A new synthetic agent with potent but selective cytotoxic activity against cancer

Article abstract of DOI:10.1021/jm048946x

The synthesis of novel unsymmetrical bifunctional antitumor agents was accomplished by linking an imidazoacridone moiety to another polycyclic heteroaromatic moiety via linkers of various length and rigidity. These compounds bind to cellular DNA, but it is hypothesized that biological effects become manifested when the drug-DNA complexes interact with critical DNA binding proteins that are involved in repair and transcription. The most promising compound of the series, 4ad (WMC79), consists of an imidazoacridone linked to a 3-nitronaphthalimide moiety via a 1,4-dipropanopiperazine linker. It was found to be potently, but selectively, cytotoxic against colon cancers (GI ₅₀ = 0.5 nM, LC₅₀ = 32 nM) and leukemias (GI₅₀ = 3.5 nM, LC₅₀ = 33 nM). Compound 4ad, which appears to be a candidate for further development as an anticancer drug, kills sensitive cells by induction of apoptosis. It also showed significant in vivo activity against HCT-116 colon cancer xenografts in nude mice. Other compounds in the series also exhibited antitumor properties, but they were significantly lower than that of 4ad.

Full text of DOI:10.1021/jm048946x

4
474
J. Med. Chem. 2005, 48, 4474-4481
A New Synthetic Agent with Potent but Selective Cytotoxic Activity against
Cancer
†
†
‡
†
Wieslaw M. Cholody, Teresa Kosakowska-Cholody, Melinda G. Hollingshead, Humcha K. Hariprakasha, and
Christopher. J. Michejda*,†
Molecular Aspects of Drug Design, Structural Biophysics Laboratory, Center for Cancer Research,
National Cancer InstitutesFrederick, Frederick, Maryland 21702, and Developmental Therapeutics Program,
National Cancer InstitutesFrederick, Frederick, Maryland 21702
Received December 30, 2004
The synthesis of novel unsymmetrical bifunctional antitumor agents was accomplished by
linking an imidazoacridone moiety to another polycyclic heteroaromatic moiety via linkers of
various length and rigidity. These compounds bind to cellular DNA, but it is hypothesized
that biological effects become manifested when the drug-DNA complexes interact with critical
DNA binding proteins that are involved in repair and transcription. The most promising
compound of the series, 4ad (WMC79), consists of an imidazoacridone linked to a 3-nitronaph-
thalimide moiety via a 1,4-dipropanopiperazine linker. It was found to be potently, but
selectively, cytotoxic against colon cancers (GI50 ) 0.5 nM, LC50 ) 32 nM) and leukemias (GI50
)
3.5 nM, LC50 ) 33 nM). Compound 4ad, which appears to be a candidate for further
development as an anticancer drug, kills sensitive cells by induction of apoptosis. It also showed
significant in vivo activity against HCT-116 colon cancer xenografts in nude mice. Other
compounds in the series also exhibited antitumor properties, but they were significantly lower
than that of 4ad.
Introduction
Many clinically effective anticancer drugs, regardless
of their primary cellular targets (see Table 1 in ref 1),
kill tumor cells either by apoptosis in susceptible cell
2-7
types
ing “slow cell death” or “aponecrosis”.
or by various nonapoptotic mechanisms includ-
8
-10
Cytotoxic
agents, however, frequently exhibit unspecific toxicities,
which has limited enthusiasm for new cytotoxic agents.
Nevertheless, the ability to selectively kill tumor cells
remains a highly desirable property of potential new
anticancer drugs.
Several years ago our laboratory reported on a new
class of promising antitumor agents with remarkable
selectivity against colon cancers s the bisimidazoacri-
Figure 1. Structures of symmetrical bifunctional agents 1
and 2.
1
1,12
dones (BIAs).
ited potent and highly selective antiviral properties.
Some of these compounds also exhib-
13
presence of two aromatic ring systems and the proper-
ties of the linker that joins the aromatic moieties.
Structures 1 and 2 are typical BIAs, where 1, called
WMC-26, is a potently cytostatic antitumor compound
Spectroscopic studies on the mode of binding of repre-
11
19,21
that targets the GI tract and leukemias, while 2, called
temacrazine, is an anti-HIV drug that targets the
transcription of the integrated viral genome.13 These
compounds were originally designed to be bisinterca-
lating agents, and their molecular construction was
based on previous findings about the structural require-
sentative BIAs to various types of DNA
revealed
that only one of the aromatic residues intercalates into
DNA, while the other aromatic moiety and the linker
reside in the minor groove. These conclusions were
based on a combination of steady state and time-
resolved fluorescence spectroscopy studies, which indi-
cated that only one chromophore was intercalated to
DNA and fluorescence anisotropy measurements showed
that the entire molecule was associated with the DNA
helix. The best model to account for this behavior
involves the intercalation of one chromophore while the
other, together with the linker, resides in the minor
groove. This model was additionally supported by DNA
melting experiments and gel shift studies.19 These and
other results led us to hypothesize that the unique
biological properties of BIAs reflect a highly specific
interaction with DNA that allows the BIA-DNA com-
14-18
ments for bisintercalation.
However, we found that
although BIAs did bind to DNA, the mode of binding
1
1,19
did not involve bisintercalation.
Likewise, their
biological mode of action was substantially different
from that of ditercalinium, a structurally related com-
20
pound that is a classical bisintercalator. Two structural
features are crucial for antitumor activity of BIAs: the
*
To whom correspondence should be submitted. Tel: 301-846-1216.
Fax: 301-846-6231. E-mail: michejda@ncifcrf.gov.
†
Molecular Aspects of Drug Design Section.
Developmental Therapeutics Program.
‡
1
0.1021/jm048946x This article not subject to U.S. Copyright. Published 2005 by the American Chemical Society
Published on Web 06/08/2005

New Synthetic Agent Against Cancer
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 13 4475
Table 1. Cytotoxic Activity of New Agents against Selected Cancer Cell Lines
tumor cell lines
SK-MEL-2
HCT-116
A549
HL-60
MCF-7
compound
GI50a
LC50b
GI50
LC50
GI50
LC50
GI50
LC50
GI50
LC50
4
4
5
6
ac
ad
ad
8
220
32
25
0.6
2.5
8
500
150
4
>1000
60
120
3.5
5
400
45
500
50
3.2
32
400
245
>1000
>1000
>1000
>1000
>1000
0.5
2.2
3.8
250
35
33
80
100
15
750
450
>1000
>1000
>1000
300
150
>1000
700
>1000
mitonafide
65
800
400
a
GI50: concentration of drug (nM) resulting in inhibition of cell growth to 50% of controls. b LC50: concentration of drug (nM) required
to reduce by 50% the initial cell number. The GI50 and LC50 values are the average of at least three independent determinations; the
coefficients of variation were between 8 and 20%.
plex to interact with a critical DNA binding protein. The
symmetrical BIAs, for example, are extremely potent
inhibitors of HIV-1 integrase in a biochemical assay,
where they interfere with 3′-processing and the strand
transfer reaction. The biochemical results are best
accounted for by the formation of a ternary complex of
DNA, integrase and the BIA, which results in the
midazoacridones 3ac and 3ad were described previous-
11,21
ly.
Compounds 3 were condensed with 3-nitro-1,8-
naphthalic anhydride in N,N-dimethylformamide (DMF)
to give compounds 4. Compound 4ad (which is com-
monly known as WMC79) was further transformed into
5ad by reduction of the nitro group with stannous chlor-
ide. Compound 6 was synthesized by the reaction of 3ad
with 3-chloro-7-methoxy-9-phenoxyacridine in phenol.
The new compounds in their neutral state were
completely insoluble in aqueous media. Consequently,
they were transformed into tri(methanesulfonate) salts,
which were sufficiently soluble in water and in dilute
buffers for biological testing. Aqueous solutions were
used for in vitro biological testing while 5% dextrose
solutions were used for in vivo experiments. Other salts
(hydrochlorides, hydrobromides, and trifluroacetates)
were also prepared, but as was found with the sym-
metrical BIAs, the methanesulfonate salts appeared to
have the most favorable properties for subsequent
1
3
inactivation of the enzyme. It is clear that the two
aromatic moieties play different roles, or in other words,
the unsymmetrical DNA molecule induces asymmetry
in the binding of the symmetric BIAs. Thus, the
postulated model proposes that one of the aromatic
moieties provides the means for docking the drug
molecule to DNA by intercalation, while the second,
together with the linker, interacts with a critical protein.
Since the role of the two aromatic moieties of sym-
metrical BIAs was different, it was interesting to
examine the biological properties of molecules where the
aromatic residues were in fact different. We expected
that replacement of one imidazoacridone moiety with
another planar ring system with a different tendency
for intercalation to DNA could provide compounds with
novel antitumor properties. To test this possibility, we
exchanged one imidazoacridone chromophore of the
BIAs with a variety of planar aromatic molecules. In
this paper, we present the results of substitution of the
imidazoacridone moiety with either a 1,8-naphthalimide
or a 3-chloro-7-methoxyacridine moiety.
testing. The new compounds are highly fluorescent, as
was the case for the symmetrical BIAs.1
1,12
The fluo-
rescent properties were very useful for studying cellular
localization of the compounds and also for a variety of
biophysical measurements. Spectroscopic experiments
(Tarasov et al., unpublished results) show that 4ad
binds to DNA in an analogous manner to BIAs.19 One
aromatic moiety intercalates to DNA while the other,
together with the linker, resides in the minor groove.
However, unlike the symmetrical BIAs, 4ad did not
exhibit enhanced fluorescence in the presence of DNA.
All of the compounds reported in this paper are very
stable in the solid state, especially in the methane-
Chemistry
The general synthetic route to the new compounds is
presented in Scheme 1. The starting monofunctional im-
Scheme 1a
a
Reagents and conditions: (a) 4, 3-nitro-1,8-naphthalic anhydride, DMF, 80 °C; 6, 3-chloro-7-methoxy-9-phenoxyacridine, phenol, 90
°
C. (b) SnCl2, HCl, AcOH, 60 °C.

4
476 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 13
Cholody et al.
Figure 2. In vitro activity of new compounds against human
colon cancer HCT-116 measured by a 5 day MTT assay as
described in the Experimental Section. Similar experiments
were run for other tumor cell lines, and the data were used to
calculate the GI50 and LC50 values presented in Table 1. Values
shown are means ( SD of at least of three independent
experiments.
sulfonate form, but appear to suffer some photochemical
degradation when dilute solutions are exposed to ambi-
ent light for prolonged periods of time. The compounds
are also adsorbed onto glass and quartz surfaces, which
affects the concentrations of dilute solutions signifi-
cantly. Curiously, these solutions were not affected by
polyethylene or polycarbonate surfaces; hence solutions
in plastic containers could be stored without significant
loss of material. All of the new compounds were sub-
jected to biological testing.
Figure 3. (A-D) The LIVE/DEAD two-color cell viability
experiment: A, untreated colon tumor HCT-116 cells (live,
negative control); B, cells treated with 70% ethanol for 30 min
(
dead, positive control); C and D, cells exposed to 100 nM 4ad
for 48 and 96 h, respectively. Both attached and floating cells
were included in the assay.
showed only moderate cytotoxicity against two cell lines
A549 and SK-MEL-2 cells and was not significantly
active against the other lines tested.
Biological Results
Cytotoxicity. Initial evaluation of 4ac (NSC 695941)
and 4ad (NSC 695942) in the standard National Cancer
Institute (NCI) 60 human tumor cell line screen22
showed a very high degree of growth inhibition (GI50)
but the capacity to produce cell killing (LC50) was much
less pronounced. The readings were made after a 48 h
exposure of the cells to the chemicals. Consequently, we
examined the effect of the compounds on several tumor
cell lines utilizing the MTT assay for cytotoxicity but
extending the drug exposure time to 120 h. Figure 2
presents typical dose-response curves for HCT-116
colon tumor. The GI50 and LC50 values were determined
from these and similar data for other tumor cell lines
LIVE/DEAD Viability/Cytotoxicity Assay. This
two-color test clearly distinguishes between live cells
with intracellular esterase activity (green fluorescence
from the calcein produced by the enzymatic conversion
of nonfluorescent calcein AM, Figure 3A) and dead cells
with damaged membranes (red fluorescence from the
ethidium dimer bound to the exposed cellular DNA,
Figure 3B). This experiment confirmed the potent
cytotoxicity of 4ad. First, dead cells were detected after
only 48 h of exposure to 100 nM of the drug, while 96 h
of treatment caused the death of almost the entire
population, as demonstrated by the red fluorescence
from the ethidium dimer (Figure 3D).
2
3
(
Table 1). These results showed that 4ad, which con-
DNA Fragmentation. The DNA from HL-60 cells
treated with 4ad and control untreated cells was
extracted and analyzed by electrophoresis on agarose
gels. Figure 4 shows that the genomic DNA of HL-60
cells treated with 100 nM 4ad was severely fragmented
at 72 and 96 h timepoints (lanes 4 and 5). Compound
4ad causes the fragmentation of HL-60 DNA much
more slowly than camptothecin, which induced maximal
nucleosomal fragmentation of DNA after only 5 h of
treatment (lane 6). The first signs of internucleosomal
laddering were visible after 48 h of treatment in cells
treated with 4ad (lane 3).
DNA Staining. We used DAPI (4′,6-diamino-2-phe-
nylindole), a DNA dye to examine morphological changes
in the nuclei of untreated and drug-treated cells as well
as to determine the percentage of apoptotic cells by
fluorescence activated cell sorting (FACS) analysis. Treat-
ment of HL-60 cells with 100 nM 4ad results in the
accumulation of cells in the sub-G0/G1 phase in a time-
dependent manner. As shown in Figure 5, the popula
tains a 3-nitronaphthalimide chromophore and pipera-
zine moiety in the linker, exhibited outstanding cyto-
toxic activity, with LC50 values (drug concentration
required to reduce the initial cell number by 50% after
1
20 h) below 100 nM, against all of the cell lines except
the breast cancer line MCF-7. Compound 4ac a struc-
tural analogue with a somewhat shortened and more
flexible linker was significantly less active at both the
GI50 and LC50 levels. On the other hand, compound 6,
which contains the piperazine linker but has the naph-
thalimide chromophore replaced with a substituted
acridine, resembled the original bisimidazoacridones
and was devoid of the cell killing activity. Compound
5
ad, in which the nitro group was reduced to an amino
group, was also cytotoxic but significantly less so than
the oxidized nitro form. It is worth noting that mitona-
fide, a 3-nitronaphthalimide derivative (Scheme 1),
which was tested in phase II clinical trials against
2
4
25
colorectal cancer and non-small-cell lung cancers,

New Synthetic Agent Against Cancer
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 13 4477
Figure 4. Internucleosomal DNA fragmentation and morphological changes induced by 4ad in HL-60 cells. Agarose gel
electrophoresis of DNA isolated from HL-60 untreated cells (lane 1), cells treated with 100 nM 4ad for 24, 48, 72, and 96 h (lanes
2, 3, 4, and 5, respectively) or 100 nM camptothecin for 5 and 24 h (lanes 6 and 7). DNA was stained with ethidium bromide after
electrophoresis on a 1.8% agarose gel and then visualized under UV light. Oligonucleosome-sized DNA fragmentation can be
clearly seen in drug-treated cells after 72 and 96 h of treatment. Sizes of the molecular weight marker (M) are indicated to the
left. DAPI staining shows morphological changes in the nuclei associated with DNA fragmentation: A, untreated cells and B, C,
and D, cells exposed to 100 nM 4ad for 48, 72, and 96 h, respectively. In the treated cells, the nuclei were fragmented, and this
was visualized as apoptotic bodies.
Figure 5. HL-60 nuclei stained with DAPI. HL-60 cells were incubated with and without 4ad or camptothecin. The percent of
apoptotic cells (M1) was determined by FACS after DAPI staining, as described in the Experimental Section: A, untreated cells;
B, cells exposed to 100 nM camptothecin for 24 h; C, D, and E, cells treated with 100 nM 4ad for 24, 72, and 96 h, respectively.
tion of apoptotic cells distinctly increased after 72 h of
exposure to the drug (from 5.97% in untreated to 45.34%
in drug-treated cells). These results are consistent with
data obtained from the experiments described above.
Activation of Caspase-3. To examine the activity
of the proapoptotic caspase-3, we used a specific fluo-
rogenic caspase-3 substrate, Ac-DEVD-. Compound 4ad
caused a time-dependent increase in 7-amino-4-meth-
ylcoumarin (AMC) fluorescence that is consistent with
the activation in caspase-3. As shown in Figure 6,
caspase-3 activity was dramatically increased (sevenfold
increase relative to untreated cells) in HCT-116 colon
cancer after 48 h in response to 100 nM 4ad. Only
marginal activity was detected in the presence of the
caspase-3 inhibitor (Ac-DEVD-CHO). In contrast, HL-
60 cell lysates had a peak (10-fold) in enzyme activity
after 72 h of treatment (data not shown).
In Vivo Studies. Compound 4ad was administered
intravenously (iv) to athymic mice (nu/nuNCr) im-
planted subcutaneously with HCT-116 human colon
tumor xenografts. The drug was solubilized in 5%
dextrose in water and given once every third day for a
total of 5 doses. On this schedule, 40 mg/kg was toxic
to 6 out of 10 animals while the 20 mg/kg and 10 mg/kg
doses were well tolerated and showed significant inhibi-
tion of tumor growth. No significant dose response was
evident from the data, which are presented in Figure
7.

4
478 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 13
Cholody et al.
lines of evidence suggest that the biological activities
of BIAs, including the compounds presented in this
paper, are the result of ternary complex formation
involving the BIA, DNA, and a relevant protein. BIAs
bind to DNA with KD in the micromolar range,19 but
other experiments have shown that they do not bind to
plasma proteins (Stinson, unpublished data), which
suggest that binding to protein is weak. However, the
biological activity of 4ad is manifested at low nanomolar
concentrations (see Table 1), several orders of magni-
tude lower than would be predicted from DNA binding
data. This fact, together with the demonstration that
ternary complexes can be seen in biochemical experi-
ments, strongly suggests the formation of such com-
plexes in cellular DNA and their involvement in the
biological activity of the compounds.
Figure 6. Time-dependent activity of caspase-3 in HCT-116
cells treated with 4ad. Untreated and 4ad treated HCT-116
cells were harvested at the indicated times. Cellular lysates
were incubated with Ac-DEVD-AMC and caspase-3 substrate
and analyzed for the presence of liberated AMC by spectrof-
luorometry (excitation at 380 nm, emission at 440 nm). Lysates
from untreated cells were incubated with Ac- DEVD-AMC
alone to determine the basal levels of apoptosis at each
timepoint (A). Lysates from cells treated with 100 nM 4ad
were incubated with the substrate alone (B) and the substrate
and the Ac-DEVD-CHO caspase-3 inhibitor (C). The highest
level of caspase-3 activity was seen after 48 h of treatment
with the drug.
As stated above, evaluation of 4ac and 4ad in the NCI
0 cell line screen showed very high, but selective,
6
growth inhibition (GI50) activity against some cancer
types (colon, leukemia), while the level of cell killing
(
LC50 level) was much less pronounced.22 This screening
system is based on a short (48 h) exposure of tumor cells
to drugs and frequently is too short to fully reveal the
cytotoxic potency at physiologically relevant concentra-
tions. Because induction of apoptosis in solid tumor cells
often requires prolonged incubation, we used an ex-
tended drug exposure time (120 h) in a cytotoxicity
assay. This was a more suitable means for evaluating
the cell killing potency of the new chemicals. We
selected five human tumor cell lines (colon tumor HCT-
1
16, lung cancer A549, melanoma SK-MEL-2, breast
cancer MCF-7, and leukemia HL-60) from the NCI panel
and used them to test the target compounds in an MTT-
based cytotoxicity assay. The strong cytotoxic effect
detected at relatively low concentrations for 4ad by
MTT and LIVE/DEAD Viability/Cytotoxicity assays
suggests that this agent may trigger apoptosis in
sensitive cells. The occurrence of programmed cell death
can be documented by many criteria, such as endonu-
cleosomal DNA breakdown (DNA ladder), flow cytomet-
ric analysis of DNA content, induction of caspases, and
morphological changes in dying cells among others.
Usually, several different lines of evidence are required
to establish apoptosis. We selected 4ad as the most
active compound in the series to evaluate the apoptosis-
inducing properties. We used a concentration of 100 nM
Figure 7. In vivo activity of 4ad against human colon
adenocarcinoma HCT-116 xenografts implanted subcutane-
ously in the flanks of athymic nude mice. Tumor cells were
implanted 4 days prior to the initiation of treatment. The drug
in 5% dextrose solution at the concentrations indicated or 5%
dextrose control solution were administered iv 5 times every
third day (Q3Dx5 schedule). The experiment was terminated
at 38 days.
Discussion
4
ad, since that resulted in the complete killing of the
cell lines under consideration. One hallmark of apop-
tosis is the demonstration of a “ladder” pattern in
electrophoretic gels of nuclear DNA,26 where a portion
of nuclear DNA is fragmented to sizes equivalent to
mono- or oligonucleosomes in cells that are undergoing
apoptosis. We detected the formation of an apoptotic
ladder in HL-60 leukemia cells (Figure 4) but not in
HCT-116 colon tumor cells. The absence of DNA ladder
in HCT-116 cells does not indicate the absence of
apoptosis, since ladder formation occurs at a late stage
of cell death and its presence is very dependent on the
cellular model. This experiment also revealed that, in
contrast to camptothecin, which induced apoptosis in
HL-60 cells after only 5 h of treatment, 4ad at an
equivalent concentration required a much longer time
(72 h). Apoptotic cell death was also detected by flow
cytometric analysis of stained cell nuclei as well as by
The new unsymmetrical analogues of the BIAs,
especially compound 4ad, have substantially different
biological behavior from their symmetrical progenitors.
The symmetrical BIAs were purely cytostatic agents,
exhibiting no cytotoxicity, even in concentrations as high
as 10 µM. In contrast, the unsymmetrical compounds,
except 6, all showed significant cytotoxicity in the
sensitive cell lines. Biophysical measurements of bind-
ing of the unsymmetrical agents to DNA revealed
(
Tarasov et al., unpublished data) that the mode of
binding is similar to that of the one previously deter-
mined for the symmetrical BIAs,19 which was described
in the Introduction. We hypothesize that the difference
in activity of the two series is the result of the differ-
ences in the ability of the DNA-drug complexes to
“
hijack” relevant proteins involved in DNA repair and/
or transcription. Numerous, although circumstantial,

New Synthetic Agent Against Cancer
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 13 4479
the activation of caspase-3, a member of the interleukin
70-230 mesh). Melting points were determined on an electro-
thermal capillary melting point apparatus and are uncor-
1
â-converting enzyme family involved in the apoptotic
1
27
rected. H NMR spectra were recorded on a Varian VXR-S
cascade. As presented in Figure 4, HL-60 cells treated
spectrometer operating at 500 MHz. Elemental analyses were
performed by Atlantic Microlab, Inc., Norcross, GA.
with 4ad show fragmentation of the nuclei. After 72 and
96 h of drug treatment, 45 and 85% of cells, respectively,
Compounds 3ac and 3ad were prepared as described
were found to be undergoing apoptosis compared to 6%
in the control cells (Figure 5). Caspase-3 is one of the
key executioners of apoptosis, being responsible for the
proteolytic activation of DNA-cleaving endonucleases.28
Compound 4ad was also able to induce caspase-3
activity in the colon line HCT-116 as well as in the
leukemia line HL-60. Figure 6 presents the results of
these experiments in HCT-116 cells. A noticeable eleva-
tion of caspase-3 activity over the background level was
evident after 24 h of treatment, and the peak activity
11,21
earlier
and were purified by reverse phase semipreparatory
high-performance liquid chromatography on C-18 columns or
by automated flash chromatography on silica gel.
General Procedure for the Preparation of 4. A mixture
of 3-nitro-1,8-naphthalenedicarboxylic anhydride (0.001 mol)
and the corresponding substituted 5-amino-2,10b-diazaacean-
thrylen-6-one (3) was stirred at 80 °C in dimethylformamide
(
8 mL) until the reaction was completed (thin layer chroma-
tography). The precipitated solid was filtered, washed with
methanol, and dried and could be purified by column chroma-
tography or by crystallization to yield the corresponding
compound 4.
(
more than 10 times the level of untreated cells) was
observed after 48 h.
Example: 2-(3-{4-[3-(6-Oxo-6H-2,10b-diaza-aceanthryl-
en-5-ylamino)-propyl]-piperazin-1-yl}-propyl)-5-nitro-
2-aza-phenalene-1,3-dione (4ad). Orange crystals after
The in vitro activity of compound 4ad was recapitu-
lated in vivo. Figure 7 presents the results of an
experiment wherein athymic (nu/nuNCr) mice with
subcutaneous HCT-116 human colon cancer xenografts
were treated iv with various concentrations of 4ad.
Preliminary experiments with intraperitoneal (ip) ad-
ministration of 4ad into tumor-bearing mice showed no
antitumor activity, presumably because of first pass
clearance of the drug in the liver (ip injected 4ad had a
plasma half-life of ∼8 min). These results indicate that
the drug inhibited the growth of the xenograft following
crystallization from dimethylformamide-water: yield 82%; mp
1
2
27-230 °C; H NMR (CDCl
3
) 9.31 (d, 1H), 9.13 (d, 1H), 8.97
(
t, 1H), 8.76 (m, 1H), 8.56 (m, 1H), 8.55 (s, 1H), 8.42 (m, 1H),
7
2
2
.94 (m, 3H), 7.89 (m,1H), 7.54 (m, 1H), 6.78 (d, 1H), 4.29 (m,
H), 3.45 (qt, 2H), 2.53 (t, 2H), 2.48 (br m, 4H), 2.39 (t, 2H),
.34 (br m, 4H), 1.96 (m, 2H), 1.88 (m, 2H). Anal. (C36
33 7 5
H N O )
C, H, N.
Example: 2-(3-{Methyl-[3-(6-oxo-6H-2,10b-diaza-acean-
thrylen-5-ylamino)-propyl]-amino}-propyl)-5-nitro-2-aza-
phenalene-1,3-dione (4ac). Purified by silica gel column
chromatography using chloroform-methanol (8:1) mixture as
5
doses administered every third day. The absence of a
1
the eluent: yield 74%; mp 218-221 °C; H NMR (CDCl
3
) 9.23
dose response may be due to the toxicity of the drug.
The highest dose (40 mg/kg) was toxic to 6 out of 10
animals, the intermediate dose (20 mg/kg) resulted in
about a 10% body weight loss but no lethality, while
the lowest dose (10 mg/kg) was well tolerated and just
as effective as the higher doses. It is expected that
additional in vivo experiments, with doses based on
pharmacokinetic data, will allow us to determine an
optimal dosing regimen. Nevertheless, the present data
suggest that the drug is active in vivo and is relatively
well tolerated.
(
d, 1H), 9.04 (d, 1H), 8.97 (t, 1H), 8.71 (m, 1H), 8.50 (s, 1H),
8
1
.34 (m, 1H), 7.94 (d, 1H), 7.88 (m, 2H), 7.77 (m,1H), 7.49 (m,
H), 6.77 (d, 1H), 4.27 (m, 2H), 3.50 (qt, 2H), 2.56 (t, 4H), 2.30
(s, 3H), 1.95 (m, 4H). Anal. (C H N O ) C, H, N.
3
3
28
6
5
5-Amino-2-(3-{4-[3-(6-oxo-6H-2,10b-diaza-aceanthrylen-
5-ylamino)propyl]-piperazin-1-yl}-propyl)-2-aza-phena-
lene-1,3-dione (5ad). Stannous chloride(1.52 g, 0.008 mol)
dissolved in concentrated hydrochloric acid (5 mL) was added
to a stirred solution of 4ad (0.644 g, 0.001 mol) in glacial acetic
acid (25 mL). The mixture was stirred at 60 °C for 2 h. After
the mixture was cooled, acetone (50 mL) was added and it was
stirred vigorously. Precipitate was collected by filtration,
washed with acetone, and suspended in water (250 mL). The
suspension was made basic (pH ∼12) with sodium hydroxide,
and the product was extracted with chloroform (5 × 50 mL).
The crude product was chromatographed on silica gel column
with chloroform-methanol (10:1) mixture containing 0.5%
isopropylamine. The main fraction, after evaporation of sol-
vents, yielded 0.550 g (89%) of yellow 5ad: mp 219-222 °C;
Conclusion
The synthesis of unsymmetrical bifunctional agents,
derived from the previously described BIAs, resulted in
the discovery of a very potent but selective antitumor
agent, 4ad.This compound induced apoptosis in sensi-
tive cells at low nanomolar concentrations in vitro.
Preliminary data also indicated positive in vivo activity
in a human colon cancer xenograft in nude mice.
Compound 4ad is a DNA-binding agent. We had previ-
ously shown that the symmetrical BIAs bind to DNA
by intercalation and minor groove interaction. We
believe that compounds such as 4ad bind in a similar
fashion. However, DNA binding is insufficient for
biological activity. We hypothesize that compounds such
as 4ad exert their activity when the DNA-drug complex
is able to “capture” a critical DNA-binding protein in a
stable ternary complex. Ongoing studies will define the
mechanistic basis for the potent, but selective, cytotox-
icity of 4ad.
1
3
H NMR (CDCl ) 8.99 (m, 1H), 8.57 (m, 1H), 8.55 (s, 1H), 8.31
(m, 1H), 8.02 (d, 1H), 7.97 (m, 1H), 7.92 (m, 2H), 7.80 (m, 1H),
7.60 (m, 1H), 7.54 (d, 1H), 7.29 (m, 1H), 6.79 (d, 1H), 4.24 (m,
2
2
H), 4.17 (s, 2H), 3.46 (qt, 2H), 2.51 (t, 2H), 2.48 (br m, 8H),
.42 (t, 2H), 2.34 (br m, 4H), 1.93 (m, 2H), 1.87 (m, 2H). Anal.
1
(
36
C H
35
7
N O
3
)‚ /
2
H
2
O, C, H, N.
5-(3-{4-[3-(6-Chloro-2-methoxy-acridin-9-ylamino)-pro-
pyl]-piperazin-1-yl}-propylamino)-2,10b-diaza-aceanthryl-
en-6-one (6). A mixture of 3ad (0.419 g, 0.001 mol), 6-chloro-
2
-methoxy-9-phenoxyacridine (0.364 g, 0.001 mol), and phenol
(4 g) was heated at 90 °C for 12 h. After the mixture was
cooled, 100 mL of chloroform was added and then poured into
100 mL of 1 N sodium hydroxide and crushed ice. The mixture
was shaken, and the chloroform layer was separated, con-
densed, and chromatographed on a silica gel column with
chloroform-methanol (5:1) mixture as an eluent. The major
1
fraction gave 0.383 g (58%) of yellow 6: mp 188-192 °C; H
Experimental Section
3
NMR (CDCl ) 9.05 (t, 1H), 8.55 (s, 1H), 8.54 (m, 1H), 8.15 (d,
Chemical Synthesis. General. All of the commercial
solvents and reagents were used without further purification.
Column chromatography was performed on silica gel (Aldrich,
1H), 8.04 (d,1H), 7.99 (d, 1H), 7.97 (d, 1H), 7.93 (d, 1H), 7.79
(m, 1H), 7.54 (m, 1H), 7.42 (m, 2H), 7.28 (m, 1H), 6.82 (d, 1H),
3.95 (s, 3H), 3.89 (m, 2H), 3.54 (m, 2H), 2.75 (t, 2H), 2.52 (m,

4
480 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 13
Cholody et al.
1
0H), 2.42 (t, 2H), 1.96 (m, 2H), 1.65 (m,2H). Anal. (C38
H
38
-
in Tris-acetate buffer (TAE) at 25 V for approximately 6 h.
After electrophoresis, DNA was visualized by ethidium bro-
mide staining (0.5 µg/mL). The gel was viewed using a UV
transilluminator and photographed using Polaroid Polapan 57
film.
1
ClN
7
O
2
)‚ /
2
H
2
O, C, H, N.
Biological Studies. Chemicals. All of the mammalian cell
culture reagents, DNA Typing Grade Agarose, DNA Typing
Grade TAE (Tris-Acetate/EDTA) buffer, ethidium bromide,
Tris-HCl, and Trypan Blue Stain were purchased from GIBCO-
BRL (Grand Island, NY). Tissue culture materials were
purchased from Corning Inc. (Acton, MA). Ammonium acetate,
DAPI ′, dimethyl sulfoxide (DMSO), ethylenediaminetetraace-
tic acid (EDTA), gel loading solution, methanesulfonic acid,
paraformaldehyde, proteinase K, RNase A, sodium chloride,
and sodium dodecyl sulfate (SDS) were from Sigma-Aldrich
DAPI Staining and DNA Content Analysis. At a desired
time, untreated and drug-treated cells were harvested by
centrifugation, rinsed with phosphate buffered saline (PBS),
and fixed on ice for 15 min with 2% paraformaldehyde. After
washes with PBS containing 1% FBS, the cells were fixed in
70% alcohol and stored at -20 °C. When all of the samples
were collected, they were rinsed with ice-cold PBS supple-
mented with 10% FBS, stained with DAPI (4′,6-diamino-2-
phenylindole) in the dark for 30 min (final concentration 20
µg/mL), washed again, and analyzed by FACS. The blue
fluorescence of the DNA-bound DAPI of individual cells was
measured with a FACStar Plus Becton Dickinson Immuno-
cytometry System (San Jose, CA).
(
St. Louis, MO). Polaroid Polopan film was from the Polaroid
Corporation (Cambridge, MA).
Cell Culture. Human colon carcinoma HCT-116, human
melanoma SK-MEL-2, human small cell lung cancer A549,
human leukemia HL-60, and human breast cancer MCF-7 cells
were purchased from the American Type Culture Collection
(
Rockville, MD). HCT-116, SK-MEL-2, and A549 cells were
For morphological examination, 50 µL of cell suspension
from each sample was dropped onto a slide, air-dried, mounted
with Gel/Mount (Biomeda, Foster City, CA), visualized, and
photographed with a Zeiss LSM 310 (Carl Zeiss, Jena, Ger-
many) confocal microscope.
grown in Dulbecco’s modified Eagle’s medium. HL-60 and
MCF-7 cells were cultured in RPMI 1640 medium. Media were
supplemented with 10% heat inactivated fetal bovine serum
(
FBS), 2 mM L-glutamine, 100 U/mL of penicillin, and 100 µg/
mL of streptomycin. Cells were maintained in a humidified
incubator at 37 °C in 5% CO atmosphere.
Analysis of Caspase-3 Activity. Intracellular caspase-3
activity was measured using a Caspase-3 Assay Kit (Pharm-
ingen, San Diego, CA) according to the manufacturer’s manual
2
Cytotoxicity Assay. Cells were seeded in six wells for each
studied concentration into 96-well microtiter plates (100 µL
of medium containing 1000-1500 cells per well) and were
precultured for 1 day. Stock solutions (2.5 mM) of test
compounds were prepared freshly by dissolving their free base
forms in a distilled water-DMSO (50:50) mixture containing
2
9
as described previously. Briefly, tested and untreated cells
were harvested at the indicated times and washed with PBS.
Pelleted cells were treated with a lysis buffer provided in the
6
kit (10 × 10 cells/mL). Cell lysate (100 µL) was added to the
reaction tubes, one containing 10 µL of reconstituted Ac-
DEVD-AMC (caspase-3 fluorogenic substrate) in 1 mL of
HEPES buffer and a second with the same amount of Ac-
DEVD-AMC and 10 µL of Ac-DEVD-CHO (caspase-3 inhibitor)
in 1 mL of HEPES the buffer. Reaction mixtures were
incubated for 1 h at 37 °C, and the emitted fluorescence from
AMC that was cleaved and released by active caspase-3 was
quantified by a UV specrofluorometer (FluoroMax2) with an
excitation of 380 nm and an emission of 440 nm.
3
equiv of methanesulfonic acid and further diluted with
distilled water to the concentration of 500 µM. These solutions
were used to prepare 2 µM of working solution and its 10-fold
serial dilutions in appropriate media. An amount of 100 µL of
drug containing medium or vehicle (control) was added to each
well. The cytotoxicity was determined by the MTT-based,
CellTiter96 Non-Radioactive Cell Proliferation Assay (Prome-
ga, Madison, WI) according to instructions provided by the
supplier of the kit. While the drugs were added, assays were
performed on extra reference plates to determine the cell
In Vivo Experiments. HCT-116 human tumor fragments
(3 mm × 3 mm) were obtained from donor mice and implanted
subcutaneously into the experimental athymic (nu/nuNCr)
mice using a tumor implant trocar (Popper and Sons, New
Hyde Park, NY). Treatment was initiated 4 days following
implantation. Mice were housed in an AAALAC-accredited
barrier facility with sterilized chow and hyperchlorinated
water provided ad libitum. Tumor growth was monitored using
a twice or thrice weekly caliper measurement of the tumor in
two dimensions. These measurements were converted to tumor
population density at time 0 (T
0
). After 120 h of incubation at
, the
3
7 °C in a humidified atmosphere containing 5% CO
2
assays were performed on test (T) and control (C) cells. The
absorbance of the wells was determined at 544 nm by a
FLUOstar/POLARstar Galaxy (BMG Labtechnologies GmbH)
MicroplateReader. Cellular responses were calculated from the
2
3
data as described previously: 100 × [(T - T
and 100 × [(T - T )/T ] for T < T0.
LIVE/DEAD Viability/Cytotoxicity Assay. This experi-
0 0
)/(C - T )] for T
>
T
0
0
0
weights using a formula for a prolate ellipsoid: [(length in mm)
2
×
(width in mm) ]/2 ) tumor weight (mg).
ment was performed according to the manufacturer’s fluores-
cence microscopy protocol provided with the kit (Molecular
Probes, Eugene, OR).
DNA Fragmentation. At specified times, untreated or
drug-treated cells were collected by centrifugation at 2500g
for 5 min. The resulting cell pellet (approximately 2 × 10 cells)
Acknowledgment. We thank the Structural Bio-
physics Resource Laboratory, NCIsFrederick for the
use of the mass spectrometer. We are grateful to Dr.
Dominic Scudiero, Developmental Therapeutics Pro-
gram, NCI for the NCI 60 cell line assay data and Dr.
Sergey Tarasov for the discussions.
6
was resuspended in a lysis buffer (0.5 M Tris, pH 8.0; 20 mM
EDTA; 10 mM NaCl; 1% SDS). Lysates were transferred to
Eppendorf tubes and incubated overnight at 37 °C with
proteinase K (final concentration 0.5 mg/mL). The next day
one-fourth of the volume of saturated NaCl solution (ap-
proximately 6 M) was added to the warm lysate. It was cooled
in an ice slurry and centrifuged for 30 min at 500g. The
supernatant was poured into a fresh tube and 0.2 vol of 11 M
ammonium acetate, and 2 vol of cold 96% ethanol were added,
mixed gently by inversion, and kept at -20 °C for several
hours. After centrifugation at 14500g for 15 min at 4 °C, the
pellet was washed twice in cold 70% ethanol, dried at room
temperature, and dissolved in TE buffer (Tris-HCl pH 8.0,
EDTA pH 8.0, NaCl), followed by treatment with DNase-free
RNase A (100 µg/mL) for 3 h at 37 °C. DNA was quantitated
using UV spectroscopy at an optical densityof 260 nm. DNA
Supporting Information Available: Elemental analysis
data for the compounds 4ac, 4ad, 5ad, and 6 and NCI 60
human tumor cell line screening data for 4ac and 4ad. This
material is available free of charge via the Internet at http://
pubs.acs.org.
References
(1) Schmitt, C. A.; Lowe, S. W. Apoptosis and therapy. J. Pathol.
1
999, 187, 127-137.
(
2) Sen, S.; D’Incalci, M. Apoptosis. Biochemical events and rel-
evance to cancer chemotherapy. FEBS Lett. 1992, 307, 122-
1
27.
(
10 µg) was mixed with loading buffer and transferred to the
(
3) Hannun, Y. A. Apoptosis and the dilemma of cancer chemo-
gel. Electrophoresis was then performed in 1.8% agarose gel
therapy. Blood 1997, 89, 1845-53.

New Synthetic Agent Against Cancer
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 13 4481
(
4) Mesner, P.; Budihardjo, I.; Kaufman, S. H. Chemotherapy-
induced apoptosis. Adv. Pharmacol. 1997, 41, 461-499.
5) Kamesaki, H. Mechanisms involved in chemotherapy-induced
apoptosis and their implications in cancer chemotherapy. Int.
J. Hematol. 1998, 68, 29-43.
(18) Gaugain, B.; Barbet, J.; Oberlin, R.; Roques, B. P.; Le Pecq, J.-
B. DNA Bifunctional Intercalators. 1. Synthesis and Conforma-
tional Properties of an Ethidium Homodimer and of an Acridine
Ethidium Heterodimer. Biochemistry 1978, 17, 5071-5078.
(19) Tarasov, S. G.; Casas-Finet, J. R.; Cholody, W. M.; Kosakowska-
Cholody, T.; Gryczynski, Z. K.; Michejda, C. J. Bisimidazoacri-
dones: 2. Steady-state and time-resolved fluorescence studies
of their diverse interactions with DNA. Photochem. Photobiol.
(
(
(
(
6) Kaufman, S. H.; Earnshaw, W. C. Induction of apoptosis by
cancer chemotherapy. Exp. Cell Res. 2000, 256, 42-49.
7) Debatin, K.-M. Activation of apoptotic pathways by anticancer
treatment. Toxicol. Lett. 2000, 112-113, 41-48.
2003, 78, 313-322.
8) Nicotera, P.; Leist, M.; Single, B.; Volbracht, C. Execution of
apoptosis: converging or diverging pathways? Biol. Chem. 1999,
(
20) Segal-Bendirdjian, E.; Coulaud, D.; Roques, B. P.; Le Pecq, J.
B. Selective loss of mitochondrial DNA after treatment of cells
with ditercalinium (NSE335153), an antitumor bis-intercalating
agent. Cancer Res. 1988, 48, 4982-4992.
21) Tarasov, S. G.; Casas-Finet, J. R.; Cholody, W. M.; Michejda, C.
J. Bisimidazoacridones: Effect of molecular environment on
conformation and photophysical properties. Photochem. Photo-
biol. 1999, 70, 568-578.
380, 1035-1040.
(
9) Blagosklonny, M. V. Cell death beyond apoptosis. Leukemia
2
000, 14, 1502-1508.
(
(
10) Formigli, L.; Papucci, L.; Tani, A.; Schiavone, N.; Tempestini,
A.; Orlandini, G. E.; Capaccioli, S.; Zecchi Orlandini, S. Apone-
crosis: morphological and biochemical exploration of a syncretic
process of cell death sharing apoptosis and necrosis. J. Cell.
Physiol. 2000, 182, 41-49.
11) Cholody, W. M.; Hernandez, L.; Hassner, L.; Scudiero, D. A.;
Djurickovic, D. B.; Michejda, C. J. Bisimidazoacridones and
Related Compounds: New Antineoplastic Agents with High
Selectivity against Colon Tumors. J. Med. Chem. 1995, 38,
(
22) Results from the standard NCI screening can be accessed at the
Developmental Therapeutic Program, NCI/NIH web site (http://
dtp.nci.nih.gov) with the Search by NSC number option.
23) Monks, A.; Scudiero, D.; Skehan, P.; Shoemaker, R.; Paull, K.;
Vistica, D.; Hose, C.; Langley, J.; Cronise, P.; Vaigro-Wolff, A.
Feasibility of a high-flux anticancer drug screen using a diverse
panel of cultured human tumor cell lines. J. Natl. Cancer Inst.
(
(
3043-3052.
(
12) Hernandez, L.; Cholody, W. M.; Hudson, E. A.; Resau, J. H.;
Pauly, G.; Michejda, C. J. Mechanism of Action of Bisimida-
zoacridones, New Drugs with Potent, Selective Activity against
Colon Cancer. Cancer Res. 1995, 55, 2338-2345.
13) Turpin, J. A.; Buckheit, Jr. R. W.; Derse, D.; Hollingshead, M.;
Williamson, K.; Palamone, C.; Osterling, M. C.; Hill, S. A.;
Graham, L.; Schaeffer, C. A.; Bu, M.; Huang, M.; Cholody, W.
M.; Michejda, C. J.; Rice, W. G. Inhibition of acute, latent and
chronic phase HIV-1 replication by a bistriazoloacridone ana-
logue that selectively inhibits HIV-1 transcription. Antimicrob.
Agents Chemother. 1998, 42, 487-494.
1991, 83, 757-766.
(
24) Abad, A.; Gravalos, C.; Font, A.; Molina, F.; Diaz-Puente, M. T.;
Fabrega, X.; Benavides, A.; Martin, M. Phase II study of
Mitonafide in advanced and relapsed colorectal cancer. Invest.
New Drugs 1996, 14, 223-225.
(
(
25) Casado, A,; Rosell, R.; Garcia-Gomez, R.; Diaz-Rubio, E.; Perez-
Manga, G.; Font, A.; Benavides, A.; Martin, M. Phase II study
of mitonafide in non-small cell lung cancer (NSCLC). Invest. New
Drugs 1996, 14, 415-417.
(26) Gong, J.; Traganos, F.; Darzynkiewicz. Z. A selective procedure
for DNA extraction from apoptotic cells applicable for gel
elecrophoresis and flow cytometry. Anal. Biochem. 1994, 218,
314-319.
(27) Aranha, O.; Gringnon, R.; Fernandes, N.; McDonnell, T. J.; Wood,
D. P.; Sarkar, F. Suppression of human prostate cancer cell
growth by ciprofloxacin is associated with cell cycle arrest and
apoptosis. Int. J. Oncol. 2003, 22, 787-794.
(28) Cohen, G. M. Caspases: the executioners of apoptosis. Biochem.
J. 1997, 326, 1-16.
(29) Cholody, W. M.; Kosakowska-Cholody, T.; Michejda, C. J.
Bisimidazoacridones induce a potent cytostatic effect in colon
tumor cells that sensitizes them to killing by UCN-01. Cancer
Chemother. Pharmacol. 2001, 47, 241-249.
(
(
(
(
14) Canellakis, E. S.; Shaw, Y. H.; Hanners, W. E.; Schwartz, R. A.
Diacridines: bifunctional intercalators. I. Chemistry, physical
chemistry and growth inhibitory properties. Biochim. Biophys.
Acta 1976, 418, 277-289.
15) Wright, R. G. McR.; Wakelin, L. P. G.; Fieldes, A.; Acheson, R.
M.; Waring, M. J. Effects of Ring Substituents and Linker
Chains on the Bifunctional Intercalation of Diacridines into
Deoxyribonucleic Acid. Biochemistry 1980, 19, 5825-5836.
16) Markovits, J.; Garbay-Jaureguiberry, C.; Roques, B. P.; Le Pecq,
J.-B. Acridine dimers: influence of the intercalating ring and of
the linking-chain nature on the equilibrium and kinetic DNA-
binding parameters. Eur. J. Biochem. 1989, 180, 359-366.
17) Le Pecq, J.-B.; Le Bret, M.; Barbet, J.; Roques, B. P. DNA
polyintercalating drugs: DNA binding of diacridine derivatives.
Proc. Natl. Acad. Sci. U.S.A. 1975, 72, 2915-2919.
JM048946X