Cross-linking and sequence-specific alkylation of DNA by aziridinylquinones. 3. Effects of alkyl substituents

Article abstract of DOI:10.1021/jm991007y

The cytotoxicities and DNA cross-linking abilities of several alkyl- substituted diaziridinylquinones have been investigated. The cytotoxicities were determined in DT-diaphorase-rich (H460 and HT29) and -deficient (H596 and BE) cell lines. It was shown that the cytotoxicities in these cell lines correlated with the relative rates of reduction by the purified human enzyme and with the cross-linking efficiencies. The rates of reduction by DT- diaphorase were more dependent on the structures of the compounds than the reduction potentials, as determined by cyclic voltammetry. A computer model was also used to explain high efficiency of cross-linking and the GNC sequence selectivity of the reduced methyl-substituted diaziridinylquinones.

Full text of DOI:10.1021/jm991007y

J . Med. Chem. 1999, 42, 2245-2250
2245
Cr oss-Lin k in g a n d Sequ en ce-Sp ecific Alk yla tion of DNA by Azir id in ylqu in on es.
3. Effects of Alk yl Su bstitu en ts
Robert H. J . Hargreaves,† C. Caroline O’Hare,‡ J ohn A. Hartley,‡ David Ross,§ and J ohn Butler*,†,∇
CRC Section of Drug Development and Imaging, Paterson Institute for Cancer Research, Christie Hospital NHS Trust,
Manchester M20 9BX, U.K., CRC Drug-DNA Interactions Research Group, Department of Oncology, University College
London Medical School, 91 Riding House Street, London W1P 8BT, U.K., Department of Pharmaceutical Sciences, School of
Pharmacy, University of Colorado Health Sciences Center, Denver, Colorado 80262, and Department of Biological Sciences,
Salford University, Salford M5 4WT, U.K.
Received J anuary 25, 1999
The cytotoxicities and DNA cross-linking abilities of several alkyl-substituted diaziridinylquino-
nes have been investigated. The cytotoxicities were determined in DT-diaphorase-rich (H460
and HT29) and -deficient (H596 and BE) cell lines. It was shown that the cytotoxicities in
these cell lines correlated with the relative rates of reduction by the purified human enzyme
and with the cross-linking efficiencies. The rates of reduction by DT-diaphorase were more
dependent on the structures of the compounds than the reduction potentials, as determined
by cyclic voltammetry. A computer model was also used to explain high efficiency of cross-
linking and the GNC sequence selectivity of the reduced methyl-substituted diaziridinylquino-
nes.
Ta ble 1. Quinones Used in This Study and Reduction
Potentials
In tr od u ction
Several aziridinylquinones have undergone clinical
trials as potential antitumor drugs.1-3 These types of
compounds can be activated toward alkylation as a
result of bioreduction by the one-electron reducing
enzymes (e.g. NADPH:cytochrome P450 reductase, cy-
tochrome b5 reductase, xanthine oxidase) or by the two-
compd
R1
R2
E2(Q/Q2-) (mV)a
electron reducing enzyme (NAD(P)H:oxidoreductase
(DT-diaphorase, E.C. 1.6.99.2)).4-6
1
2
3
4
5
6
7
8
CH3
CH3
CH3
CH3
C2H5
n-C3H7
i-C3H7
H
CH3
-1330
-1355
-1344
-1396
-1350
-1407
-1396
-1065
Our previous studies on diaziridinylquinones concen-
trated on determining the structural requirements for
the compounds to be substrates for DT-diaphorase
(DTD) and the mechanisms whereby they alkylate
DNA.7-9 During the course of these studies, MeDZQ
(2,5-diaziridinyl-3,6-dimethyl-1,4-benzoquinone (1)) was
identified as being an excellent substrate for DTD and
a good correlation was obtained for the toxicity of
MeDZQ and the DTD levels in several non-small-cell
lung cancer (NSCLC) and breast cell lines.10 MeDZQ
also inhibits the growth of DTD-rich human cell animal
xenografts (unpublished results). As a consequence of
these and other studies, an analogue of MeDZQ is
currently being considered for phase I/II clinical trials
by the NCI and the CRC.
C2H5
n-C3H7
i-C3H7
C2H5
n-C3H7
i-C3H7
H
a
Values ( 30 mV.
The aim of the present work was to study the
biological and physical properties of a series of alkyl-
substituted MeDZQ analogues in order to establish the
structural requirements which are important for deter-
mining the cytotoxic properties of MeDZQ.
Resu lts
The previous works showed that diaziridinylbenzo-
quinones with a hydrogen in the 6-position cross-link
DNA at TGC sequences after reduction.9,11,12 This is
essentially because the hydrogen is small enough to be
accommodated between the stacked bases of DNA. In
contrast, MeDZQ cross-links the bases in DNA at 5′-
GNC-3′ sequences after reduction.13 However, the in-
teractions responsible for the selectivity of reduced
MeDZQ have not, as yet, been studied.
Ch em istr y. The quinones used in this study are
shown in Table 1.
In Vitr o Cytotoxicities. The cytotoxicities for the
quinones 1-9 are shown in Table 2.
Cyclic Volta m m etr y. The two-electron reduction
potentials, as measured by cyclic voltammetry, are
included in Table 1. The trends in reduction potentials
are as would be expected from the Hammett functions
of the substituted groups.14 Thus, as the σp values for
the alkyl groups are relatively similar, the reduction
potentials are also very similar. This is in contrast to
the more positive σp value for a hydrogen which means
that the reduction potentials for DZQ are more positive
(DZQ is more readily reduced).
* Corresponding author. E-mail: jbutler@picr.man.ac.uk. Fax: (0161)
446 3109. Tel: (0161) 446 3150.
† Christie Hospital NHS Trust.
‡ University College London Medical School.
§ University of Colorado Health Sciences Center.
∇ Salford University.
10.1021/jm991007y CCC: $18.00 © 1999 American Chemical Society
Published on Web 05/22/1999

2246 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 12
Hargreaves et al.
F igu r e 1. DNA interstrand cross-linking of plasmid DNA by the reduced quinones determined by the agarose gel assay. Compounds
were reacted at the concentration shown over 2 h at 37 °C, pH 4, following reduction with dithiothreitol. C and Cd are nondenatured
and denatured controls, respectively, DS is double-stranded DNA, and SS is single-stranded DNA.
Ta ble 2. In Vitro Cytotoxicity Values (nM) for the Different
Ta ble 4. Relative DNA Interstrand Cross-Linking Efficiencies
of the Reduced Quinones
Cell Lines and Cytotoxic Differentialsa
compd
H460
H596
HT29
BE
% double-stranded (cross-linked) DNA
1
2
3
4
5
6
7
8
0.77
1.6
3.3
3.6
4.1
20.1
170
4.6
48.8 (63.4)
98.5 (61.6)
228 (69.1)
1230 (341.7)
1160 (282.9)
1290 (64.2)
2620 (15.4)
12.9 (2.8)
3.4
3.9
5.5
25.9 (7.6)
84 (21.5)
67 (12.2)
1300 (39.4)
290 (13.3)
590 (10.1)
3130 (1.1)
10.2 (2)
compd
10 µM
100 µM
1
2
3
4
5
6
7
8
79.8
69.0
58.4
28.1
60.9
16.4
0
100
100
100
87.8
100
95.6
0
33
21.8
58.2
2740
5.1
a
100
100
Errors within (15%. The cytotoxicity differentials (H596/H460
and HT29/BE) are in parentheses.
cross-linking agent (Table 4), it shows the poorest
differential between the DTD-rich and -deficient cell
lines. This is probably because DZQ has a more positive
one-electron reduction potential and thus can be more
readily reduced by other enzymes.15,16 Furthermore, as
DZQ has labile hydrogens, it can be detoxified by
undergoing Michael additions with other cell compo-
nents.17 We have observed similar trends with other
diaziridinylquinones.10
The cytotoxicity data shows that there is a trend such
that as the size of the alkyl group is increased, the
toxicities in all of the cell lines generally decrease
(compare 1, 5, 6, and 7). A similar trend is found for
the DNA cross-linking efficiencies (Table 4). However,
the cross-linking data was obtained after chemical
reduction by dithiothreitol and hence is not influenced
by the relative rates of reduction by DTD. Indeed, the
conditions of the reduction were chosen such that all of
the quinones were reduced to the same extent. Hence,
the implication is that it is the size or shape of the
hydroquinones which determines the cross-linking ef-
ficiencies.
Ta ble 3. Rates of Reduction by Purified DTD and BE Cell
Extracts
compd
DTDa
BE extractsb
1
2
3
4
5
6
7
8
21.0
12.5
7.9
2.7
3.0
1.0
0.22
35.6
9.2
4.0
3.4
1.1
2.3
1.7
NDc
25.3
a
b
µM/min/mg of protein. nM/min/mg of protein. c ND, not
detected.
Red u ction Ra tes by DTD. The rates of reduction
of the compounds by human recombinant DTD and in
the BE cell extracts are shown in Table 3.
DNA In ter str a n d Cr oss-Lin k in g. The abilities of
the reduced forms of the quinones to produce inter-
strand cross-links in the plasmid DNA are shown in
Figure 1, and the relative cross-link efficiencies are
given in Table 4.
The results in Table 4 show that the cross-linking
efficiencies decrease as the size of the side chain
increases (compare 8, 1, 5, 6, and 7). It is also apparent
that if a compound contains a methyl group, it can cross-
link DNA more efficiently than the disubstituted ana-
logue (compare 2 and 5, 3 and 6, 4 and 7). Quinone 7,
Discu ssion
The cytotoxicity values for MeDZQ (1) and DZQ (8),
given in Table 2, are of similar magnitude to those
reported previously.8,11 Interestingly, although DZQ is
the best substrate for DTD (Table 3) and is the best

DNA Cross-Linking by Aziridinylquinones
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 12 2247
similar trend as the cross-linking efficiencies (Table 4).
Unfortunately, the crystal structure of the human DTD
enzyme is not available, but the rat enzyme with bound
duroquinone has been published and is believed to be
similar to that of the human.19 The rat enzyme structure
shows that a methyl group from duroquinone can be
accommodated close to the Y126, Y128, and F178 amino
acids, with the -OH group of the hydroquinone being
hydrogen bonded to the -OH groups of Y126 and Y128.
It would appear that the methyl group in the para
position to this methyl is free to rotate around the C-C
bond. It can be speculated from these positions that
groups larger than a simple hydrogen or methyl group
will have difficulty in locating the space available
around the Y126, Y128, and F178 amino acids. If this
is true, it would explain why the ethyl and propyl groups
of compounds 5 and 6 are much poorer substrates than
1. Similarly, the methyl groups of compounds 2-4 can
be accommodated as with duroquinone with the more
bulkier alkyl groups at the para positions having a much
less effect on steric hindrance. This general concept
wherein only one side of the quinone has to interact with
the electron-transfer sites of DTD could explain why
other apparently more bulkier quinones (e.g. certain
indolequinones such as EO9 analogues and strepto-
nigrin11,18,20) can also be good substrates for the enzyme
and also would explain why adriamycin and other
anthraquinones are poor substrates for DTD.21
Although 1 is the best substrate for DTD, the best
DNA cross-linker, and the most cytotoxic of the alkyl-
substituted quinones, it does not produce the highest
cytotoxic differential (Table 2). It is generally believed
that these differentials are due to the differences in
activation of the quinones by DTD in the DTD-rich cell
lines (H460 and HT29) compared to the amount of
activation by the one-electron reducing enzymes in the
DTD-deficient cell lines (H596 and BE).
The relative rates of reduction by DTD and the
reduction in the BE extracts generally follow similar
trends (Table 3). However, it can be seen from Table 2
that 4 and 5 are inordinately less toxic in the BE and
H596 cell lines and hence produce relatively high
differentials. It would appear that these quinones or
their semiquinones are more readily detoxified in these
DTD-deficient cell lines. At the present time, we cannot
explain these findings. However, the possible mecha-
nisms of detoxification involving glutathione addition,
phenol UDP glucuronosyltransferases, sulfotransferas-
es, and MRP1 are currently being investigated for all
of these quinones.
F igu r e 2. Initial interactions between MeDZQ hydroquinone
and DNA at a 5′-GNC-3′ sequence. The dotted lines indicate
the hydrogen bonds between the hydroquinone OH’s and the
guanine O-6’s and the interactions between the protonated
aziridines and the guanine N-7’s.
which has two isopropyl side chains, did not produce
any measurable cross-links, even at the highest dose
used in the assay. Our previous studies have shown that
MeDZQ cross-links DNA at GNC sequences after reduc-
tion.13 Computer models were therefore generated in
order to study the initial interaction between the
hydroquinones of the alkyl-substituted diaziridinylben-
zoquinones and DNA. A simplified diagram for the
hydroquinone of MeDZQ is shown in Figure 2. In this
model, the two hydroquinone OH groups hydrogen bond
to the two O-6’s of the guanine residues. In this position,
the two protonated aziridines can interact with the two
N-7’s of the same guanines (the positions where the
GNC alkylations will occur). The overall interactions
occur in the major groove of DNA. However, the model
also predicts that due to the helical structure of the
DNA, the hydroquinone has to be twisted in the major
groove. This twisting forces one of the methyl groups
to point slightly into the DNA and the other group
points slightly out. It is this slight twisting which can
account for the increased cross-linking efficiencies of the
methyl-substituted analogues. The models predict that
the hydroquinones with a methyl group can fit closer
to the DNA and have less steric hindrance with the N
bases. If this side chain is increased, the GNC cross-
linking abilities of these compounds will decrease. The
models also predict that as the second side chain can
point away from the DNA, all of the methyl-substituted
quinones can be accommodated at this alkylation site.
Hence the analogues 2-4 are relatively efficient cross-
linkers.
The rates of reduction of the different quinones by
the human recombinant DTD (Table 3) show that DZQ
is the best substrate for the enzyme and there is a
decrease in the rates as the alkyl groups are extended.
DZQ has the most positive two-electron reduction
potential (Table 1) and thermodynamically should be
the most readily reduced. The two-electron potentials
of the alkyl-substituted quinones are very similar.
Hence, although there is an indication that rates of
reduction by DTD follow a similar trend as the two-
electron reduction potentials, it is probable that the
dramatic changes in the rates of reduction are due to
the changes in structure of the quinones. A recent study
has shown that indolequinones behave similarly.18
Interestingly, the rates of reduction by DTD follow a
Exp er im en ta l Section
Ch em ica ls. The quinones used in the study are shown in
Table 1. 1,4-Diacetyl-2,5-dimethoxybenzene was purchased
from Apin Chemicals Ltd., Abingdon, Oxon, U.K.
Aziridine (ethylenimine) was prepared according to the flash
distillation method of Reeves et al.22 Initial stock solutions for
the cross-linking assays were made up in DMSO at 10 mM.
Electrophoresis grade acrylamide and bis(acrylamide) were
purchased from Sigma; ultrapure urea and agarose were from
BRL and pBR322 plasmid DNA was obtained from Northum-
bria Biologicals. All other reagents were of the highest purity
commercially available.
2,5-Diaziridinyl-3,6-dimethyl-1,4-benzoquinone, MeDZQ (1),23
and 2,5-diaziridinyl-1,4-benzoquinone (8)24 were prepared
according to previous methods.

2248 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 12
Hargreaves et al.
1,4-Dieth yl-2,5-d im eth oxyben zen e. A stirred solution of
1,4-diacetyl-2,5-dimethoxybenzene (2.5 g, 11.3 mmol), hydra-
zine (80% solution in water, 9.4 mL, 0.23 mol) and potassium
hydroxide (4.9 g, 87.5 mmol) in ethylene glycol (18 mL) was
refluxed for 13 h. The solution was then distilled until the
reaction temperature reached 195 °C. The distillate was then
extracted with ether. The ether extracts were dried (Na2SO4),
and the solvent was removed in vacuo to yield a colorless liquid
which solidified to colorless crystals (0.9 g, 41%). A further
crop was obtained by extracting the reaction mixture in the
same way (0.8 g, 36%): 1H NMR (60 MHz, CDCl3) δ 6.67 (s,
2H, Ar-H), 3.75 (s, 6H, OCH3), 2.60 (q, J ) 7 Hz, 4H, CH2),
1.19 (t, J ) 7 Hz, 6H, CH3); MS EI m/z 194 (M+), 193, 179,
165, 135.
(2,5-Dim eth oxy-4-m eth ylp h en yl)eth en e. To a stirred
suspension of methyltriphenylphosphonium bromide (6.54 g,
18.3 mmol) in dry THF (120 mL) was added butyllithium (11.5
mL, 1.6 M solution in hexanes, 18.4 mmol). After 20 min, a
solution of 2,5-dimethoxy-4-methylbenzaldehyde25 (3 g, 16.7
mmol) in dry THF (20 mL) was added and the stirring
continued. After a further 30 min, the reaction was quenched
with water (150 mL) and extracted three times with diethyl
ether. The ether extract was washed twice with water and once
with brine and dried (Na2SO4) and the solvent removed in
vacuo. The product was then passed down a short silica column
(eluent, 9:1 petroleum ether 40/60:EtOAc) to remove tri-
phenylphosphonium oxide, and the solvent was again removed
in vacuo (2.49 g, 84%): MS EI m/z 178 (M+), 163, 135.
1-(2,5-Dim et h oxy-4-m et h ylp h en yl)p r op -1-en e. To a
stirred suspension of ethyltriphenylphosphonium bromide (6.8
g, 18.3 mmol) in dry THF (120 mL) was added butyllithium
(11.5 mL, 1.6 M solution in hexanes, 18.4 mmol). After 20 min,
a solution of 2,5-dimethoxy-4-methylbenzaldehyde (3 g, 16.7
mmol) in dry THF (20 mL) was added and the stirring
continued. After a further 30 min, the reaction was quenched
with water (150 mL) and extracted three times with diethyl
ether. The ether extract was washed twice with water and once
with brine and dried (Na2SO4) and the solvent removed in
vacuo. The product was then passed down a short silica column
(eluent, 9:1 petroleum ether 40/60:EtOAc) to remove any
triphenylphosphonium oxide, and the solvent was again
removed in vacuo (2.53 g, 79%): MS EI m/z 192 (M+), 177,
149, 115, 91.
2-Eth yl-5-m eth yl-1,4-d im eth oxyben zen e. To a stirred
solution of (2,5-dimethoxy-4-methylphenyl)ethene (1 g, 5.6
mmol) and 1,4-cyclohexadiene (10.61 mL, 8.98 g, 112 mmol)
in ethanol (30 mL) under N2 was added 10% palladium on
charcoal (1 g, 1 wt equiv). After 1 h, the resulting mixture was
filtered through Celite and the solvent removed in vacuo. The
compound was not purified further (837 mg, 83%): 1H NMR
(60 MHz, CDCl3) δ 6.61 (s, 2H, Ar-H), 3.73 (s, 6H, OCH3),
2.60 (q, J ) 7.5 Hz, 2H, CH2), 2.16 (s, 3H, ArCH3), 1.17 (t, J
) 7.5 Hz, 3H, CH2CH3); MS EI m/z 180 (M+), 179, 165, 151,
135.
2-Eth yl-5-m eth yl-1,4-ben zoqu in on e. To a stirred solution
of 1,4-dimethoxy-2-ethyl-5-methylbenzene (0.75 g, 4.17 mmol)
in acetonitrile (15 mL) was added a solution of ceric am-
monium nitrate (6.88 g, 12.5 mmol) in water (15 mL) over a
period of 5 min. After a further 30 min, the solution was
extracted with diethyl ether and dried (Na2SO4) and the
solvent removed in vacuo, yielding a yellow solid. This solid
was then dissolved in dichloromethane and passed down a
small silica column; the solvent was again removed in vacuo
to yield yellow crystals (379 mg, 61%): 1H NMR (60 MHz,
CDCl3) δ 6.53 (m, 2H, CH), 2.45 (q, J ) 7 Hz, 2H, CH2CH3),
2.01 (d, J ) 1.5 Hz, 3H, CH3), 1.11 (t, J ) 7 Hz, 3H, CH2CH3);
MS EI m/z 150 (M+), 122, 107, 79.
2-Meth yl-5-p r op yl-1,4-ben zoqu in on e. To a stirred solu-
tion of 1,4-dimethoxy-2-propyl-5-methylbenzene (0.8 g, 4.12
mmol) in acetonitrile (15 mL) was added a solution of ceric
ammonium nitrate (6.8 g, 12.36 mmol) in water (15 mL) over
a period of 5 min. After a further 30 min the solution was
extracted with diethyl ether and dried (Na2SO4) and the
solvent removed in vacuo to yield a brown oil. This oil was
then dissolved in dichloromethane and passed down a small
silica column; the solvent was again removed in vacuo to yield
a yellow oil (601 mg, 89%): 1H NMR (60 MHz, CDCl3) δ 6.53
(m, 2H, CH), 2.39 (t, J ) 7.5 Hz, 2H, CH2CH2CH3), 2.00 (d, J
) 1.5 Hz, 3H, CH3), 1.46 (m, 2H, CH2CH2CH3), 0.96 (t, J ) 7
Hz, 3H, CH2CH2CH3); MS EI m/z 164 (M+), 149, 137, 121, 91.
2,5-Dip r op yl-1,4-ben zoqu in on e. To a stirred solution of
2,5-diallylhydroquinone (1 g, 5.3 mmol) and 1,4-cyclohexadiene
(10.0 mL, 8.47 g, 106 mmol) in ethanol (40 mL) under N2 was
added 10% palladium on charcoal (2 g, 2 wt equiv). After 13
h, the resulting mixture was filtered through Celite and the
solvent removed in vacuo to yield 2,5-dipropylhydroquinone
as a white solid (1.01 g, 99%). This solid was redissolved in
diethyl ether (50 mL), and silver(I) oxide (2.39 g, 10.3 mmol)
was added. This was stirred as a slurry for 12 h after which
time the silver residue was filtered off and the solvent removed
in vacuo to yield a yellow oil (0.92 g, 93%): MS EI m/z 192
(M+), 177, 164, 163, 149, 135, 121.
2,5-Dia zir id in yl-3,6-d ieth yl-1,4-ben zoqu in on e (5). To a
solution of 2,5-diethyl-1,4-benzoquinone (420 mg, 2.56 mmol)
in ethanol (15 mL) under N2 was added aziridine (0.40 mL,
0.33 g, 7.68 mmol). This was then placed in a refrigerator for
72 h. The resulting red precipitate was then filtered and
recrystallized from THF to yield dark-red needles (203 mg,
1
32%): mp 183-184 °C (lit. mp 179 °C26); H NMR (60 MHz,
CDCl3) δ 2.55 (q, J ) 7 Hz, 4H, CH2CH3), 2.27 (s, 8H, Az),
1.09 (t, J ) 7 Hz, 6H, CH2CH3); MS EI m/z 246 (M+), 231,
203, 189, 175, 94, 80.
2,5-Dia zir id in yl-3,6-d ip r op yl-1,4-ben zoqu in on e (6). To
a solution of 2,5-dipropyl-1,4-benzoquinone (0.9 g, 4.7 mmol)
in dry ethanol (30 mL) under N2 was added aziridine (0.73
mL, 0.61 g, 14.1 mmol). This was placed in a refrigerator for
24 h. The resulting orange precipitate was filtered, washed
with cold ethanol, and dried (324 mg, 25%): mp 121-122 °C
(lit. mp 119 °C26); 1H NMR (200 MHz, CDCl3) δ 2.50 (t, J ) 8
Hz, 4H, CH2CH2CH3), 2.39 (s, 8H, Az), 1.51 (m, 4H, CH2CH2-
CH3), 0.99 (t, J ) 7.5 Hz, 6H, CH2CH2CH3); MS EI m/z 274
(M+), 259, 245, 231, 217, 194, 189, 175, 165, 108, 91, 80.
2-Meth yl-5-p r op yl-1,4-d im eth oxyben zen e. To a stirred
solution of 1-(2,5-dimethoxy-4-methylphenyl)prop-1-ene (1 g,
5.2 mmol) and 1,4-cyclohexadiene (9.8 mL, 8.3 g, 104 mmol)
in ethanol (30 mL) under N2 was added 10% palladium on
charcoal (1 g, 1 wt equiv). After 1 h, the resulting mixture was
filtered through Celite and the solvent removed in vacuo. The
compound was not purified further (914 mg, 91%): 1H NMR
(60 MHz, CDCl3) δ 6.62 (s, 2H, ArH), 3.74 (s, 6H, OCH3), 2.56
(t, J ) 8 Hz, 2H, ArCH2), 2.18 (s, 3H, ArCH3), 1.61 (m, 2H,
CH2CH2CH3), 0.93 (t, J ) 7 Hz, 3H, CH2CH3); MS EI m/z 194
(M+), 179, 165, 151, 135.
2,5-Dieth yl-1,4-ben zoqu in on e. To a stirred solution of 1,4-
diethyl-2,5-dimethoxybenzene (0.8 g, 4.12 mmol) in acetonitrile
(15 mL) was added a solution of ceric ammonium nitrate (6.8
g, 12.36 mmol) in water (10 mL) over a period of 5 min. After
30 min, water (100 mL) was added to the reaction mixture
and the resulting precipitate was filtered and dried to give a
yellow solid (593 mg, 88%): 1H NMR (60 MHz, CDCl3) δ 6.48
(s, 2H, C-H), 2.43 (q, J ) 7 Hz, 4H, CH2), 1.12 (t, J ) 7 Hz,
6H, CH3); MS EI m/z 164 (M+), 149, 120, 91, 73.
2,5-Diazir idin yl-3-eth yl-6-m eth yl-1,4-ben zoqu in on e (2).
To a solution of 2-ethyl-5-methyl-1,4-benzoquinone (370 mg,
2.47 mmol) in ethanol (15 mL) under N2 was added aziridine
(0.39 mL, 0.32 g, 7.4 mmol). This was then placed in a
refrigerator for 72 h. The resulting orange precipitate was then
filtered and passed down a silica column. Removal of the
solvent in vacuo yielded red crystals (233 mg, 41%): mp 157-
1
158 °C (lit. mp 156 °C26); H NMR (200 MHz, CDCl3) δ 2.53
(q, J ) 7.5 Hz, 2H, CH2CH3), 2.27 (s, 4H, Az), 2.26 (s, 4H, Az),
1.95 (s, 3H, CH3), 1.09 (t, J ) 7.5 Hz, 3H, CH2CH3); MS EI
m/z 232 (M+), 217, 203, 189, 161, 94, 80.
2,5-Dia zir id in yl-3-m et h yl-6-p r op yl-1,4-b en zoq u in on e
(3). To a solution of 2-methyl-5-propyl-1,4-benzoquinone (580
mg, 3.53 mmol) in ethanol (15 mL) under N2 was added
aziridine (0.55 mL, 0.45 g, 10.61 mmol). This was then placed
in a refrigerator for 72 h. The resulting orange precipitate was

DNA Cross-Linking by Aziridinylquinones
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 12 2249
then filtered and dried (270 mg, 31%): mp 107-108 °C; 1H
NMR (200 MHz, CDCl3) δ 2.48 (t, J ) 7.5 Hz, 2H, CH2CH2-
CH3), 2.28 (s, 4H, Az), 2.26 (s, 4H, Az), 2.00 (s, 3H, CH3), 1.49
(m, 2H, CH2CH2CH3), 0.98 (t, J ) 7 Hz, 3H, CH2CH2CH3); MS
EI m/z 246 (M+), 231, 217, 203, 189, 137, 80. Anal. (C14H18N2O3)
C, N; H: calcd, 7.37; found, 8.09. HREIMS: found, 246.1368;
(typical errors ( 15%). The concentration of protein was
determined using the Biorad assay.
Deter m in a tion of DNA In ter str a n d Cr oss-Lin k in g. The
method has been previously described in detail.30 Briefly, 5′-
end-labeled DNA (∼5000 cpm/sample) was incubated with
drug in 25 mM triethanolamine, 1 mM EDTA (pH 4) with 2
mM dithiothreitol at 37 °C for 2 h. Reactions were terminated
by the addition of an equal volume of 0.6 M sodium acetate,
20 mM EDTA, 100 µM/mL tRNA, and the DNA was im-
mediately precipitated by the addition of 3 vol of 95% ethanol.
Following centrifugation and removal of the supernatant, the
DNA pellet was dried by lyophilization. Samples were dis-
solved in 10 mL of 30% dimethyl sulfoxide, 1 mM EDTA, 0.04%
bromophenol blue, 0.04% xylene cylanol, heated at 90 °C for 2
min, and chilled immediately in an ice-water bath prior to
gel loading.
Control undenatured samples were dissolved in 10 mL of
6% sucrose, 0.04% bromophenol blue and loaded directly.
Electrophoresis was performed on 20-cm 0.8% submerged
horizontal agarose gels at 40 V for 16 h with Tris-acetate
running buffer.
Gels were dried at 90 °C onto filter paper; an autoradiog-
raphy was performed at -70 °C. Quantitation was achieved
by microdensitometry of the autoradiograph using a LKB
Ultrascan-XL laser densitometer. For each lane the amount
of single- and double-stranded DNA was determined, and the
percent cross-linked (double-stranded) DNA was calculated.
Molecu la r Mod elin g. The computer molecular modeling
was carried out on a Silicon Graphics Iris 4D/310GTX work-
station using QUANTA 4.0 software (including CHARMm
22.2) working under IRIS 4.0.5.
C
14H18N2O3 requires 246.1368.
2,5-Dia zir id in yl-3-isop r op yl-6-m eth yl-1,4-ben zoqu in o-
n e (2,5-Dia zir id in ylth ym oqu in on e) (4). To a solution of
thymoquinone (1 g, 6.09 mmol) in ethanol (30 mL) under N2
was added aziridine (0.94 mL, 0.78 g, 18.27 mmol). This was
then placed in a refrigerator for 72 h. The resulting red
precipitate was then filtered and recrystallized from methanol
to yield purple needles (317 mg, 21%): mp 119-120 °C (lit.
mp 118 °C26); 1H NMR (300 MHz, CDCl3) δ 3.22 (sept, J ) 6.5
Hz, 1H, CH(CH3)2), 2.27 (s, 8H, Az), 1.97 (s, 3H, CH3), 1.39 (d,
J ) 6.5 Hz, 6H, CH(CH3)2); MS EI m/z 246 (M+), 231, 203,
189, 175, 108, 94, 81.
2,5-Dia zir id in yl-3,6-d iisop r op yl-1,4-ben zoqu in on e (7).
To a solution of 2,5-diisopropyl-1,4-benzoquinone (350 mg, 1.82
mmol) in ethanol (10 mL) was added aziridine (0.28 mL, 0.23
g, 5.34 mmol). This was placed in a refrigerator for 14 days.
The resulting red/orange precipitate was filtered, washed with
cold ethanol and dried (176 mg, 35%): mp 168-169 °C (lit.
1
mp 170 °C26); H NMR (300 MHz, CDCl3) δ 3.23 (sept, J ) 7
Hz, 2H, CH(CH3)2), 2.29 (s, 8H, Az), 1.32 (d, J ) 7 Hz, 6H,
CH(CH3)2); MS EI m/z 274 (M+), 259, 245, 231, 216, 203, 201,
189, 175, 160.
Toxicity Testin g. The cell lines used in this work were
H596, H460, HT29 and BE. These were obtained from the
American Tissue Type Culture Collection (ATTCC) or from the
European Collection for Animal Cell Cultures (ECACC). HT29
is a human colon cancer cell line, and H460 is a human large-
cell lung carcinoma line. These two cell lines were used as they
express relatively high levels of DTD. The corresponding
matched cell lines, H596 (human lung carcinoma) and BE
(human colon carcinoma), were chosen because the H596 and
BE lines do not have detectable DTD activity.8,27 Continuous
challenge cytotoxicity studies were carried out on these cells
(600/well) in 96-well plates using the MTT method.28 After 5
days, the absorbances were read on a multiscan plate reader
at 540 and 640 nm. Plots were then drawn for inhibition of
cell growth as a function concentration. Each compound was
tested in triplicate at each concentration on at least two
occasions.
Ack n ow led gm en t. This work was supported by
grants from the Cancer Research Campaign, U.K., and
by NIH Grant CA51210. We are grateful to J ohn
Hadfield, Dee Evans, and Sally Haran for advice and
technical assistance.
Su p p or tin g In for m a tion Ava ila ble: Stereoview of the
initial interactions of the protonated hydroquinone of MeDZQ
(1) with DNA. This material is available free of charge via the
Internet at http://pubs.acs.org.
Refer en ces
(1) Schold, S. C.; Friedman, H. S.; Bjornsson, T. D.; Falleta, J . M.
Treatment of patients with recurrent primary brain tumours
with AZQ. Neurology 1984, 34, 615-619.
Cyclic Volta m m etr y. The voltametric studies were carried
out using an EG and G K0264 microcell; current was applied
by a PAR 362 scanning potentiostat operated by Condecon 310
software. All quinones were tested at 100 µM in dimethyl-
formamide. Tetra-n-butylammonium perchlorate was used as
the supporting electrolyte. Oxygen was removed from the
system by purging with nitrogen. The cyclic voltammograms
were recorded using potential sweep rates ranging from 10 to
(2) Bender, J . F.; Grillo-Lopez, A.; Posada, J . G. Diaziquone. Invest.
New Drugs 1983, 1, 71-84.
(3) Obe, G.; Beek, B. Trenimon: Biochemical, physiological and
genetic effects on cells and organisms. Mutat. Res. 1979, 65, 21-
70.
(4) Szmigiero, L.; Kohn, K. W. Mechanisms of DNA strand breakage
and interstrand cross-linking by diaziridinylbenzoquinone
(Diaziquone) in isolated nuclei from human cells. Cancer Res.
1984, 44, 4453-4457.
(5) Siegel, D.; Gibson, N. W.; Preusch, P. C.; Ross, D. Metabolism
of diaziquone by NAD(P)H:(quinone acceptor) oxireductase (DT-
diaphorase); role in diaziquone-induced DNA damage and cy-
totoxicity in human colon carcinoma cells. Cancer Res. 1990, 50,
7293-7300.
(6) Butler, J .; Dzielendziak, A.; Lea, J . S.; Ward, T. H.; Hoey, B. M.
Contrasting cytotoxic mechanisms of similar antitumour diaz-
iridinybenzoquinones. Free Radical Res. Commun. 1990, 8, 231-
239.
(7) Dzielendziak, A.; Butler, J .; Hoey, B. M.; Lea, J . S.; Ward, T. H.
A comparison of the structural and cytotoxic activity of novel
AZQ analogues. Cancer Res. 1990, 5, 2003-2007.
(8) Mayalarp, S. P.; Hargreaves, R. H. J .; Butler, J .; O’Hare, C. C.;
Hartley, J . A. Cross-linking and sequence specific alkylation of
DNA by diaziridinylquinones. 1. Quinone Methides. J . Med.
Chem. 1996, 39, 531-537.
(9) Hargreaves, R. H. J .; Mayalarp, S. P.; Butler, J .; McAdam, S.
R.; O’Hare, C. C.; Hartley, J . A. Cross-linking and sequence
specific alkylation of DNA by aziridinylquinones. 2. Structure
requirements for sequence selectivity. J . Med. Chem. 1997, 40,
357-362.
(10) Gibson, N. W.; Hartley, J . A.; Butler, J .; Seigel, D.; Ross, D. The
relationship between DT-diaphorase mediated metabolism of a
series of diaziridinylbenzoquinones and DNA damage and cy-
totoxicity. Mol. Pharmacol. 1992, 42, 531-536.
500 mVs-1, but it was routinely set at 100 mVs-1
.
Ra tes of Red u ction of th e Qu in on es by DTD. The
reduction of the quinones by human DTD was followed by the
corresponding reduction of cytochrome c.29 The quinones (50
µM) were put into a 1-mL quartz cuvette containing 50 mM
tris-HCl buffer (pH 7.4), 70 µM cytochrome c, and 10 µL of
DTD (specific activity 653 µmol/min/mg of protein). The
reaction was initiated by the addition of NADH (250 µM). The
reduction of the cytochrome c was followed at 550 nm (∆R550nm
) 21 × 103 M-1 cm-1). Each quoted value is the average of
three determinations (typical errors ( 12%).
R a t es of R ed u ct ion of t h e Qu in on es b y BE Cell
Extr a cts. The extract used in these experiments was made
from a suspension of 2 × 107 BE cells in 2 mL of phosphate-
buffered saline. This was sonicated and then centrifuged to
remove the cell debris. The quinones (100 µM) were placed in
a 1-mL cuvette containing phosphate-buffered saline, acety-
lated cytochrome c (100 µM), and NADPH (200 µM). The
reaction was initiated by the addition of 50 µL of cell extract.
The reduction of acetylated cytochrome c was followed at 550
nm. Each quoted value is the average of three determinations

2250 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 12
Hargreaves et al.
(11) Beall, H. D.; Murphy, A. M.; Siegel, D.; Hargreaves, R. H. J .;
Butler, J .; Ross, D. Nicotinamide adenine dinucleotide (phos-
phate): Quinone oxireductase (DT-diaphorase) as a target for
bioreductive antitumor quinones: Quinone cytotoxicity and
selectivity in human lung and breast cancer cell lines. Mol.
Pharmacol. 1995, 48, 499-504.
(12) Hartley, J . A.; Berardini, M.; Ponti, M.; Gibson, N. W.; Thomp-
son, A. S.; Thurston, D. E.; Hoey, B. M.; Butler, J . DNA cross-
linking and sequence selectivity of aziridinylbenzoquinones: A
unique reaction at 5′-GC-3′ sequences with 2,5-diaziridinyl-1,4-
benzoquinone upon reduction. Biochemistry 1991, 30, 11719-
11724.
(13) Berardini, M. D.; Souhami, R. L.; Lee, C.-S.; Gibson, N. W.;
Butler, J .; Hartley, J . A. Two structurally related diaziridinyl-
benzoquinones preferentially cross-link DNA at different sites
upon reduction with DT-Diaphorase. Biochemistry 1993, 32,
3306-3312.
(14) Hansch, C.; Leo, A.; Taft, R. W. A survey of Hammett substituent
constants and resonance and field parameters. Chem. Rev. 1991,
91, 165-195.
(15) Butler, J .; Hoey, B. M. The one electron reduction potential of
several substrates can be related to their reduction rates by
cytochrome P450 reductase. Biochim. Biophys Acta 1993, 1161,
73-77.
(16) Butler, J .; Redox Cycling antitumour drugs. In DNA and Free
Radicals: Techniques, Mechanisms and Applications; Halliwell,
B., Aruoma, O. I., Eds.; OICA International: London, 1998; pp
131-154.
(17) Butler, J .; Hoey, B. M. Reactions of glutathione and glutathione
radicals with benzoquinones. Free Radical Biol. Med. 1992, 12,
337-345.
(20) J affar, M.; Naylor, M. A.; Robertson, N.; Lockyer, S. D.; Phillips,
R. M.; Everett, S. A.; Adams, G. E.; Stratford, I. J . 5-Substituted
analogues of 3-hydroxymethyl-5-aziridinyl-1-methyl-2-[1H-in-
dole-4,7-dione]prop-2-en-1-ol (EO9, NSC 382459) and their re-
gioisomers as hypoxia-selective agents: structure-cytotoxicity in
vitro. Anticancer Drug Des. 1998, 13, 105-123.
(21) Wallin, R. Adriamycin and DT-diaphorase. Cancer Lett. 1986,
30, 97-101.
(22) Reeves, W. A.; Drake, G. L.; Hoffpauir, C. L. Ethylenimine by
flash distillation. J . Am. Chem. Soc. 1951, 73, 3522.
(23) Cameron, D. W.; Giles, R. G. F. Photochemical formation of
benzoxazoline derivatives from aminated quinones. J . Chem.
Soc. C 1968, 1461-1464.
(24) Petersen, S.; Gauss, W.; Urbschat, E. Synthesis of simple
quinone derivatives with fungicidal, bacteriostatic and cytostatic
properties. Angew. Chem. 1955, 67, 217-231.
(25) Standridge, R. T.; Howell, H. G.; Gylys, J . A.; Partyka, R. A.;
Shulgin, A. T. Phenyalkylamines with potential psychothera-
peutic utility. 1. 2-Amino-1-(2,5-dimethoxy-4-methylphenyl-
butane. J . Med. Chem. 1976, 19, 1400-1404.
(26) Nakao, H.; Arakawa, M.; Nakamura, T.; Fukushima, M. Anti-
leukemic agents. II. New 2,5-Bis(1-aziridinyl)-p-benzoquinone
derivatives. Chem. Pharm. Bull. 1972, 20, 1968-1979.
(27) Traver, R. D.; Siegel, D.; Beall, H. D.; Phillips, R. M.; Gibson,
N. W.; Franklin, W. A.; Ross, D. Characterization of a polymor-
phism in NAD(P)H: quinone oxidoreductase (DT-diaphorase) Br.
J . Cancer 1997, 75, 69-75.
(28) Carmichael, J .; DeGraff, W. G.; Gazdar, A. F.; Minna, J . D.;
Michael, J . B. Evaluation of tetrazolium- based semiautomatic
colorimetric assay; assessment of chemosensitivity testing.
Cancer Res. 1987, 47, 936-942.
(29) Beall, H. D.; Mulcahy, R. T.; Siegel, D.; Traver, R. D.; Gibson,
N. W.; Ross, D. Metabolism of bioreductive antitumor compounds
by purified rat and human DT-diaphorase. Cancer Res. 1994,
54, 3196-3201.
(30) Hartley, J . A.; Berardini, M.; Souhami, R. L. An agarose gel
method for the determination of DNA interstrand cross-linking
applicable to the measurement of total and second-arm cross-
link reactions. Anal. Biochem. 1991, 193, 131-134.
(18) Beall, D. H.; Winski, S.; Swann, E.; Hudnott, A. R.; Cottrill, A.
S.; O’Sullivan, N.; Green, S. J .; Bien, R.; Siegel, D.; Ross, D.;
Moody, C. J . Indoloquinone antitumour agenst: Correlation
between quinone structure, rate of metabolism by rcombinant
human NQO1 and in vitro cytoxicity. J . Med. Chem. 1998, 41,
4755-4766.
(19) Li, R.; Bianchet, M. A.; Talalay, P.; Amzel, L. M. The three-
dimensional structure of NAD(P)H: quinone reductase, a fla-
voprotein involved in cancer chemoprotection and chemo-
therapy: mechanism of the two-electron reduction. Proc. Natl.
Acad. Sci. U.S.A. 1995, 92, 8846-8850.
J M991007Y