Synthesis and cytotoxicity of 2-methyl-4,9-dihydro-1-substituted-1H- imidazo[4,5-g]quinoxaline-4,9-diones and 2,3-disubstituted-5,10-pyrazino[2,3- g]quinoxalinediones

Article abstract of DOI:10.1021/jm970695n

Series of 2-methyl-4,9-dihydro-1H-imidazo[4,5-g]quinoxaline-4,9-diones and 5,10-pyrazino[2,3-g]quinoxalinediones have been synthesized from 6,7- dichloro-5,8-quinoxalinedione for developing new anticancer drugs. Intercalation complex of 2-methyl-4,9-dihyd

Full text of DOI:10.1021/jm970695n

4716
J . Med. Chem. 1998, 41, 4716-4722
Syn th esis a n d Cytotoxicity of
2-Meth yl-4,9-d ih yd r o-1-su bstitu ted -1H-im id a zo[4,5-g]qu in oxa lin e-4,9-d ion es a n d
2,3-Disu bstitu ted -5,10-p yr a zin o[2,3-g]qu in oxa lin ed ion es
Hee-Won Yoo,†,‡ Myung-Eun Suh,† and Sang Woo Park*,‡
Division of Medicinal Chemistry, College of Pharmacy, Ewha Womans University, Seoul 120-750, Korea, and Medicinal
Chemistry Research Center, Korea Institute of Science and Technology, P.O. Box 131, Cheongryang, Seoul 130-650, Korea
Received October 14, 1997
Series of 2-methyl-4,9-dihydro-1H-imidazo[4,5-g]quinoxaline-4,9-diones and 5,10-pyrazino[2,3-
g]quinoxalinediones have been synthesized from 6,7-dichloro-5,8-quinoxalinedione for developing
new anticancer drugs. Intercalation complex of 2-methyl-4,9-dihydro-1-methyl-1H-imidazo-
[4,5-g]quinoxaline-4,9-dione with GC/GC base pairs by a computer graphics-aided method was
parallel to the axis of the helix. The interaction energies of synthetic compounds were low.
2-Methyl-4,9-dihydro-1-(4-bromophenyl)-1H-imidazo[4,5-g]quinoxaline-4,9-dione, which has the
lowest interaction energy of the test compounds, was identified as being more cytotoxic against
human gastric adenocarcinoma cells (MKN 45) than adriamycin and cis-platin in vitro using
the MTT assay. The IC50 value of this compound was 1.30 µM, whereas those of adriamycin
and cis-platin were 3.13 and 86.5 µM, respectively. The cytotoxicity of 2,3-diethyl-5,10-pyrazino-
[2,3-g]quinoxalinedione was comparable to that of adriamycin in the in vitro assay. However
these compounds showed loss of activity on human lung adenocarcinoma cells (PC 14) and
human colon adenocarcinoma cells (colon 205). These results suggest that 2-methyl-4,9-dihydro-
1-(4-bromophenyl)-1H-imidazo[4,5-g]quinoxaline-4,9-dione, with the lowest interaction energy,
might be an excellent and selective antitumor agent against MKN 45.
In tr od u ction
Streptonigrin (1), obtained from Streptomyces floccu-
lus, was known as an antitumor agent by Rao and
Cullen in 1959,1 but application is limited because of
its toxicity.2,3 The 7-amino-6-methoxy-5,8-quinolinedi-
one moiety in 1 is responsible for antitumor activity.4,5
Studies on the activity of heterocyclic quinones contain-
ing nitrogen such as quinolinedione showed that the
number and position of nitrogens are considerably
important for the cytotoxicity.6 Thus the diazanaphtho-
quinones were proved to be the most active compounds
in comparison with naphthoquinone and quinolinedione.
It was published that 5,8-quinoxalinediones that have
one additional nitrogen in the nucleus different from
quinolinedione showed antitumor activity.7 Another
structural requirement for the antitumor activity is the
p-quinone moiety in the nonheterocyclic ring, whereas
o-quinone gave decreased activity.8,9 The electron-
withdrawing groups at the 6 and 7 positions of quino-
linediones also contributed to the activity,10 and more
annelated heterocyclic quinones were reported to in-
crease the antitumor activity.11 Giorgi-Renault pre-
pared benzoquinoxalinediones and pyridoquinoxalinedi-
ones, which are quinoxalinediones fused with benzene
or pyridine, and examined their antitumor activities.12
These tricyclic quinones exhibited cytotoxicity; espe-
cially 2,3-bis(bromomethyl)-5,10-benzo[g]quinoxalinedi-
one (2) was highly active against sarcoma 180. This led
us to think that replacement of the benzene ring of
benzoquinoxalinedione with pyrazine or imidazole het-
erocyclic rings might improve the activity. We report
here the synthesis of chemical modifications on hetero-
cycles like imidazoquinoxalinedione and pyrazinoqui-
noxalinedione derivatives. In a previous paper13 6,7-
dichloro-5,8-quinoxalinedione (3) was prepared in rela-
tively good yield, while synthesis of imidazoquinoxa-
linediones and pyrazinoquinoxalinediones had not yet
been described. 6,7-Dichloro-5,8-quinoxalinedione was
substituted with nucleophiles easily. Therefore we
could obtain heterocyclic compounds with more nitro-
gens from 6,7-dichloro-5,8-quinoxalinedione as a start-
ing material.
One of the cytostatic action mechanisms of coplanar
polycyclic compounds is their intercalation with human
DNA. This caused enzymatic blockade and reading
errors during the replication process.14 If the com-
pounds have three to four rings which are coplanar, they
would give the optimal intercalation. The intercalation
of a synthetic compound between two base pairs and
their interaction energies were investigated by computer-
aided molecular modeling. The synthetic compounds
described here are coplanar annelated quinoxalinedione
* Corresponding author: Sang Woo Park. Phone: 82-2-958-5132.
Fax: 82-2-958-5189.
† Ewha Womans University.
‡ Korea Institute of Science and Technology.
10.1021/jm970695n CCC: $15.00 © 1998 American Chemical Society
Published on Web 10/24/1998

Synthesis and Cytotoxicity of Quinoxalinediones
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 24 4717
Sch em e 1
Sch em e 2
analogues and might interact with DNA. On the basis
of this consideration, they might be expected to possess
antitumor activity. In order to examine the effect of side
chains, the alkyl and aryl groups were substituted on
the nitrogen nucleus.
Resu lts An d Discu ssion
Syn th etic Ch em istr y. 6-Azido-7-chloro-5,8-quinox-
alinedione (4) was obtained by treating 6,7-dichloro-5,8-
quinoxalinedione (3) with sodium azide in acetic acid
and then was reduced by sodium borohydride to 6-amino-
7-chloro-5,8-quinoxalinedione (5). When 6,7-dichloro-
5,8-quinoxalinedione was treated with ammonia, it was
reduced to 6,7-dichloro-5,8-dihydroxyquinoxaline. This
was confirmed by acetylation of 6,7-dichloro-5,8-dihy-
droxyquinoxaline which gave 6,7-dichloro-5,8-diacetoxy-
quinoxaline identified from NMR and MS spectrometry.
Acetylation of 5 produced 6-acetamido-7-chloro-5,8-
quinoxalinedione (6). 6-Acetamido-7-amino-5,8-quinox-
alinedione (7) was prepared by reaction of (acylamino)-
quinone with sodium azide and reduction with sodium
borohydride. 6-Acetamido-7-[(4-bromophenyl)amino]-
5,8-quinoxalinedione (8) was obtained by reacting com-
pound 6 with 4-bromoaniline in ethanol.
The conversion of the 6-acetamido-7-amino-5,8-quinox-
alinedione (7) and 6-acetamido-7-[(4-bromophenyl)-
amino]-5,8-quinoxalinedione (8) to the corresponding
imidazoles (9, 14) was readily performed under acidic
or basic conditions by the method of Fries and Billig.15
However 2-methyl-4,9-dihydro-1-alkyl-1H-imidazo[4,5-
g]quinoxaline-4,9-diones (10-12) and 2-methyl-4,9-di-
hydro-1-phenyl-1H-imidazo[4,5-g]quinoxaline-4,9-di-
one (13) were prepared directly by reaction of 6-acet-
amido-7-chloro-5,8-quinoxalinedione (6) with alkylamine
and phenylamine without catalyst (Scheme 1).
Reaction of diaminoquinoxalinedione (16) with glyoxal
in water18,19 gave 5,10-pyrazino[2,3-g]quinoxalinedione
(17). 2,3-Dialkyl-5,10-pyrazino[2,3-g]quinoxalinediones
(18-20) were prepared by reaction of diaminoquinoxa-
linedione with 1,2-diketones in 50% aqueous acetic
acid.20 Acetic acid was necessary in this reaction.
Attempts to prepare pyrazinoquinoxalinedione deriva-
tives without acid21 were unsuccessful. 2,3-Diphenyl-
5,10-pyrazino[2,3-g]quinoxalinedione (21) was obtained
by reaction of 6,7-diamino-5,8-quinoxalinedione (16)
with benzil in ethanol using concentrated sulfuric acid
as a catalyst (Scheme 2). Reaction of diaminoquinoxa-
linedione with 1,4-dibromo-2,3-butanedione did not give
2,3-bis(dibromomethyl)-5,10-pyrazino[2,3-g]quinoxalinedi-
one but 2,3-dimethyl-5,10-pyrazino[2,3-g]quinoxalinedi-
one (18). The bromomethylated quinone with, the great
lability of bromine atoms, easily reacted with nucleo-
philes like water.12 Dehalogenation occurred under this
condition using 50% aqueous acetic acid as a solvent.
Molecu la r Mod elin g of In ter ca la tion Com p lexes
a n d Th eir In ter a ction En er gies. Intercalation of
compounds with human DNA is the insertion of the
planar part of a molecule between two stacked base
pairs. Hydrogen bonds make a major contribution to
the stability of the complex. The primary and secondary
structures of DNA in the complex remain unchanged.
However the DNA tertiary structure in the intercalation
complex is somewhat lengthened and unwound in
comparison with the original structure.22 Complete
energy maps can be represented by molecular modeling,
and it is possible to follow molecular changes visually
With respect to the pyrazinoquinoxalinedione deriva-
tives, 6,7-diamino-5,8-quinoxalinedione (16) was syn-
thesized by reduction of diazidoquinoxalinedione (15)
with sodium borohydride which was produced by react-
ing 6,7-dichloro-5,8-quinoxalinedione (3) with sodium
azide in dimethylformamide.16 6,7-Diamino-5,8-qui-
noxalinedione (16) was also obtained by deacetylation
of 6-acetamido-7-amino-5,8-quinoxalinedione (7) under
basic conditions.17

4718 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 24
Yoo et al.
Ta ble 1. Interaction Energies (∆E) of DNA-Intercalator
complexes
intercalator
MAXIMIN2
energy (kcal/mol)
DNA-intercalator
complex energy
(kcal/mol)
interaction
energy
(kcal/mol)
compd
no.
10
11
12
14
18
19
20
39.232
37.212
41.228
39.222
14.695
6.478
27.307
29.187
29.881
24.574
3.678
-11.925
-8.025
-11.347
-14.648
-11.017
-10.929
-14.464
-4.457
0.836
15.300
and to calculate defined interaction energies. This
would allow us to expect their activities. Molecular
spectroscopic methods have contributed considerably to
the molecular resolution of an intercalation complex.
The molecule should have three to four planar rings
(minimum surface area of 28 Å2), and the intercalation
complex should be parallel to the axis of the helix for
an ideal intercalation.23 Normally the compound which
has a lower interaction energy is more cytotoxic than
the compound which has a higher interaction energy.
This is true only when binding energy is more important
than other factors such as cellular uptake and bioacti-
vation.
At first, a theoretical study from the intercalation
complex of synthetic compounds with cytidylyl(3′-5′)-
guanosine was examined using molecular mechanics
method, here especially MAXIMIN2 force field in the
Sybyl software program (hardware: Indy, 6.2 version,
Tripos Co., St. Louis, MO). The structures of com-
pounds were constructed with the option Crysin in
Sybyl, and the low-energy conformers of compounds
were determined by the standard Sybyl energy mini-
mizer, MAXIMIN2, which was used under default
conditions. The cocrystal structure of the synthetic
compound with cytidylyl(3′-5′)guanosine was built on
a molecular graphics system using the geometry data
of a GC base pair model taken from the published
structure,24 and then geometric optimization of the
complex was carried out. The partial charges used in
the calculations were generated by the Gasteiger-
Hu¨ckel method using the Tripos force field in the Sybyl
molecular modeling package. Gradient norm was 0.01
kcal/mol, and 400 iterations were performed for the
optimization. Each unit of the structure consists of one
synthetic compound and two dinucleoside monophos-
phate CpG molecules. The solvent and counterions
were not used in the models. The interaction energies
of DNA-intercalator complexes are summarized in
Table 1.
F igu r e 1. Molecular structure of 2-methyl-4,9-dihydro-1-
methyl-1H-imidazo[4,5-g]quinoxaline-4,9-dione (10) with cyti-
dylyl(3′-5′)guanosine.
complexes. Interestingly the interaction energy of
2-methyl-4,9-dihydro-1-(4-bromophenyl)-1H-imidazo-
[4,5-g]quinoxaline-4,9-dione (14) substituted with a
4-bromophenyl at the 1 position is -14.648 kcal/mol
which is the lowest energy examined. 2,3-Diethyl-5,10-
pyrazino[2,3-g]quinoxalinedione (20), substituted with
diethyl groups at the 2 and 3 positions, also has an
interaction energy of -14.464 kcal/mol. This suggests
that 14 and 20 with bulky side chains would interact
with DNA more strongly than the others. Thus they
would form stable intercalation complexes and might
have high activity.
The compound 10, as a representative of synthetic
compounds, has been tested to see if it interacts with
DNA, using ultraviolet spectrophotometric means and
an unwinding test.25 The UV spectroscopic properties
of 10 in the presence and absence of DNA were studied
by conventional optical spectroscopy. Characterization
of this mixture was done by comparing the ultraviolet
absorption spectrum of the complex with the spectrum
of the sample. The DNA-bound 10 showed hyper-
chromism and a red shift in the absorption band. These
spectral changes suggest binding of the compound on
the base pairs (data not shown). In general the ability
of a drug to unwind circular supercoiled DNA molecules
is also considered to be one of the sensitive criterion for
intercalative binding. If the substance intercalates with
DNA, it induces negative supercoiled DNA unwinding
characteristic of DNA intercalators, and then DNA
becomes supercoiled positively. The unwinding test of
10 using electrophoresis with pGEm7GF plasma DNA
and topoisomerase I indicated that 10 might intercalate
with DNA. The unwinding of circular DNA by 10 was
detected at 100 µM (data not shown). Amsacrine (N-
[4-(9-acridinylamino)-3-methoxyphenyl]methane sul-
fonamide) which intercalates DNA in the range of 2.5-
50 µM was used as a control.
A modeling study of the 2:1 complex between cytidyl-
yl(3′-5′)guanosine and 2-methyl-4,9-dihydro-1-methyl-
1H-imidazo[4,5-g]quinoxaline-4,9-dione (10) has led to
a general DNA binding model in which the compound
chromophore intercalates between base pairs (Figure 1).
As expected the DNA intercalation complex is parallel
to the axis of the helix, and the synthetic compound is
perpendicular to the line joining the base pairs. This
intercalator, which is the coplanar compound with three
rings, especially shows the advantageous arrangement,
and its interaction energy is -11.925 kcal/mol. We
could see only hydrogen atoms out of plane. It proves
that the coplanar imidazoquinoxalinediones and pyrazi-
noquinoxalinediones would make good intercalation
Cytotoxicity. MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide) assay developed by Mos-
mann26 is a rapid and precise colorimetric tetrazolium
assay for mammalian cell survival. The test compounds
were dissolved in DMSO at a concentration of 1 mM
and diluted with RPMI-FCS to various concentrations,
respectively. Human lung adenocarcinoma cell lines

Synthesis and Cytotoxicity of Quinoxalinediones
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 24 4719
Ta ble 2. Cytotoxicity Data on Human Lung Adenocarcinoma
Cell Line (PC 14) and Human Gastric Adenocarcinoma Cell
Line (MKN 45)
these two compounds are -14.648 and -14.464 kcal/
mol, and these values were the lowest ones for the tested
compounds. These results may support that the com-
pounds with lower interaction energies show higher
activity. Interestingly the quinoxalinediones with sub-
stituted bulky groups were more potent, which suggests
that enhancement of molecular size might contribute
to the cytotoxicity. However 2,3-diphenyl-5,10-pyrazino-
[2,3-g]quinoxalinedione (21) bearing two phenyl groups
is not potently active against MKN 45. The poor
solubility of this compound seems to cause this low
activity.
In the previous section, we concluded that 2-methyl-
4,9-dihydro-1-methyl-1H-imidazo[4,5-g]quinoxaline-4,9-
dione (10), as a representative of synthetic compounds,
intercalates with DNA. Unfortunately this compound
with an IC50 value of 34.12 µM, which showed almost
10-fold loss of activity compared to adriamycin but is
2.5 times more active than cis-platin, turned out to not
be significantly cytotoxic against MKN 45. 2-Methyl-
4,9-dihydro-1-(4-bromophenyl)-1H-imidazo[4,5-g]qui-
noxaline-4,9-dione (14) was considerably less active than
adriamycin and cis-platin against human lung adeno-
carcinoma cell lines (PC 14). This indicates that 14
could be more selective for MKN 45 than for PC 14 and
colon 205. All of the synthetic compounds did not show
high cytotoxicity on human colon adenocarcinoma cells
(colon 205). The pyrazinoquinoxalinediones substituted
with alkyl groups at the 2 and 3 positions exhibited
comparable cytotoxicity relative to the imidazoquinoxa-
linediones.
IC50 (µM)a
compd no.
PC 14
MKN 45
9
10
11
12
13
14
17
18
82.50
38.68
130.25
78.32
125.4
45.50
76.23
74
67.52
20.49
119.6
0.29
19.81
34.12
93.14
61.13
22.83
1.30
20.08
16.67
12.99
7.61
19
20
21
82.52
3.13
86.5
adriamycin
cis-platin
3.87
a
Values are the average from three independent dose-response
curves; variation was generally (15%.
(PC 14), human gastric adenocarcinoma cell lines (MKN
45), and human colon adenocarcinoma cell lines (colon
205) were used for the cytotoxicity test in vitro. Single-
cell suspensions were treated with each sample, adria-
mycin, and cis-platin in 96-well flat-bottom microplates,
and the plates were incubated at 37 °C for 4 days in a
5% CO2 incubator. The samples in RPMI-FCS were
used as a blank. MTT solution was added to cells, the
plates were incubated at 37 °C for an additional 4 h in
a 5% CO2 incubator, and then acid-isopropanol was
added to all wells. The optical densities of plates were
observed by a Microplate reader (Titertek, Multiscan)
at 540 nm. Some of the compounds were colored;
however, they did not show absorption at 540 nm in the
highest concentration for testing. Fractional inhibition
is calculated as 1 - (the ratio of optical density of sample
and control); 50% inhibition concentration (IC50) is
defined as the compound concentration required to
reduce the viability of human adenocarcinoma cells by
50%. The IC50 values for each compound were deter-
mined by dose-effect analysis with microcomputers
(J oseph Chou and Ting-Chao Chou)27 for comparison
with adriamycin and cis-platin, which are summarized
in Table 2. Each value is the mean of triplicate
experiments. Dose-effect relationships were analyzed
by the median-effect equation derived by Chou: fa/fu )
(D/Dm)m, where D is dose of samples, Dm is the median-
effect dose that is required for 50% cytotoxicity (e.g.,
IC50), fa is the fraction affected, fu is the fraction
unaffected, and m is the Hill-type coefficient signifying
the degree of the sigmoidal shape of the dose-effect
curve. The correlation coefficients (R) were greater than
0.9 for all median-effect lines.
Con clu sion
In the present report we summarize results of our
survey of the cytotoxicity of synthetic compounds. This
survey has revealed that imidazoquinoxalinediones
bearing 4-bromophenyl substituents at the 1 position
and pyrazinoquinoxalinediones bearing diethyl substit-
uents at the 2 and 3 positions exhibited high cytotoxicity
on human gastric adenocarcinoma cells (MKN 45). The
IC50 values of 2-methyl-4,9-dihydro-1-(4-bromophenyl)-
1H-imidazo[4,5-g]quinoxaline-4,9-dione (14) and 2,3-
diethyl-5,10-pyrazino[2,3-g]quinoxalinedione (20) were
1.30 and 7.61 µM, respectively, whereas the IC50 values
of adriamycin and cis-platin (used as anticancer drugs)
were 3.13 and 86.5 µM, respectively. It seems that
enhanced molecular area might increase the cytotoxic-
ity. Meanwhile 14 showed a considerable loss of activity
on PC 14 and colon 205. This suggests that 14 could
be a very potent and selective compound against MKN
45. All synthetic compounds are coplanar annelated
heterocyclic compounds with four nitrogens.
Molecular modeling study showed that 2-methyl-4,9-
dihydro-1-methyl-1H-imidazo[4,5-g]quinoxaline-4,9-di-
one (10) inserted between GC/GC base pairs completely.
Although modeling and experimental data indicated the
possibility that 10 intercalates with DNA, it seems to
not be significantly cytotoxic. The interaction energy
of 10 is -11.925 kcal/mol, while the compounds 14 and
20 which are markedly cytotoxic against MKN 45 have
the lowest energies of -14.648 and -14.464 kcal/mol.
There is no doubt that 10 with the higher interaction
energy shows lower activity than 14 and 20. The
compound 11 with the highest energy of -8.025 kcal/
Most of the synthetic compounds (9-14, 17-21)
showed more potent cytotoxicity on MKN 45 than cis-
platin. Within the series, the most cytotoxic compound
against human gastric adenocarcinoma cells (MKN 45)
was 2-methyl-4,9-dihydro-1-(4-bromophenyl)-1H-imidazo-
[4,5-g]quinoxaline-4,9-dione (14) with an IC50 value of
1.30 µM, whereas those of adriamycin and cis-platin are
3.13 and 86.5 µM, respectively. The cytotoxicity of 2,3-
diethyl-5,10-pyrazino[2,3-g]quinoxalinedione (20) with
an IC50 value of 7.61 µM was comparable to that of
adriamycin, and 20 is almost 10 times more active
against human gastric adenocarcinoma cells (MKN 45)
than cis-platin in vitro. The interaction energies for

4720 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 24
Yoo et al.
residue was crystallized from ethanol: yield, 370 mg (80.4%);
mp >280 °C; 1H NMR δ 2.29 (s, 3H, CH3), 8.80 (s, 1H, H-6,7);
13C NMR δ 174.7, 162.5, 146.9, 146.6, 146.2, 146.1, 145.9, 17.5;
IR 3480, 1683.9 cm-1; MS m/z for C10H6N4O2 214. Anal.
(C10H6N4O2‚H2O) C, H, N.
mol has an IC50 value of 93.14 µM. It could be evident
that there is a relationship between the interaction
energy of an intercalation complex and cytotoxicity on
MKN 45.
2-Meth yl-4,9-d ih yd r o-1-(4-br om op h en yl)-1H-im id a zo-
[4,5-g]qu in oxa lin e-4,9-d ion e (14): yield, 250 mg (86.2%);
mp 279-281 °C; 1H NMR δ 2.36 (s, 3H, CH3), 7.57 (d, 2H, J )
7.7 Hz, H-3′,5′), 7.85 (d, 2H, J ) 7.7 Hz, H-2′,6′), 9.04 (d, 2H,
H-6,7); 13C NMR δ 176.06, 171.7, 153.8, 147.7, 147.5, 134.2,
133.9, 132.8, 132.5, 132.1, 129.2, 129.1, 122.9, 13.6; IR 1680
cm-1; MS m/z for C16H9N4O279Br 368, m/z for C16H9N4O281Br
370.
Gen er a l P r oced u r e for th e P r ep a r a tion of 2-Meth yl-
1-su bstitu ted -im id a zo[4,5-g]qu in oxa lin e-4,9-d ion es 10-
13 a n d 6-Aceta m d o-7-[(4-br om op h en yl)a m in o]-5,8-qu i-
n oxa lin ed ion e (8). 2-Meth yl-4,9-d ih yd r o-1-m eth yl-1H-
im id a zo[4,5-g]qu in oxa lin e-4,9-d ion e (10). Methylamine
(40% in water, 0.19 mL) was added to a suspension of
acetaminoquinone 6 (300 mg, 1.3 mmol) in ethanol (20 mL)
at 70 °C, and the reaction mixture was refluxed for 30 min.
The solvent was evaporated under reduced pressure, and the
residue was loaded on a silica gel column (Kieselgel 9385, 230-
400 mesh) and eluted with n-hexane/ethyl acetate/ethanol (50:
45:8): yield, 210 mg (77.8%); mp 278-280 °C; 1H NMR δ 2.52
(s, 3H, COCH3), 3.97 (s, 3H, NCH3), 9.03 (s, 2H, H-6,7); IR
1630 cm-1. Anal. (C11H8N4O2) C, H, N.
2-Meth yl-4,9-d ih yd r o-1-eth yl-1H-im id a zo[4,5-g]qu in ox-
a lin e-4,9-d ion e (11): yield, 220 mg (75.9%); mp 278-280 °C;
1H NMR δ 1.36 (t, 3H, J ) 7.14 Hz, CH3), 2.56 (s, 3H, CH3),
4.43 (q, 2H, J ) 7.14 Hz, CH2), 9.02 (s, 2H, H-6,7); IR 1680
cm-1. Anal. (C12H10N4O2) C, H, N.
2-Meth yl-4,9-dih ydr o-1-isopr opyl-1H-im idazo[4,5-g]qu i-
n oxa lin e-4,9-d ion e (12): yield, 230 mg (74.2%); mp >280 °C;
1H NMR δ 1.58 (d, 6H, dime.), 2.63 (s, 3H, CH3), 5.08 (m, 1H,
CH), 9.03 (d, 2H, H-6,7); IR 1670 cm-1; MS m/z for C13H12N4O2
256. Anal. (C13H12N4O2‚3/2H2O) C, H, N.
2-Met h yl-4,9-d ih yd r o-1-p h en yl-1H -im id a zo[4,5-g]q u i-
n oxa lin e-4,9-d ion e (13): yield, 230 mg (65.7%); mp >300 °C;
1H NMR δ 2.33 (s, 3H, CH3), 7.61-7.55 (m, 5H, ArH), 9.02 (d,
2H, H-6,7); IR 1643.5 cm-1; MS m/z for C16H10N4O2 290. Anal.
(C16H10N4O2‚H2O) C, H; N: calcd, 19.31; found, 20.01.
6-Acet a m d o-7-[(4-b r om op h en yl)a m in o]-5,8-q u in oxa -
lin ed ion e (8): yield, 420 mg (91.3%); mp 270-272 °C; 1H
NMR δ 1.42 (s, 3H, CH3), 6.91 (d, 2H, J ) 8.7 Hz, H-3′,5′),
7.40 (d, 2H, J ) 8.7 Hz, H-2′,6′), 9.00 (d, 1H, J ) 9.89 Hz,
H-2), 9.01 (d, 1H, J ) 9.89 Hz, H-3), 9.32 (s, 1H, NH, exchanges
with D2O), 9.41 (s, 1H, NH (4-bromoaniline), exchanges with
D2O); MS m/z for C16H11N4O379Br 386, m/z for C16H11N4O381Br
388.
Exp er im en ta l Section
Ma ter ia ls a n d Meth od s. 1H and 13C NMR spectra were
recorded on a 300-MHz Varian Gemini NMR spectrometer.
Samples were dissolved in DMSO-d6 and CDCl3. Mass spectra
were obtained on Hewlett Packard 5790 (70 eV) GC-mass and
GC-mass 5985B spectrometers. Elemental analyses were
performed using a Perkin-Elmer model 240C elemental ana-
lyzer. IR spectra were recorded on a Perkin Elmer model 1710
spectrometer. Most reagents were purchased from Aldrich
Chemical Co. 6,7-Dichloro-5,8-quinoxalinedione was prepared
by Park’s method.13
6-Azido-7-ch lor o-5,8-qu in oxalin edion e (4). Sodium azide
(210 mg, 3.3 mmol) was added to a solution of 6,7-dichloro-
5,8-quinoxalinedione (3) (500 mg, 2.2 mmol) in acetic acid (20
mL), and the reaction mixture was stirred for 30 min. The
precipitate was collected by filtration and washed with wa-
ter: yield, 470 mg (90.4%); mp >280 °C; IR 2129, 1675, 1695
cm-1
.
6-Am in o-7-ch lor o-5,8-qu in oxalin edion e (5). Sodium boro-
hydride (120 mg, 3.2 mmol) was added to a solution of
azidoquinone 4 (500 mg, 2.1 mmol) in ethanol (20 mL), and
the reaction mixture was stirred for 30 min. The precipitate
was collected by filtration and washed with ethanol: yield, 445
mg (quantitative); mp >280 °C; 1H NMR δ 7.90 (br s, 2H, NH2,
exchanges with D2O), 8.93 (d, 2H, H-2,3); 13C NMR δ 144.7,
147.1, 147.3, 147.4, 148.0, 148.3, 177.4, 177.2; IR 3444, 1710,
1598 cm-1; MS m/z for C8H4N3O235Cl 209; m/z for C8H4N3O2
-
37
Cl 211.
6-Aceta m id o-7-ch lor o-5,8-qu in oxa lin ed ion e (6). A so-
lution of aminoquinone 5 (500 mg, 2.4 mmol) and 5 drops of
concentrated sulfuric acid in acetic anhydride (30 mL) was
stirred in an ice bath for 1 h. After quenching with methanol
slowly, the solvent was evaporated under reduced pressure.
The residue was loaded on a silica gel column (Kieselgel 9385,
230-400 mesh) and eluted with n-hexane/ethyl acetate/ethanol
1
(50:45:8): yield, 370 mg (61.7%); mp 230-231 °C; H NMR δ
2.16 (s, 3H, CH3), 9.08 (s, 2H, H-2,3), 10.28 (s, 1H, NH,
exchanges with D2O); IR 3220, 1700, 1665 cm-1; MS m/z for
C
10H6N3O335Cl 251, m/z for C10H6N3O337Cl 253. Anal. (C10H6-
N3O3Cl) C, H, N.
6-Acet a m id o-7-a m in o-5,8-q u in oxa lin ed ion e (7). So-
dium azide (390 mg, 6.0 mmol) was added to a solution of
acetaminoquinone 6 (500 mg, 1.99 mmol) in acetic acid (20
mL). The reaction mixture was stirred for 3 h, and the solvent
was evaporated under reduced pressure. The residue was
dissolved in ethanol (20 mL), and sodium borohydride (110 mg,
2.99 mmol) was added to this solution. The reaction mixture
was stirred for 1 h. The precipitate was collected by filtration
and washed with ethanol: yield, 250 mg (54.3%); mp 236-
237 °C; 1H NMR δ 2.04 (s, 3H, CH3), 7.08 (br s, 2H, NH2,
exchanges with D2O), 8.95 (d, 2H, H-2,3), 9.20 (br s, 1H, NH,
exchanges with D2O); IR 3286, 1695.5, 1668.5 cm-1; MS m/z
for C10H8N4O3 232.
Gen er a l P r oced u r e for th e P r ep a r a tion of 2-Meth -
ylim idazo[4,5-g]qu in oxalin e-4,9-dion es 9 an d 14. 2-Meth -
yl-4,9-dih ydr o-1H-im idazo[4,5-g]qu in oxalin e-4,9-dion e (9).
(a ) NaOH (2 N) (1 mL) was added to a suspension of
(acylamino)quinone 7 (500 mg, 2.2 mmol) in ethanol (20 mL)
at 70 °C, and the suspension was refluxed for 30 min. The
reaction mixture was poured into water (5 mL) with 2 N HCl
(1 mL). After cooling, the precipitate was collected by filtra-
tion, loaded on a silica gel column (Kieselgel 9385, 230-400
mesh), and eluted with chloroform/acetone (9:1).
6,7-Dia zid o-5,8-qu in oxa lin ed ion e (15). Sodium azide
(430 mg, 10 mmol) was added to a solution of 6,7-dichloro-
5,8-quinoxalinedione (3) (500 mg, 2.19 mmol) in DMF (20 mL),
and the solution was stirred at room temperature for 30 min.
The precipitate was collected by filtration and washed with
water: yield, 500 mg (94.3%); mp >280 °C; IR 2080, 1660 cm-1
.
6,7-Diam in o-5,8-qu in oxalin edion e (16). (a) Sodium boro-
hydride (200 mg, 6.2 mmol) was added to a solution of
diazidoquinone 15 (500 mg, 2.07 mmol) in ethanol (20 mL),
and the solution was stirred for 30 min at 70-80 °C. The
precipitate was collected by filtration and washed with aque-
ous ethanol.
(b) Potassium hydroxide (100 mg, 3.2 mmol) was added to
a suspension of (acylamino)quinone 7 (500 mg, 2.2 mmol) in
ethanol (20 mL), and the reaction mixture was refluxed for 1
h. After cooling, the precipitate was collected by filtration and
recrystallized from ethanol: yield, 290 mg (74.4%); mp >280
1
°C; H NMR δ 5.84 (s, 4H, each NH2, exchanges with D2O),
8.70 (s, 2H, H-2,3); 13C NMR δ 144.94, 146.08, 147.51, 195.27;
IR 3460, 3380, 1630 cm-1; MS m/z for C8H6N4O2 190.
(b) The HCl gas was passed through a suspension of
(acylamino)quinone 7 (500 mg, 2.2 mmol) in methanol (20 mL)
for 1 h, and the suspension was refluxed for 1.5 h. The
reaction mixture was neutralized with sodium bicarbonate,
and the solvent was evaporated under reduced pressure. The
5,10-P yr azin o[2,3-g]qu in oxalin edion e (17). Sodium boro-
hydride (200 mg, 6 mmol) was added to a solution of diazido-
quinone 15 (500 mg, 2 mmol) in ethanol (20 mL), and the
reaction mixture was stirred at 40 °C for 1 h. The solvent

Synthesis and Cytotoxicity of Quinoxalinediones
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 24 4721
was evaporated under reduced pressure, and the residue was
suspended in water (50 mL). Glyoxal (40% in water, 2 mL)
was added to this suspension at 90 °C, and the reaction
mixture was stirred for 1 h. After cooling, the precipitate was
collected by filtration and recrystallized from ethanol: yield,
210 mg (47.7%); mp >280 °C; 1H NMR δ 9.18 (s, H-2, 3, 6,
and 7); IR 1732.2 cm-1; MS m/z for C10H4N4O2 212.
Ack n ow led gm en t. The authors wish to thank Il
Dong Pharm. Co. for its donation of the Chair Fund to
KIST, Dr. Yun-Sil Lee (Laboratory of Radiation Effect,
Korea Cancer Hospital) for assistance in obtaining and
interpreting the cytotoxicity data, and Dr. Seog K. Kim
(Department of Chemistry, Yeungnam University) for
assistance in obtaining and interpreting the UV spec-
trum. This work was partially supported by the Korea
Science Engineering Foundation (KOSEF).
Gen er a l P r oced u r e A for th e P r ep a r a tion of 2,3-
Disu bstitu ted-5,10-pyr azin o[2,3-g]qu in oxalin edion es 18-
20. 2,3-Dim eth yl-5,10-p yr a zin o[2,3-g]qu in oxa lin ed ion e
(18). Sodium borohydride (200 mg, 6 mmol) was added to a
solution of diazidoquinone 15 (500 mg, 2 mmol) in ethanol (20
mL), and the reaction mixture was stirred at 30-40 °C for 1
h. After cooling, the precipitate was filtered and washed with
50% aqueous ethanol (50 mL). The ethanol was evaporated
under reduced pressure, and the residue was diluted with
acetic acid to 50 mL. 2,3-Butanedione (0.5 mL, 6 mmol) was
added to this solution, and the reaction mixture was stirred
at room temperature for 30 min. Water was added to this
mixture, the mixture was extracted with chloroform, and the
organic layer was dried over anhydrous magnesium sulfate.
The solvent was evaporated under reduced pressure, and the
residue was loaded on a silica gel column (Kieselgel 9385, 230-
400 mesh) and eluted with n-hexane/ethyl acetate/ethanol (50:
45:8). After evaporation of solvent from the pooled fractions
the precipitate was recrystallized from n-hexane and ethyl
acetate.
Refer en ces
(1) Rao, K. V.; Cullen, W. P. Streptonigrin, An antitumor substance,
I. Isolation and characterization. Antibiot. Ann. 1959, 950-953.
(2) Ebert, P. S.; Chirigos, M. A.; Ellsworth, P. A. Differential
response of Friend leukemia virus and lactate dehydrogenase
virus to chemotherapy and in vitro neutralization. Cancer Res
1968, 28, 363-367.
(3) Kaung, D. T.; Whittington, R. M.; Spencer, H. H.; Patno, M. E.
Comparison of chlorambucil and streptonigrin (NSC-45383) in
the treatment of malignant lymphomas. Cancer 1969, 23, 597-
600.
(4) Chirigos, M. A.; Pearson, J . W.; Papas, T. S.; Woods, W. A.; Wood,
H. B.; Spahn, G. Effect of streptonigrin (NSC-45383) and analogs
on oncornavirus replication and DNA polymerase activity.
Cancer Chemother. Rep., Part 1 1973, 57, 305-309.
(5) Inouye, Y.; Take, Y.; Oogose, K.; Kubo, A.; Nakamura, S. The
quinolinequinone as the minimum entity for reverse tran-
scriptase-inhibitory activity of streptonigrin. J . Antibiot. 1987,
40, 105-107.
(6) Shaikh, I. A.; J ohnson, F.; Grollman, A. P. Streptonigrin. 1.
Structure-activity relationships among simple bicyclic analogues.
Rate dependence of DNA degradation on quinone reduction
potential. J . Med. Chem. 1986, 29, 1329-1340.
(7) Rao, K. V.; Rock, C. P. Streptonigrin and related compounds. 6.
Synthesis and activity of some quinoxaline analogues. J . Het-
erocycl. Chem. 1996, 33, 447-458.
(8) Rao, K. V. Quinone natural products: streptonigrin (NSC-45383)
and lapachol (NSC-11905) structure-activity relationships. Can-
cer Chemother. Rep., Part 2 1974, 4, 11-17.
(9) Lin, A. J .; Lillis, B. J .; Sartorelli, A. C. Potential bioreductive
alkylating agents. 5. Antineoplastic activity of quinoline-5,8-
diones, naphthazarins, and naphthoquinones. J . Med. Chem.
1975, 18, 917-921.
(10) Lown, J . W.; J oshus, A. V.; Lee, J . S. Molecular mechanisms of
binding and single strand scisson of DNA by the antitumor
antibiotics saframycins A and C. Biochemistry 1982, 21, 419-
428.
(11) Yamashita, Y.; Tsubata, Y.; Suzuki, T.; Miyashi, T.; Mukai, T.;
Tanaka, Benazo[g][1,2,5]thiadiazolo[3,4-b]quinoxaline-5,10-dione
and its selenium analog. An unusual type of quinones. Chem.
Lett. 1990, 1990, 445-448.
(12) Renault, S. G.; Renault, J .; Baron, M.; Servolles, P.; Paoletti,
C.; Cros, S. Heterocyclic quinones. VI: synthesis and antitumoral
effects of 5,10-benzo[g]quinoxalinediones and aza-analogues.
Eur. J . Med. Chem.-Chim. Ther. 1985, 20 (2), 144-148.
(13) Han, G. H.; Shin, K. J .; Kim, D. C.; Yoo, K. H.; Kim, D. J .; Park,
S. W. A new synthetic route to 6,7-dichloro-5,8-quinoxalinedione
and synthesis of its derivatives. Heterocycles 1996, 43, 2495-
2502.
(14) Saenger, W. Principles of Nucleic Acid Structure; Springer: New
York, 1983.
(15) Hoover, J . R. E.; Day, A. R. Preparation of some imidazole
derivatives of 1,4-naphthoquinone. J . Am. Chem. Soc. 1954, 76,
4148-4152.
(16) Fieser, L. F.; Martin, E. L. A comparison of heterocyclic systems
with benzene. VII. Isologs of anthraquinone containing one and
two triazole rings. J . Am. Chem. Soc. 1935, 57, 1844-1849.
(17) Winkelmann, E. 2,3,5,6-Tetraamino-1,4-benzoquinone. Prepara-
tion, properties, and reactions. Tetrahedron 1969, 25, 2427-
2454.
(18) Nakazumi, H.; Agawa, T.; Kitao, T. Synthesis and absorption
spectra of 1,4-diazaanthraquinone derivatives. Bull. Chem. Soc.
J pn. 1979, 52 (8), 2445-2446.
(19) Suschitzky, H.; Kramen, W.; Neidlein, R.; Uhl, H. 2H-Benz-
imidazoles (isobenzimidazoles). Part 2. Synthesis and reacions
of 5-azido-2H-benzimidazoles-2-spirocyclohexane. J . Chem. Soc.
Perkin Trans 1 1988, 983-989.
(20) Efimova, G. A.; Efros, L. S. Heterocyclic derivatives based on
substituted 1,4-naphthoquinones. II. Synthesis and properties
of 2,3-disubstituted-benzoquinoxalin-5,10-diones. Zh. Organ.
Khim. (Russ.) 1966, 2, 531-536.
Gen er a l P r oced u r e B. Sodium borohydride (200 mg, 6
mmol) was added to a solution of diazidoquinone 15 (500 mg,
2 mmol) in ethanol (20 mL), and the reaction mixture was
stirred at 40 °C for 1 h. After cooling, the precipitate was
filtered and washed with 50% aqueous ethanol (50 mL). The
ethanol was evaporated under reduced pressure, and the
residue was diluted with acetic acid to 50 mL. 1,4-Dibromo-
2,3-butanedione (1.5 g, 6 mmol) was added to this solution,
and the reaction mixture was refluxed for 2.5 h. Water was
added to this mixture, the mixture was extracted with
chloroform, and the organic layer was dried over anhydrous
magnesium sulfate. The solvent was evaporated under re-
duced pressure, and the residue was crystallized from metha-
1
nol: yield, 450 mg (90.9%); mp 245-246 °C; H NMR δ 9.18
(s, 2H, H-7,8), 2.89 (s, 6H, dime.); 13C NMR δ 179.5, 160.3,
149.7, 144.7, 142.3, 23.2; IR 1701.3 cm-1; MS m/z for C12H8N4O2
240. Anal. (C12H8N4O2‚1/2H2O) C, H, N.
2-Met h yl-3-et h yl-5,10-p yr a zin o[2,3-g]q u in oxa lin ed i-
on e (19). 19 was prepared using general procedure A: yield,
420 mg (80.8%); mp 232-233 °C; 1H NMR δ 9.17 (s, 2H, H-7,8),
3.16 (q, 2H, J ) 7.4 Hz, CH2 of C-3), 2.9 (s, 3H, CH3), 1.45 (t,
3H, J ) 7.4 Hz, CH3 of C-3); 13C NMR δ 179.58, 179.46, 164.44,
159.79, 149.65, 144.71, 144.62, 142.43, 142.04, 32.9, 29.2,
11.92; IR 1703.3 cm-1; MS m/z for C13H10N4O2 254. Anal.
(C13H10N4O2‚1/2H2O) C, H, N.
2,3-Dieth yl-5,10-p yr a zin o[2,3-g]qu in oxa lin ed ion e (20).
20 was prepared using general procedure A: yield, 460 mg
1
(83.6%); mp 256-257 °C; H NMR δ 9.17 (s, 2H, H-7,8), 3.19
(q, 4H, J ) 7.4 Hz, each CH2), 1.47 (t, 6H, J ) 7.4 Hz, each
CH3); 13C NMR δ 179.5, 163.9, 149.8, 144.6, 142.2, 29.7, 28.6,
12.6; IR 1703.3 cm-1. Anal. (C14H12N4O2‚3/5H2O) C, H, N.
2,3-Diph en yl-5,10-pyr azin o[2,3-g]qu in oxalin edion e (21).
Sodium borohydride (200 mg, 6 mmol) was added to a solution
of diazidoquinone 15 (500 mg, 2 mmol) in ethanol (20 mL),
and the reaction mixture was stirred at 30-40 °C for 1 h.
Benzil (700 mg, 6 mmol) and 5 drops of concentrated sulfuric
acid were added to this solution, and the reaction mixture was
refluxed for 1 h. The solvent was evaporated under reduced
pressure, and the residue was loaded on a silica gel column
(Kieselgel 9385, 230-400 mesh) and eluted with n-hexane/
ethyl acetate/ethanol (50:45:8). After evaporation of solvent
from the pooled fractions, the precipitate was recrystallized
from DMSO: yield, 490 mg (65.3%); mp 263-265 °C; 1H NMR
δ 9.20 (d, 2H, H-7,8), 8.38 (m, 4H, each H-2′,6′), 7.7 (m, 6H,
each H-3′,4′,5′); 13C NMR δ 175.31, 155.21, 151.22, 147.49,
137.53, 136.62, 132.26, 131.33, 125.65, 117.54, 114.5; IR 1701.3
cm-1
.

4722 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 24
Yoo et al.
(21) Kitahara, Y.; Nakahara, S.; Tanaka, Y.; Kubo, A. Synthesis of
5,8-quinoxalinediones and 5,8-quinazolinediones. Heterocycles
1992, 34 (8), 1623-1630.
(25) Keller, W. Determination of the number of superhelical turns
in simian virus 40 DNA by gel electrophoresis. Proc. Natl. Acad.
Sci. U.S.A. 1975, 72 (12), 4876-4880.
(26) Mosmann, T. Rapid colorimetric assay for cellular growth and
survival: application to proliferation and cytotoxicity assays. J .
Immunol. Methods 1983, 65, 55-63.
(27) Chou, J .; Chou, T. C. Dose-effect analysis with microcomputers:
quantiation of ED50, LD50, synergism, antagonism, low-dose risk,
receptor-ligand binding and enzyme kinetics; manual and soft-
ware; Biosoft: Cambridge, England, 1987.
(22) Berman, H. M.; Young, P. R. The interaction of intercalating
drugs with nucleic acids. Annu. Rev. Biophys. Bioeng. 1981, 10,
87-114.
(23) Pindur, U.; Haber, M.; Sattler, K. Antitumor active drugs as
intercalators of deoxyribonucleic acid. J . Chem. Educa. 1993,
70 (4), 263-272.
(24) J ain, S. C.; Bhandary, K. K.; Sobell, H. M. Visualization of drug-
nuclei acid interactions at atomic resolution. J . Mol. Biol. 1979,
135, 813-840.
J M970695N