, 457.1, 519.0 nm. Anal.
(C21H19N3O3S) C, H, N.
Cleavage assay mixtures (10 µL) contained drug at the
indicated concentrations in cleavage buffer (50 mM Tris-HCl,
pH 7.5; 50 mM NaCl; 10 mM MgCl2; 2 mM DTT, 0.5 mM
EDTA, 2.5 mM ATP; 0.125 mg pBR322; 30 nM topo IIR).
Reactions were incubated for 10 min at 25 °C and stopped by
addition of 1 µL of 10% SDS. Mixtures were then incubated
with proteinase K (0.1 mg/mL) for 30 min at 37 °C, 10 µL of
STEB was added and samples were finally extracted with 50
µL of phenol-CHCl3-isoamyl alcohol (25:24:1).
Completed reactions were analyzed by 0.8% agarose gel
electrophoresis with gels containg 40 µg/mL ethidium bromide.
Gels were scanned and DNA was quantitated using a video
camera linked to a Syngene gel analysis program.
Activity of Com p ou n d s a ga in st S. cer evisia e Exp r ess-
in g Recom bin a n t Hu m a n Top o II. This experiment used
logarithmically growing yeast strain J N394 top2-4 transformed
with a human TOP2R, human TOP2â or yeast TOP2 plasmid,
and activity of compound 14 was assayed according to the
method of Meczes et al.30
Dr u g Up ta k e a n d In tr a cellu la r Loca liza tion . Cells were
trypsinized and collected in fresh RPMI 1640 medium and the
cell density was adjusted to 5 × 105-106 cells/mL. The cells
were kept on ice or 37 °C prior to introduction of drugs. To 1
mL of cells in 12 × 75-mm tubes, compound 14 was added to
a final concentration of 50 nM. The average fluorescence (Fl1)
of 2000 cells was recorded using a Becton-Dickinson FACScan
flow cytometer at diffrent time periods. The kinetic uptake of
drugs was expressed as the change of mean cellular fluores-
cence with time.
The tetr a flu or obor ic a cid sa lt, from 17e and 48% fluo-
roboric acid, had: mp 245 °C dec (from aqueous 2-propanol);
UV 207.5, 238.4, 276.6, 326.1, 520.0 nm.
In Vitr o Cytotoxicities of P olycyclic Acr id in es. All the
tested polycyclic acridines and m-AMSA were dissolved in
DMSO to make stock solutions of 10 mM. Stock solutions of
etoposide (10 mM) and doxorubicin (1 mM) were prepared in
saline (0.9%). All drug solutions were stored at -20 °C.
NSCLC and breast cancer cells were grown in RPMI 1640
culture medium supplemented with 2 mM L-glutamine, 100
IU/mL penicillin, 100 µg/mL streptomycin and 10% heat-
inactivated FCS. Initial seeding densities per well for 7-day
cytotoxicity assays were: A549 (250), NCI-H460 (250), NCI-
H647 (300), NCI-H226 (1000), NCI-H322 (1000), NCI-H358
(1000), NCI-H460pv8 (500), MCF-7 (300), SKBr-3 (1000), T47D
(1000), MDA-231 (800), ZR-75-1 (1000), MDA-468 (800), MCF-
7/adr (800).
MTT Color im etr ic Assa ys. The fraction of viable cells
after drug treatment was determined by the ability of cells to
metabolize MTT. To each well was added 50 µL of MTT
solution (2 mg/mL) to yield a final concentration of 0.4 mg/
mL and the plates were further incubated at 37 °C (95% air,
5% CO2) for 4 h to allow viable cells to convert soluble MTT to
insoluble purple formazan. The medium containing MTT was
aspirated and the formazan was dissolved by adding DMSO
(100 µL) and glycine buffer (25 µL). The absorbance of the
formazan solutions was determined at 550 nm using an Anthos
Labtec System plate reader. The IC50 values (concentration of
drug to produce a 50% reduction in the absorbance of the drug
untreated controls) were determined from the dose-response
curves.
For determination of drug intracellular distribution in
adherent cells A549 and MCF-7 cells were grown on cover slips
in 60 × 10-mm Petri dishes containing 10 mL of RPMI medium
at 37 °C in a 5% CO2 and 95% air atmosphere. Cells were
incubated with compound 14 (10 µM) for 30 min. Cells were
washed free of drugs with cold PBS and the coverslips were
mounted onto microscope slides using Fluoromount mountant
(BDH Chemicals). Slides were visualized under blue light
(503-530 nm) using a Leitz Dialux 20 fluorescence microscope.
Fluorescence micrographs were captured using 400 or 800 ASA
film.
Ack n ow led gm en t. This work was supported by the
Cancer Research Campaign (U.K.). The authors thank
Dr. Caroline A. Austin (University of Newcastle, U.K.)
for yeast cells transfected with human topo II isoforms
and for experimental assistance and Dr. R. A. Robins
(University of Nottingham) for assistance with FACS
studies and helpful discussions. The University Putra
Malaysia is thanked for financial support (to J .S.).
NCI in Vitr o Cytotoxicity Assa ys. Sulforhodamine B
assays were used for assessing the cytotoxicity of test agents

1572 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 8
Stanslas et al.
1988, 88, 298-368. Stevens, M. F. G.; Shi, D.-F.; Castro, A.
Antitumour benzothiazoles. Part 2. Formation of 2,2′-diamino-
biphenyls from the decomposition of 2-(4-azidophenyl)benzothi-
azoles in trifluoromethane sulfonic acid. J . Chem. Soc., Perkin
Trans. 1 1996, 83-93.
Refer en ces
(1) Part 6: Bostock-Smith, C. E.; Gime´nez-Arnau, E.; Missailidis,
S.; Laughton, C. A.; Stevens, M. F. G.; Searle, M. S. Molecular
recognition between a new pentacyclic acridinium salt and DNA
sequences investigated by optical spectroscopic techniques,
proton nuclear magnetic resonance spectroscopy and molecular
modeling. Biochemistry 1999, 38, 6723-6731.
(2) Albert, A. The Acridines, 2nd ed.; Edward Arnold: London, 1966.
(3) Albert, A. Selective Toxicity, 7th ed.; Chapman & Hall: London,
1985.
(4) Denny, W. A.; Baguley, B. C. In Molecular Aspects of Anti-Cancer
Drug-DNA Interaction; Neidle, S., Waring, M. J ., Eds.; Mac-
millan: London, 1994; pp 270-311.
(5) Baguley, B. C. DNA intercalating anti-tumour agents. Anti-
Cancer Drug Des. 1991, 6, 1-35.
(6) Gamage, S. A.; Spicer, J . A.; Atwell, G. J .; Finaly, G. J .; Baguley,
B. C.; Denny, W. A. Structure-activity relationships for sub-
stituted bis(acridine-4-carboxamides): a new class of anticancer
agents. J . Med. Chem. 1999, 42, 2383-2393.
(7) Kohn, K. W. Beyond DNA cross-linking: history and prospects
of DNA-targeted cancer treatment - Fifteenth Bruce F. Cain
Memorial Award Lecture. Cancer Res. 1996, 56, 5533-5546.
(8) Molinski, T. F. Marine pyridoacridine alkaloids: structure,
synthesis, and biological chemistry. Chem. Rev. 1993, 93, 1825-
1838.
(9) Alvarez, M.; Salas, M.; J oule, J . A. Marine, nitrogen-containing
heterocyclic natural products. Structures and synthesis of
compounds containing quinoline and/or isoquinoline units.
Heterocycles 1991, 32, 759-794. Alvarez, M. Synthesis of pyrido-
acridines. Heterocycles 1992, 34, 2385-2405.
(10) Copp, B. R.; J ompa, J .; Tahir, A.; Ireland, C. M. New tetracyclic
pyridoacridine alkaloids from the Indonesian Ascidian, Eusyn-
styela latericius. J . Org. Chem. 1998, 63, 8024-8026.
(11) Fugman, B.; Steffan, B.; Steglich, W. Necatorone, an alkaloidal
pigment from the gilled toadstool Lactarius necator (Agaricales).
Tetrahedron Lett. 1984, 25, 3575-3578. Hilger, C. S.; Fugman,
B.; Steglich, W. Synthesis of necatorone. Tetrahedron Lett. 1985,
26, 5975-5978. Klamann, J .-D.; Fugmann, B.; Steglich, W.
Alkaloidal pigments from Lactarius necator and L. atroviridis.
Phytochemistry 1989, 28, 3519-3523.
(12) Hagan, D. J .; Gime´nez-Arnau, E.; Schwalbe, C. H.; Stevens, M.
F. G. Antitumour polycyclic acridines. Part 1. Synthesis of 7H-
pyrido- and 8H-quino-[4,3,2-kl]acridines by Graebe-Ullmann
thermolyses of 9-(1,2,3-triazol-1-yl)acridines: application of dif-
ferential scanning calorimetry to predict optimum cyclisation
conditions. J . Chem. Soc., Perkin Trans. 1 1997, 2739-2746.
(13) Hagan, D. J .; Chan, D.; Schwalbe, C. H.; Stevens, M. F. G.
Antitumour polycyclic acridines. Part 3. A two-step conversion
of 9-azidoacridine to 7H-pyrido[4,3,2-kl]acridines by Graebe-
Ullmann thermolysis of substituted 9-(1,2,3-triazol-1-yl)acridines.
J . Chem. Soc., Perkin Trans. 1 1998, 915-923.
(14) J ulino, M.; Stevens, M. F. G. Antitumour polycyclic acridines.
Part 5. Synthesis of 7H-pyrido[4,3,2-kl]acridines with exploitable
functionality in the pyridine ring. J . Chem. Soc., Perkin Trans.
1 1998, 1677-1684.
(15) Fixler, N.; Demeunynck, M.; L’homme, J . Regioselective elec-
trophilic substitution of 2-hydroxy and 2-methoxy substituted
acridines. Application to the synthesis of pyrido[2,3,4-mn]-
acridine. Synth. Commun. 1997, 27, 2311-2324.
(16) Gime´nez-Arnau, E.; Missailidis, S.; Stevens, M. F. G. Antitumour
polycyclic acridines. Part 2. Physicochemical studies on the
interactions between DNA and novel polycyclic acridine deriva-
tives. Anti-Cancer Drug Des. 1998, 13, 125-143.
(17) Gime´nez-Arnau, E.; Missailidis, S.; Stevens, M. F. G. Antitumour
polycyclic acridines. Part 4. Physicochemical studies on the
interactions between DNA and novel tetracyclic acridine deriva-
tives. Anti-Cancer Drug Des. 1998, 13, 431-451.
(22) Ohta, T.; Machida, R.; Takeda, K.; Endo, Y.; Shudo, K.; Okamoto,
T. Reductive phenylation of nitroarenes. J . Am. Chem. Soc. 1980,
102, 6385-6386.
(23) Giaccone, G.; Gadzar, A. F.; Beck, H.; Zunino, F.; Capranico, G.
Multidrug sensitivity phenotype of human lung cancer cells
associated with topoisomerase II expression. Cancer Res. 1992,
52, 1666-1674. Kasahara, K.; Fujiwara, Y.; Sugimoto, Y.; Nishio,
K.; Tamura, T.; Matsuda, T.; Saijo, N. Determinants of response
to the DNA topoisomerase II inhibitors adriamycin and etopo-
side. J . Natl. Cancer Inst. 1992, 84, 113-118. Kaufman, S. H.;
Karp, J . E.; J ones, R. J .; Miller, C. B.; Schneider, E.; Zwelling,
L. A.; Cowan, K.; Wendel, K.; Burke, P. J . Topoisomerase II
levels and drug sensitivity in adult acute myelogenous leu-
kaemia. Blood 1994, 83, 517-530.
(24) Houlbrook, S.; Harris, A. L.; Carmichael, J .; Stratford, I. J .
Relationship between topoisomerase II levels and resistance to
topoisomerase II inhibitors in lung cancer cell lines. Anticancer
Res. 1996, 16, 1603-1610.
(25) Houlbrook, S.; Addison, C. M.; Davies, S. L.; Carmichael, J .;
Stratford, I. J .; Harris, A. L.; Hickson, I. D. Relationship between
expression of topoisomerase II isoforms and intrinsic sensitivity
to topoisomerase II inhibitors in breast cancer cell lines. Br. J .
Cancer 1995, 72, 1454-1461.
(26) Monks, A.; Scudiero, D. A.; J ohnson, G. S.; Paull, K. D.; Sausville,
E. A. The NCI anti-cancer drug screen: a smart screen to
identify effectors of novel targets. Anti-Cancer Drug Des. 1997,
12, 533-541.
(27) Wang, H.; J iang, Z.; Wong, Y. W.; Dalton, W. S.; Futscher, B.
W.; Chan, V. T.-W. Decreased CP-1 (NF-Y) activity results in
transcriptional down-regulation of topoisomerase IIR in a doxo-
rubicin-resistant variant of human multiple myeloma RPMI
8226. Biochem. Biophys. Res. Commun. 1997, 237, 217-224.
(28) Fairchild, C. R.; Ivy, S. P.; Kao-Shan, C.-S.; Whang-Peng, J .;
Rosen, N.; Israel, M. A.; Melera, P. W.; Cowan, K. H.; Goldsmith,
M. E. Isolation of amplified and overexpressed DNA sequences
from adriamycin-resistant human breast cancer cells. Cancer
Res. 1987, 47, 5141-5148.
(29) Wasserman, R. A.; Austin, C. A.; Fisher, L. M.; Wang, J . C. Use
of yeast in the study of anticancer drugs targeting DNA
topoisomerases: expression of a functional recombinant human
DNA topoisomerase IIR in yeast. Cancer Res. 1993, 53, 3591-
3596.
(30) Meczes, L. M.; Marsh, K. L.; Fisher, L. M.; Rogers, M. P. Austin,
C. A. Complementation of temperature sensitive topoisomerase
II mutations in S. cerevisiae by a human TOPIIâ contruct allows
study of topoisomerase IIâ inhibitors in yeast. Cancer Chemo-
ther. Pharmacol. 1997, 39, 367-375.
(31) Weinstein, J . N.; Myers, T. G.; O’Connor, P. M.; Friend, S. H.;
Fornace, A. J .; Kohn, K. W.; Fojo, T.; Bates, S. E.; Rubinstein,
L. V.; Anderson, N. L.; Buolamwini, J . K.; van Osdol, W. W.;
Monks, A. P.; Scudiero, D. A.; Sausville, E. A.; Zaharevitz, D.
W.; Bunow, B.; Viswanadhan, V. N.; J ohnson, G. S.; Wittes, R.
E.; Paull, K. D. An information-intensive approach to the
molecular pharmacology of cancer. Science 1997, 275, 343-349.
(32) Stanslas, J .; Marsh, K. L.; Austin, C. A.; Robins, R. A.; Price,
M. R.; Double, J . A.; Stevens, M. F. G. Antitumour properties of
novel aza-polycyclic compounds. Br. J . Cancer 1998, 78, (S1) P32.
Bostock-Smith, C. E.; Stanslas, J .; Gime´nez-Arnau, E.; Missai-
lidis, S.; Laughton, C. A.; Searle, M. S.; Stevens, M. F. G.
Noncovalent encounters between DNA and a new pentacyclic
heteroaromatic salt and the biological sequelae. Proc. Am. Assoc.
Cancer Res. 1999, 40, P32.
(33) Hudson, R. F.; Chopard, P. A. U.S. Patent 3294820, 1963; Chem.
Abstr. 1967, 66, 104827d.
(34) Atwell, C. J .; Cain, B. F.; Seelye, R. N. Potential antitumor
agents. 12. 9-Anilino-acridines. J . Med. Chem. 1972, 15, 611-
615.
(35) Boyd, M. R.; Paull, K. D. Some practical considerations and
applications of the National Cancer Institute in vitro anticancer
drug discovery screen. Drug Dev. Res. 1995, 13, 91-109.
(36) Hammonds, T. R.; Maxwell, A. The DNA dependence of the
ATPase activity of human DNA topoisomerase IIR. J . Biol.
Chem. 1997, 272, 32696-32703.
(18) Mair, A. C.; Stevens, M. F. G. Azidoacridines: potential nucleic
acid mutagens. J . Chem. Soc., Perkin Trans. 1 1972, 161-165.
(19) Reynolds, G. A. The reaction of organic azides with benzyne. J .
Org. Chem. 1964, 29, 3733-3734.
(20) Mitchell, G.; Rees, C. W. Cyclo-octa[def]carbazole:
a new
heterocyclic paratropic ring system. J . Chem. Soc., Perkin Trans.
1 1987, 403-412.
(21) Abramovitch, R. A.; Chinnasmay, P.; Evertz, K.; Huttner, G.
Cyclization of arylnitrenium ions to the apomorphine ring
system: remarkable formation of a sixteen-membered ring by
an intramolecular electrophilic aromatic substitution. J . Chem.
Soc., Chem. Commun. 1989, 3-5. Scriven, E. F. V.; Turnbull,
K. Azides: their preparation and synthetic uses. Chem. Rev.
J M9909490