A comprehensive study of the active site residues of DT-diaphorase: Rational design of benzimidazolediones as DT-diaphorase substrates

Article abstract of DOI:10.1021/jm0104365

A series of quinone substrates were modeled into the active site of human DT-diaphorase and minimized. Correlation of these models with the substrate specificity k_cat/K_m provided insights into the structural requirements of quinone s

Full text of DOI:10.1021/jm0104365

J . Med. Chem. 2002, 45, 1211-1220
1211
A Com p r eh en sive Stu d y of th e Active Site Resid u es of DT-Dia p h or a se: Ra tion a l
Design of Ben zim id a zoled ion es a s DT-Dia p h or a se Su bstr a tes†
Ali Suleman and Edward B. Skibo*
Department of Chemistry and Biochemistry, Arizona State University, Tempe, Arizona 85287-1604
Received September 17, 2001
A series of quinone substrates were modeled into the active site of human DT-diaphorase and
minimized. Correlation of these models with the substrate specificity kcat/Km provided insights
into the structural requirements of quinone substrates. The W105, F106, and H194 residues
can influence the position of the quinone substrate in the active site resulting in formation of
one of the two possible Michael anions resulting from hydride transfer from FADH2. Electron
withdrawing groups on the substrate can stabilize these anions resulting in excellent substrate
specificity. Inspection of models indicated that the W-105 and F-106 residues form parallel
walls that will accommodate large polycyclic substrates. Thus excellent polycyclic substrates
of DT-diaphorase were designed. However, the placement of tetrahedral centers on these
polycyclic substrates interfered with the W-105 and the F-106 residues resulting in their
exclusion from the active site. The histidine (H194) residue permits recognition of substrate
enantiomers as a result of hydrogen bonding interactions. As a result of this study, it will be
possible to design poor to excellent substrates of DT-diaphorase and take advantage of varying
levels of this enzyme in histologically different cancers.
Ch a r t 1
The enzyme DT-diaphorase is an NADPH-dependent
reducing enzyme that reduces quinones and other
substrates to the corresponding two-electron reduction
products. This enzyme is elevated in some cancers, and
the reduction process can lead to either activation or
inactivation of antitumor agents, Chart 1.1-4 The reduc-
tive activation process is well-known with substrates
such as mitomycin C,5,6 indoloquinones,7,8 and the
PBIs,9,10 which are converted to the corresponding
cytotoxic hydroquinones. In contrast, some environmen-
tal toxins1,11-13 and the APBI antitumor agents9,14 are
inactivated upon DT-diaphorase reduction. A recent
report from this laboratory revealed that the level of
substrate specificity for DT-diaphorase influences cyto-
toxicity, cancer specificity, and in vivo toxicity.15 The
rational design of DT-diaphorase substrates is therefore
essential to take advantage of DT-diaphorase levels in
cancers and control toxicity.
developed in this study will be used to develop new DT-
diaphorase activated antitumor agents.
Resu lts a n d Discu ssion
In the present study, molecular modeling and sub-
strate specificity determinations were performed with
the diverse quinone systems developed in this laboratory
over the years. These systems are all benzimidazole-
based quinones but vary with respect to size (tricyclic
to pentacyclic), stereochemistry, and substituents. From
these models it was possible to gain insights into the
mechanism of DT-diaphorase reduction as well as the
role of amino acid residues in the active site. The models
Qu in on e Su bstr a tes a n d Th eir P r ep a r a tion . The
quinone substrates 1-5 shown in Chart 2 represent a
structurally diverse series of benzimidazole-based com-
pounds that permitted the development of a computer
model able to predict substrate specificity. We main-
tained the benzimidazole structure throughout the
series so as to keep the quinone redox potentials within
the same range. Variables include stereochemistry,
hydrogen bonding capability, and substrate dimension.
The preparation of series 1a -c was recently re-
ported,16 and the new analogue 1d was readily prepared
by the reaction of 1a with N,N-dimethylcarbamyl chlo-
ride. Series 2 represents the tetrahydropyrido version
of series 1. Access to both enantiomers of 2 in pure form
was possible starting with (D) or (L) 2,5-diaminopen-
tanoic acid as outlined in Scheme 1. The synthetic
methodology outlined in Scheme 1 has been described
† Abbreviations: TMS, tetramethylsilane; TLC, thin layer chroma-
tography; HPLC, high performance liquid chromatography; PDB,
protein data bank; CVFF, consistent valence force field; RMSD, root
mean square distance; FAD, flavin adenine dinucleotide; FADH2, flavin
adenine dinucleotide reduced; NAD+, nicotinamide adenine dinucle-
otide; NADH, Nicotinamide adenine dinucleotide reduced; NADPH,
nicotinamide adenine dinucleotide phosphate reduced; NSC non-small-
cell.
* To whom correspondence should be addressed. Telephone: 480-
965-3581. Fax: 480-965-2747. E-mail: ESkibo@ASU.ed.
10.1021/jm0104365 CCC: $22.00 © 2002 American Chemical Society
Published on Web 02/09/2002

1212 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 6
Suleman and Skibo
Ch a r t 2
Sch em e 1
Sch em e 2
in the literature.16,17 The enantiomerically pure forms
of series 39 and 414 were carried out as reported in the
literature. The preparation of 5 involved the double
annulation methodology illustrated in Scheme 2.
Mod elin g of Su bstr a tes in to th e Hu m a n DT-
Dia p h or a se Active Site a n d Su bstr a te Sp ecificity
(kca t/Km ) Deter m in a tion s. This section discusses the
generation of quinone substrate models in the human
DT-diaphorase active site employing the crystal struc-
ture of rat DT-diaphorase. The rat enzyme crystal
structure has a duroquinone substrate bound to the
active site whereas the human enzyme has no quinone
substrate. Models of the human DT-diaphorase active
site bearing quinone substrates were generated by a
series of superimpositions and minimizations described
in the Experimental Section. In addition, active site
energies of enzyme-bound substrates were determined
and correlated with measured values of substrate
specificity.
(1D4A). Previously, studies of substrate reductions
by human and rat DT-diaphorases have been carried
out employing computer modeling into crystal struc-
tures.19,21-23
From computer-generated models of the respective
active sites shown in Figure 1, the human diaphorase
active site is more compact than the rat enzyme active
site in that W105 and H194 are closer to each other.
These models show that the His-194 residue in human
model is closer to the other active site residues than the
rat model and therefore better able to undergo hydrogen
bonding interactions with substrates.24 Since the posi-
tions of the important active site residues (other than
His 194) are similar for both enzymes, superimposition
of the duroquinone substrate complex of rat enzyme into
the human active site is feasible.
The active site of DT-diaphorase has been defined
from crystal structures as well as from substrate studies
of both rat and human enzymes.18-20 Human DT-
diaphorase (from non-small cell lung NCI-H460) has
been cloned18 and its crystal structure determined
The superimposition process is described below:
First, the quinone ring of 1-4 was superimposed onto
the duroquinone ring of the minimized A-chain of the

Benzimidazolediones as DT-Diaphorase Substrates
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 6 1213
Ta ble 1. DT-Diaphorase Kinetic Parameters and Calculated
Active Site Energies
kcat/Km
(M-1 s-1
×105
13.9
6.4
8.9
12
9.2
5.2
Km
(M)
active site
energy
(kcal/M)
)
kcat
compd
(s-1
)
×105
plot
1a -(S)
1a -(R)
1b-(R)
1b-(S)
1c-(R)
1c-(S)
1d -(R)
1d -(S)
2-(R)
2.64
3.26
2.7
0.19
0.51
0.3
-105
-98.3
-111
-116
-108
-117
-170
-138
-115
-113
-112
-127
-105
-106
-78
A
A
A
A
A
B
A
B
B
B
B
B
B
B
A
A
2.4
0.2
1.47
1.87
9.56
3.955
4.92
2.972
3.81
3.85
4.845
1.735
3.811
3.74
0.16
0.36
0.45
0.42
4.66
1.58
1.79
0.59
12.2
2.5
21.3
9.4
1.06
1.88
1.54
4.72
0.4
0.69
1.18
14.5
2-(S)
3a -(R)
3a -(S)
3b-(R)
3b-(S)
4c
1.18
0.26
5
-109
F igu r e 1. Active sites of (A) human cancer (from NCI-H460
non-small-lung) and (B) rat DT-diaphorase from the protein
database, 1D4A and 1QRD. Important amino acid residues are
labeled.
F igu r e 2. Plots of kcat/Km versus minimized energy (kcal/mol)
for quinones 1-5. The data reveal the presence of two linear
plots: plot A with correlation coefficient ) 0.88 and plot B with
correlation coefficient ) 0.95. These respective plots correspond
to the two possible transition states for hydride transfer.
1QRD dimer. There are two to four ways to superimpose
1-4 on duroquinone, so each possible superimposition
was made and the lowest energy structure was chosen
for the next step. The superimposition with the lowest
energy was chosen as the most likely substrate/enzyme
complex structure. This process correctly predicted that
4a cannot be a DT-diaphorase substrate since none of
the possible superimpositions afforded a stable struc-
ture. In contrast, the most stable superimposed struc-
tures of 1-2 show the aziridinyl ring in the same
position as that of EO9 in the DT-diaphorase active
site.23
Second, the 1QRD:A chain bearing the new quinone
substrate was added to the 1D4A:A crystal structure
by superimposition of the FAD cofactors. After deleting
the 1QRD:A chain and the rat FAD, the optimal binding
orientation of the 1D4A:A chain with the quinone
substrate was then minimized utilizing the consistent
valence force field (CVFF). The minimization was either
carried out for 5000 iterations or until RMS of less than
0.05 was reached.
Third, electrostatic and van der Waals energies (ac-
tive site energies) were calculated for the active site
residues (W-105, F-106, Y-155, H-161, H-194) along with
the FAD cofactor and the superimposed quinone. The
energies obtained from the 5000 iteration minimization
are listed in Table 1.
The 1D4A:A-bound quinone structures obtained as
described above were evaluated with respect to their
position with respect to the FAD cofactor. The distance
between the carbon center of the quinone (called “C2”
in Figure 3) that would accept hydride by Michael
addition and the N (5) of the isoalloxazine ring were
typically 3-4 Å in models. This distance would permit
hydride transfer to the quinone if the reduced cofactor
were present.
The catalytic parameters for quinone substrates are
also provided in Table 1. Using either kcat or Km in
correlations with active site energy is risky due to the
inclusion of nonproductive constants.25 In contrast, the
ratio kcat/Km provides the apparent second-order rate
constant (s-1 M-1) of the enzymatic transformation
without including any nonproductive processes.
The kcat/Km versus active site energy plots shown in
Figure 2 reveal a very good to excellent correlation for
analogues 1-5 (plots A & B) when the 5000 iteration
data were used. The presence of two linear plots in
Figure 2, rather than one plot with a large amount of
scatter may pertain to the two possible ways of hydride
transfer to the quinone substrate. The linear relation-
ships shown in Figure 2 (increasing kcat/Km with de-
creasing energy) is perplexing, if the calculated energy
pertains to that of the enzyme substrate complex. A
lower minimized energy for substrate binding will
translate to an unfavorable (lower) value for kcat/Km

1214 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 6
Suleman and Skibo
Sch em e 3
because more energy is required to reach the transition
state. On the other hand, the linear relationships shown
in Figure 2 can be explained if the calculated active site
energy pertains to the transition state energy. The
minimized active site structures, with quinone in posi-
tion to accept hydride, may in fact resemble the transi-
tion state complex.
Correlation of kcat/Km with energies obtained from
iteration to RMS of 0.05 afforded a zero slope plot
because the active site energies of all the quinone
substrates are nearly the same (110-120 kcal/mol). The
quinone substrates ultimately provide similar active site
structures upon complete minimization, perhaps be-
cause they are all based on the benzimidazole ring
system. These minimized structures do not account for
the different kcat/Km values for the substrates; therefore,
they must not reflect structures on the reaction path.
We conclude that minimization to RMS of less than 0.05
is not desirable, and higher energy minima actually
reflect structures on the reaction path.
F igu r e 3. Inspection of molecular models of 1-5 revealed that
plots A and B in Figure 2 are associated with different
substrate orientations in the human DT-diaphorase active site.
Illustrated in this figure are representative models of these
two substrate orientations (A: plot A hydride attack; B: plot
B hydride attack). Both active site models show the substrate
in yellow with the site of hydride attack on the quinone labeled
“C2” and the nitrogen bearing the hydride labeled “N-5.” Note
that oxidized flavin is shown in these models, and therefore a
hydride is not present on N-5.
The findings cited above led to the conclusions dis-
cussed below.
F AD a n d Tyr osin e (Y 155) Resid u es. The tyrosine
(Y 155) and histidine (H 161) residues are involved in
quinone reduction when the FAD cofactor is bound in
its reduced form (FADH2). Previously, Li et al. suggested
that hydride transfer occurs directly to the carbonyl
oxygen,20 while Schlegel et al. suggested hydride trans-
fer directly to a quinone carbonyl carbon.26 The former
hydride transfer is not electrostatically feasible due to
the electronegativity of oxygen compared to carbon. A
third mechanism for quinone reduction involves hydride
transfer by a Michael addition followed by tautomer-
ization to the hydroquinone, Scheme 3.19 In fact, this
laboratory previously reported supporting evidence for
hydride quinone reduction by Michael addition.27
The two possible hydride transfer processes shown in
Scheme 4 explain the presence of two linear relation-
ships, plots A and B in Figure 2. Inspection of the
minimized structures for enzyme-bound substrates fall-
ing on plot A and plot B in Figure 3 revealed that the
Michael acceptor carbon is different for each respective
set of substrates. Substrates on plot B of Figure 2
possess smaller kcat/Km values than plot A because the
Michael addition anions of the latter are in a more
electron rich environment.
Not all DT-diaphorase substrates bind like duro-
quinone. Thus other quinone positions with respect to
the cofactor are tolerated as long as the hydride transfer
could occur. Stereochemical and bulk changes are suf-
ficient to reposition the substrate in the active site. In
substrates 1-5, formation of only two Michael anions
is possible. The cofactor does not have access to the other
two sites for hydride attack; see Scheme 4. The Y155
residue is not strictly required for the Michael addition,
a conclusion previously obtained from active site muta-
tion studies.21 Substrates 4 and 5 are therefore reduced
even though steric bulk prevents involvement of the
Y155 residue. This residue may still have a catalytic
role, such as in a charge relay process, but only if the
substrate is too electron rich to support anion develop-
ment. The Michael hydride addition mechanism can be
used in the rational design of outstanding DT-diapho-
rase substrates. By designing a system that stabilizes
the developing anion by resonance (inset of Scheme 4),
we have achieved kcat/KM values of 600 M-1 s-1. In fact,
Phillips and co-workers have reported that similar
2-carbonyl substituted indoloquinones are excellent
substrates for DT-diaphorase.23

Benzimidazolediones as DT-Diaphorase Substrates
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 6 1215
Sch em e 4
Tr yp top h a n (W-105) a n d P h en yla la n in e (F -106)
Resid u es. The tryptophan (W-105) and phenylalanine
(F-106) residues form parallel walls by a stacking
interaction. On the basis of modeling studies with the
aziridinyl quinone EO9,23 steric interactions will cause
the aziridinyl ring to face away from the F-106 residue.
At the same time, the aziridinyl ring will undergo
hydrophobic interactions with W-105. Both W-105 and
F-106 also limit the bulk of the 7-substituent of the PBI
system, and substantial decreases in DT-diaphorase
substrate activity have been observed when a butyl
group is substituted for the methyl group normally
found at the PBI 7-position.9
It was postulated that the tryptophan (W-105)/phe-
nylalanine (F-106) “sandwich” could be exploited for
substrate design by utilizing systems that could inter-
calate between these rings, Figure 4. The dipyrroloimi-
dazobenzimidazoles 4 and dipyridoimidazobenzimida-
zole 5 systems were designed and found to be excellent
substrates for human DT diaphorase. In addition, the
analogues 4a and 4b were studied as systems that
would be excluded from the active site by steric interac-
tions with the tryptophan (W-105)/phenylalanine (F-
106) “sandwich.”
F igu r e 4. Quinone 4c modeled into human DT-diaphorase
active site showing the substrate sandwiched between tryp-
tophan (W105) and phenylalanine (F106) residues.
Quinones 4b,c were found to be excellent substrates
for human and rat DT-diaphorase while 4a could not
be reduced by either enzyme. It was concluded from the
computer-generated models of active site bound 4b,c
that the pentacyclic quinone fitted snugly in the active
site of DT-diaphorase. The binding of 4b was such that
the unsubstituted pyrrolo ring was sandwiched between
the W105/F106 residues. In the case of 4a , where both
pyrrolo rings bear acetate, steric hindrance with the
W105/F106 residues precluded substrate activity. Mod-
eling studies correctly predicted that the unsubstituted
pyrido derivative 5 would be an even better substrate
than 4c mainly due to increased van der Waals interac-
tions with the W105/F106 residues.
role in the reductive detoxification of polycyclic quinones
by DT-diaphorases.12,13
Histid in e (H194) Resid u e. Ongoing work in this
laboratory has shown that this residue is useful for the
enantioselection of cancers.24 The position of the H194
residue in the human DT-diaphorase active site (Figure
1) results in a hydrogen bonding interaction between
the histidine and the carbonyl of the 3-substituent of
1b, 1c, and 3a . Since the S enantiomers are more able
to undergo this interaction, human histological cancers
are more sensitive to the S than the R enantiomers.24
In contrast, the H194 residue of rat enzyme is not
positioned for hydrogen bonding to the substrate. Figure
5 shows the substrate S-3a in the human and the rat
DT-diaphorase active sites. The absence of this hydro-
gen bonding interaction in the rat enzyme is manifested
The conclusion of this section is that DT-diaphorase
can transfer electrons to large flat multi-ring systems
as a result of van der Waals interactions with the W105/
F106 residues. These residues very likely play the same

1216 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 6
Suleman and Skibo
Con clu sion s
Previous reports have dealt with the modeling of
quinone substrates into the DT-diaphorase active
site.19,20,28 The present study does likewise and also
shows that quantitative relationships of predictive value
can be developed from DT-diaphorase models. A series
of benzimidazole-based quinone substrates were mod-
eled into the active site of human NSC 460 DT-
diaphorase and minimized. The models and calculations
lead to the following conclusions:
The measured values of the kcat/Km were correlated
with the minimized energy to afford two linear plots
reflecting two possible Michael hydride transfer pro-
cesses. Hydride transfer by Michael addition to quinones
was suggested as a DT-diaphorase reduction mecha-
nism19 and documented in a chemical system.27 Most
complex quinone substrates (mitomycin C and EO9)
may likewise undergo one of two possible Michael
additions, depending on their position in the active site.
It is concluded that the “duroquinone” structure is not
preserved in all quinone substrates due to steric inter-
actions elsewhere in the DT-diaphorase active site. With
two hydride transfer processes possible, repositioning
of substrates will still result in reduction.
In a previous study, it was surprising to observe that
rat DT-diaphorase could reduce polycyclic quinone
substrates such as 4b,c.14 The human DT-diaphorase
does likewise and even reduces the larger substrate 5.
Minimized structures of human DT-diaphorase bound
to 4b,c and 5 are very stable with two rings of the
pentacyclic systems sandwiched between W-105 and
F-106. These amino acid residues very likely intercalate
polycyclic quinones in the DT-diaphorase active site to
facilitate reductive detoxification.12,13
Finally, the H194 residue of human DT-diaphorase
functions as a hydrogen bond donor to the carbonyl of
the 3-acetamido and carbamido substituents of quinone
series 1 and 3.
F igu r e 5. Models of quinone substrate S-3a bound to rat (A)
and human (B) DT-diaphorase active sites. These models
illustrate the different position of the histidine residue in the
active sites of both enzymes.
Exp er im en ta l Section
All analytically pure compounds were dried under high
vacuum in a drying pistol over refluxing toluene. Elemental
analyses were run at Atlantic Microlab, Inc., Norcross, GA.
Elemental analyses are reported for purified final products
namely, 1d . All TLCs were performed on silica gel plates using
a variety of solvents and a fluorescent indicator for visualiza-
tion. IR spectra were taken as thin films and the strongest
absorbances reported. 1H NMR and 13C NMR spectra were
obtained from a 300 MHz and a 75 MHz spectrometer,
respectively, unless otherwise specified. All chemical shifts are
reported relative to TMS.
by a 20-fold increase in KM value for S-3a compared to
human enzyme.
It should be pointed out that hydrogen bonding
between the 3-substituent of 1 and 3 is not required for
substrate activity in human DT-diaphorase. Thus, both
the R and S enantiomers are substrates, with the latter
possessing somewhat lower active site energies and
therefore higher kcat/Km values. In some instances,
interactions between the 3-substituent and H-194 re-
sults in repositioning the substrate. Substrates 1c and
1d fall on plot A of Figure 2 as the R enantiomers, but
the S enantiomers fall on plot B.
The findings cited above led to the conclusions dis-
cussed below. Although rat and human DT-diaphorases
differ in many ways,19 the H194 residue can readily
distinguish the enzymes stereochemically with the
appropriate substrate. The H194 residue can influence
the position of the quinone substrate in the active site
resulting in formation of one of two Michael anions.
Cancer enantioselection relies on the H194 residue in
human DT-diaphorase,24 and rat enzyme is therefore
unsuitable for substrate specificity determination of
candidate compounds.
3-N,N′-Dim eth ylca r ba m id o-6-a zir id in yl-2,3-d ih yd r o-7-
m eth yl-1H-p yr r olo[1,2-a ]ben zim id a zole-5,8-d ion e (1d ). A
solution of 3-amino-6-aziridinyl-2,3-dihydro-7-methyl-1H-pyr-
rolo[1,2-a]benzimidazole-5,8-dione 1a (35 mg, 0. 136 mmol),
anhydrous p-dioxane (9 mL), triethylamine (5 mL), and N,N-
dimethylcarbomyl chloride (0.15 mL, 1.6 mmol) was placed
under nitrogen. The reaction was stirred at 60 °C for 26 h and
then concentrated in vacuo. The residue was dissolved in 8%
methanol in methylene chloride and purified by flash column
chromatography using 8% methanol in methylene chloride as
the eluent. Evaporation of solvent gave a dark red solid which
was recrystallized from chloroform/hexane: 26.8 mg (60%)
yield; mp >260 °C; TLC (10% methanol in chloroform) Rf )
0.64; IR (KBr pellet) 3306, 3001, 2926, 2361, 1668, 1635, 1570,
1
1518, 1379, 1313, 1215, 1138 cm-1; H NMR (CDCl3) δ 6.39
(1H, d, J ) 6.6 Hz, amide proton), 5.25-5.22 (1H, m, 3-me-
thine), 4.30-4.23 and 4.06-3.99 (2H, 2m, 1-methylene), 3.11-

Benzimidazolediones as DT-Diaphorase Substrates
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 6 1217
3.01 and 2.92-2.68 (2H, 2m, 2-methylene), 2.95 (6H, s, 2-CH3),
2.38 (4H, s, aziridinyl protons), 1.97 (3H, s, 7-methyl); MS (EI)
m/z 329 (M+). The R and S enantiomers of 1d were prepared
from the enantiomerically pure 1a . Anal. Calcd (C16H19N5O3‚
1.65 H2O) C, H, N.
(S)-(-) a n d (R)-(+)-4-Aceta m id o-8-m eth yl-1,2,3,4-tetr a -
h yd r op yr id o[1,2-a ]ben zim id a zole (6) were prepared from
(L)- and (D)-2,5-diaminopentanoic acid, respectively, by the
general three-step procedure described below using either the
indicated portions of reactants or proportional amounts thereof.
To a refluxing solution of 5 g (0.03 mol) of 2,5-diaminopen-
tanoic acid in 50 mL of water was added 13.5 g (0.06 mol) of
CuCO3‚Cu(OH)2, and the resulting mixture was refluxed for
3 h. The reaction mixture was filtered hot, and the precipitate
was washed 2× with 20 mL portions of hot water. The deep
blue filtrate was allowed to cool to room temperature, and (10
g, 0.12 mol) NaHCO3 was added to the solution. To the stirred
mixture was added a solution of 9.6 g (0.04 mol) of 3-bromo-
4-nitrotoluene in 96 mL of acetone. The resulting solution was
refluxed overnight during which a dark green solid separated
out from the reaction mixture. The precipitate was washed
with water, acetone, and diethyl ether sequentially to yield
1.86 g of a dark green solid. The precipitate was suspended in
60 mL of boiling water, and then hydrogen sulfide gas was
bubbled through it for 25 min with continued heating. The
mixture was then stirred for additional 10 min, treated with
60 mL of 6 N HCl, and filtered hot. The filtrate was cooled to
room temperature and then neutralized to pH 7 by dropwise
addition of 4 N NaOH. The neutralized mixture was left in
the refrigerator overnight during which yellow precipitate was
obtained: 1.04 g (13%) yield.
The yellow precipitate was suspended in 120 mL of metha-
nol containing 4 mL of 4 N HCl. Nitrogen was bubbled through
this solution for 10 min and then 420 mg of 5% Pd on carbon
was added, and the reaction mixture was hydrogenated at 50
psi overnight. The reaction mixture was filtered through
sintered funnel, and the black solid was washed 2× with 20
mL portions of methanol. The solvent was removed under
reduced pressure to afford an off-white residue.
The residue obtained above was immediately dissolved in
160 mL of 4 N HCl, and the resulting solution was refluxed
for 16 h. At the end of the reaction time, the reaction mixture
was neutralized to pH 6-7 with cooling on an ice bath. To
this mixture were added 150 mL of acetic anhydride and 50
mL of acetic acid, and the reaction mixture was stirred for 6
h. The solution was neutralized with saturated NaHCO3
solution and then extracted 3× with 60 mL portions of
chloroform. The organic layer was dried over sodium sulfate
and then concentrated in vacuo to yield either (S)-6 or (R)-6
as light brown solids:
The deacetylation was carried out by dissolving 195 mg (0.8
mmol) of 6 in 400 mL of 1.2 N hydrochloric acid and refluxing
for 12 h. The solvent was removed in vacuo, and the off-white
residue was dried under vacuum for 2 h. Trifluoroacetylation
was carried out by dissolving the dried product in 8 mL of
trifluoroacetic acid followed by addition of 8 mL of trifluoro-
acetic anhydride. The reaction mixture was stirred for 20 min
at room temperature and then poured into 100 mL of 0.1 M
pH ) 7 phosphate buffer. This solution was extracted 3× with
35 mL portions of ethyl acetate, and the organic extracts were
then washed 2× with 100 mL portions of saturated sodium
bicarbonate and dried over anhydrous sodium sulfate. The
dried organic extract was concentrated to a residue that was
recrystallized from ethyl acetate/hexane to yield 160 mg (67%
yield) of a yellow solid: mp 195-198 °C; TLC (10% methanol
in chloroform) Rf ) 0.8; IR (KBr pellet) 3443, 2924, 2922, 2362,
1685, 1456, 1211, 1145 cm-1; 1H NMR (CDCl3) δ 8.04 (1H, bs,
amide proton) 7.58-7.55 (1H, d, 6-aromatic), 7.13-7.10 (1H,
d, 7-aromatic), 7.06 (1H, s, 9-aromatic), 5.21-5.13 (1H, m,
4-methine), 4.19-4.13 and 3.95-3.86 (2H, 2m, 1-methylene),
2.78-2.69, 2.38-2.26, 2.22-2.15 and 1.94-1.81 (4H, 4m,
2-methylene and 3-methylene), 2.49 (3H, s, 8-CH3); MS (EI)
m/z 297 (M+), 200 (M+ - COCF3). Rotation: S(-)[R]2D5
)
-15.6° (c ) 1.6, EtOH) and R(+)[R]2D5 ) +10.8°(c ) 0.74,
EtOH).
7-Br om o-4-tr iflu or oa ceta m id o-8-m eth yl-6-n itr o-1,2,3,4-
tetr a h yd r op yr id o[1,2-a ]ben zim id a zole (8) was prepared
from 7 by the following two-step procedure.
To a solution of 180 mg (0.61 mmol) of 7 in 10 mL of acetic
acid was added 0.6 mL of 0.8 M bromine in acetic acid. The
resulting mixture was stirred for 25 min at room temperature
and then quenched by addition of 200 mL of 0.1 M pH ) 7
phosphate buffer. The resulting mixture was extracted 3× with
35 mL portions of ethyl acetate. The organic extracts were
combined and washed 2× with 60 mL portions of saturated
sodium bicarbonate and then dried over anhydrous sodium
sulfate. The organic layer was concentrated, and residue was
recrystallized from ethyl acetate/hexane to afford 154 mg of a
pale yellow solid (67% yield): TLC (10% methanol in chloro-
form) Rf ) 0.47; IR (KBr pellet) 3277, 3095, 2957, 2567, 2361,
1703, 1560, 1450, 1188 cm-1; 1H NMR (CDCl3) δ 7.98 (1H, bs,
NH proton), 7.87 (1H, s, 6-aromatic proton), 7.13 (1H, s,
9-aromatic proton), 5.18-5.13 (1H, m, 4-methine), 4.18-4.12
and 3.98-3.89 (2H, 2m, 1-methylene), 2.75-2.68, 2.38-2.29,
2.27-2.18 and 1.97-1.93 (4H, 4m, 2-methylene and 3-meth-
ylene), 2.51 (3H, s, 8-CH3); MS (EI) m/z 377 and 375 (M+, 81Br
and 79Br), 280 and 278 (M+ - COCF3).
The 7-bromo-4-trifluoroacetamido-8-methyl-1,2,3,4-tetrahy-
dropyrido[1,2-a]benzimidazole (154 mg, 0.41 mmol) obtained
above was added slowly to chilled fuming nitric acid (10 mL)
with continuous stirring on an ice-salt bath. To this mixture,
acetic anhydride (0.6 mL) was added, and the reaction mixture
was stirred in an ice bath for another 5 min. The ice-salt bath
was removed, and the reaction mixture was stirred at room
temperature for 90 min. The reaction mixture was then poured
into 200 mL of ice water followed by neutralization with
saturated sodium bicarbonate solution. The aqueous layer was
extracted 3× with 60 mL portions of ethyl acetate, and the
extracts were dried over anhydrous sodium sulfate. The
solvent was removed in vacuo, and the yellow residue was
recrystallized using ethyl acetate and hexane to yield 116 mg
of a pale yellow solid (67% yield): mp 199-201 °C; TLC (10%
methanol in chloroform) Rf ) 0.62; IR (KBr pellet) 3229, 3055,
2957, 2928, 2361, 2337, 1720, 1543, 1375, 1213, 1161, 1039
(S)-6: Yield 67%; mp (dec) 243-245 °C; TLC (10% methanol
in chloroform) Rf ) 0.48; IR (KBr pellet) 3448, 3304, 2953,
2922, 2858, 2364, 1639, 1541, 1425, 1309, 1261, 1153 cm-1
;
1H NMR (CDCl3) δ 7.59-7.57 (1H, d, 6-aromatic), 7.12-7.10
(1H, d, 7-aromatic), 7.10 (1H, s, 9-aromatic), 5.22-5.15 (1H,
m, 4-methine), 4.21-4.16 and 3.96-3.86 (2H, 2m, 1-methyl-
ene), 2.71-2.66, 2.29-2.25, 2.19-2.10 and 1.78-1.71 (4H, 4m,
2-methylene and 3-methylene), 2.49 (3H, s, 8-CH3), 2.11 (3H,
s, methyl); MS (EI) m/z 243 (M+), 200 (M+ - COCH3).
Rotation: S(-) [R]2D5 ) -54° (c ) 1.8, EtOH).
(R)- 6: Yield 38%; mp (dec) 251-253 °C; TLC (10% methanol
in chloroform) Rf ) 0.44; IR (KBr pellet) 3448, 3290, 2957,
2862, 2359, 1716, 1637, 1547, 1427, 1271, 1261, 1153 cm-1
;
1H NMR (CDCl3) δ 7.617-7.587 (1H, d, 6-aromatic), 7.123-
7.099 (1H, d, 7-aromatic), 7.112 (1H, s, 9-aromatic), 5.19-5.12
(1H, m, 4-methine), 4.22-4.16 and 3.97-3.87 (2H, 2m, 1-me-
thylene), 2.78-2.72, 2.28-2.21, 2.20-2.16 and 1.74-1.61 (4H,
4m, 2-methylene and 3-methylene), 2.49 (3H, s, 8-CH3), 2.10
(3H, s, methyl); MS (EI) m/z 243 (M+), 200 (M+ - COCH3).
Rotation: R(+) [R]2D5 ) +50° (c ) 2.2, EtOH).
cm-1; H NMR (CDCl3) δ 8.34 (1H, bs, NH proton), 7.36 (1H,
1
s, 9-aromatic), 5.19-5.12 (1H, m, 4-methine), 4.29-4.24 and
4.11-4.01 (2H, 2m, 1-methylene), 2.91-2.86, 2.42-2.36, 2.26-
2.17 and 2.01-1.88 (4H, 4m, 2-methylene and 3-methylene),
2.58 (3H, s, 8-CH3); MS (EI) m/z 419 (M+), 373 (M+ - NO2),
351, 333, 323, 305, 294, 278, 224, 198. Rotation: S(-)[R]D25
)
-15° (c ) 0.88, EtOH) and R(+)[R]2D5 ) +13° (c ) 0.90,
EtOH).
4-Tr iflu or oa cet a m id o-8-m et h yl-1,2,3,4-t et r a h yd r op y-
r id o[1,2-a ]ben zim id a zole (7) was prepared from 6 by the
following two-step procedure.
4-Am in o-7-azir idin yl-8-m eth yl-1,2,3,4-tetr ah ydr opyr ido-
[1,2-a ]ben zim id a zole-6,9-d ion e (2). To a well-stirred solu-

1218 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 6
Suleman and Skibo
tion of 66 mg (0. 156 mmol) of 8 in 10 mL of ethanol was added
100 mg (2.6 mmol) of sodium borohydride portion wise. The
resulting mixture was stirred at room temperature for 1 h.
The reaction mixture was then filtered through a sintered
glass funnel. The filtrate was concentrated in vacuo and then
loaded over a silica column and eluted with 8% methanol in
methylene chloride in chloroform. The fraction containing the
product was concentrated in vacuo and the residue recrystal-
lized from chloroform/hexane to yield a crude yellow solid (26
mg): yield 51%; Rf ) 0.42 (8% methanol in chloroform); 1H
NMR (CDCl3) δ 7.31 (1H, s, 9-aromatic), 4.32-4.27 (1H, m,
4-methine), 4.15-4.11 and 4.05-4.01 (2H, 2m, 1-methylene),
2.38-2.34, 2.21-2.05 and 1.84-1.11 (4H, 4m, 2-methylene and
3-methylene), 2.58 (3H, s, 8-CH3); MS (EI) m/z 325 (M+), 309,
296, 279, 249, 198.
IR (KBr pellet) 3288, 2939, 2852, 2802, 1658, 1597, 1533, 1504,
1
1417, 1130, 1111, 1031, 906 cm-1; H NMR (CDCl3) 9.06 and
6.89 (2H, 2s, aromatic protons), 8.24 (2H, s, amide protons),
2.76-2.75 (8H, t, pyrido protons), 2.17 (6H, s, methyl protons),
1.72-1.70 (8H, t, pyrido protons), 1.58 (4H, m, pyrido protons)
m/z 358 (M+), 315 (M+ - ketene).
To a solution consisting of 6 mL of 96% formic acid and 3
mL of 30% hydrogen peroxide was added the acetylated
product prepared as described above (860 mg, 2.4 mmol). The
resulting mixture was stirred at 70 °C for 2 h and then cooled
to room temperature. The reaction mixture was diluted with
20 mL of water, and the pH of the solution was adjusted to 7
(approximately) with concentrated ammonium hydroxide. The
resulting solution was extracted 5× with 20 mL portions
chloroform. The dried organic solvent (NaSO4) was removed
in vacuo to afford a crude product that was recrystallized from
chloroform/hexane: 202 mg (31%) yield; mp 104-108 °C. TLC
(10% methanol in chloroform) Rf ) 0.23; IR (KBr pellet) 3418,
2949, 1707, 1633, 1446 cm-1; 1H NMR (CDCl3) 7.963 and 7.117
(2H, 2s, aromatic protons), 4.146-4.116 (4H, t, pyrido protons),
3.202-3.171 (4H, t, pyrido protons), 2.207-2.179 (4H, m,
pyrido protons), 2.091-2.047 (4H, m, pyrido protons) m/z 266
(M+).
The unstable product obtained above(46 mg, 0.14 mmol) was
quickly dissolved in 40 mL of methanol, and the resulting
solution was hydrogenated in the presence of 5% Pd over
carbon (80 mg) at 50 psi for 5 h. The reaction mixture was
filtered over sintered glass funnel, and the filter cake was
washed 2× with 20 mL portions of methanol. The solvent was
removed in vacuo and the residue redissolved in 20 mL of
water containing 0.3 g of KH2PO4. The resulting mixture was
combined with another 20 mL solution containing 1 g of KH2-
PO4 and 0.8 g (3 mmol) of Fremy salt. The reaction mixture
was stirred at room temperature for 5 h. The solution was then
concentrated under high vacuum and the residue loaded over
Bakerbond phenyl reverse-phase column (15 g resin) prepared
in water. The unstable aminoquinone was eluted as yellow
fraction from the column with water, and the combined
fractions were concentrated to dryness. The residue was
resuspended in 20 mL of methanol and 0.8 mL of aziridine.
The resulting mixture was stirred at room temperature for 5
h. The solution was concentrated in vacuo, and the residue
was purified by flash column chromatography using 8%
methanol in chloroform as the eluent. The organic fraction was
concentrated in vacuo, and the residue was recrystallized from
chloroform/hexane to yield a red solid (6.8 mg): yield 18%
overall; mp (dec) >235 °C. TLC (10% methanol in chloroform)
Rf ) 0.41; IR (KBr pellet) 3454, 2935, 2364, 2343, 1676, 1637,
To a solution consisting of 4 mL fuming nitric acid and 4
mL of concentrated sulfuric acid cooled in an ice-salt bath to
0 °C was added 190 mg of 5 (0.714 mmol), and the mixture
stirred for 15 min. At the end of the reaction time, the mixture
was poured over cracked ice and the pH adjusted to 6-6.5
using concentrated ammonium hydroxide. The resulting aque-
ous solution was extracted 4× with 20 mL portions of
chloroform. The chloroform extracts were combined and dried
(Na2SO4), and then concentrated to afford the crude product.
The product was recrystallized using chloroform/hexane to
yield a yellow solid: 110 mg (50% yield); mp 230-232 °C dec.
TLC (10% methanol in chloroform) Rf ) 0.19; IR (KBr pellet)
3431, 2951, 1653, 1496, 1365, 1244, 1028 cm-1 1H NMR
;
(CDCl3) 7.385 (1H, s, aromatic proton), 4.191-4.151 (4H, t,
pyrido protons), 3.265-3.223 (4H, t, pyrido protons), 2.233-
2.202 (4H, m, pyrido protons), 2.097-2.061 (4H, m, pyrido
protons); MS(EI mode) m/z 311 (M+), 281 (M+ - NO).
1,2,3,4,8,9,10,11-Octa h yd r od ip yr id o[1,2-a :1′,2′-a ′]ben zo-
[1,2-d :5,4-d ′]d iim id a zole-6, 12-d ion e(5) was prepared by the
two-step procedure outlined below.
1541, 1381, 1340, 1188 cm-1 1H NMR (CDCl3) δ 4.46-4.38
;
and 4.2-4.1 (3H, 3m, 4-methine and 1-methylene), 2.31-2.18,
2.03-1.91 and 1.74-1.66 (4H, 4m, 2-methylene and 3-meth-
ylene), 2.39 (4H, s, aziridinyl protons), 2.07 (3H, s, 8-methyl);
MS (EI) m/z 272 (M+). Anal. Calcd (C14H16N2O4) C, H, N.
A solution consisting of 98 mg (0.32 mmol) of 10, 50 mg of
5% Pd on carbon, and 50 mL methanol was shaken at 50 psi
H2 for 4 h. The mixture was filtered using sintered funnel,
and the filter cake was washed 2× with 20 mL portions of
methanol. The filtrate was removed in vacuo and then redis-
solved in 10 mL of water containing monobasic potassium
phosphate (500 mg). To this solution was then added 0.9 g of
Fremy salt and 25 mL of 0.1 M monobasic potassium phos-
phate in water. The resulting reaction mixture was then
stirred for 2 h, and the aqueous solution was extracted 3× with
20 mL portions of chloroform. The organic fractions were
combined, dried (Na2SO4), and then concentrated to a crude
solid. The solid was recrystallized from chloroform/hexane to
yield a yellowish solid: 34 mg (36% yield); mp > 260 °C; TLC
(10% methanol in chloroform) Rf ) 0.23; IR (KBr pellet) 3448,
1,3-Bis(p ip er id in o)-4,6-d in itr oben zen e (9). To 1.98 g (6.1
mmol) of 1,3-dibromo-4,6-dinitrobenzene in a round-bottom
flask was added 4 mL of piperidine dropwise. After completing
the addition, the reaction mixture was heated at 90 °C for 12
h. The reaction mixture was then poured over cracked ice and
stored in the refrigerator for 4 h. The precipitate was filtered
off, and the product was purified by flash chromatography
using silica gel with chloroform as eluent. The organic fractions
containing the product were combined, and the solvent was
removed in vacuo. The resulting solid was recrystallized from
chloroform/hexane to yield 1,3-piperidino-4,6-dinitrobenzene
9 as a yellow solid (1.77 g, 86% yield), mp 120-124 °C. TLC
(5% methanol in CHCl3) Rf ) 0.69; IR (KBr pellet) 2945, 2852,
1
2947, 1678, 1651, 1423, 1294, 1047 cm-1; H NMR (CDCl3) δ
1601, 1560, 1510, 1446, 1342, 1242, 856 cm-1 1H NMR
;
4.36-4.32 (4H, t, pyrido protons), 3.02-2.97 (4H, t, pyrido
protons), 2.06-1.94 (8H, m, pyrido protons). m/z 296 (M+).
Anal. Calcd. (C16H16N4O2‚0.8H2O) C, H, N.
(CDCL3) δ 8.67 and 6.27 (2H, 2s, aromatic protons), 3.14-3.11
(8H, t, pyrido protons), 1.78-1.74 (8H, t, pyrido protons), 1.69-
1.66 (4H, m, pyrido protons); m/z 334 (M+), 317 (M+ - OH).
Hu m a n a n d Ra t DT-Dia p h or a se Coor d in a tes. Crystal-
lographic coordinates for rat (1QRD) and human DT-diapho-
rase (1D4A)18,19 were obtained from the Protein Data Bank.
6-Nitr o-1,2,3,4,8,9,10,11-octa h yd r od ip yr id o[1,2-a :1′,2′-
a ′]ben zo[1,2-d :5,4-d ′]d iim id a zole (10) was prepared by the
four-step procedure outlined below. A solution containing 1.60
g (4.8 mmol) of 9 and 400 mg of 5% Pd on carbon in 200 mL
of acetic acid was shaken under 50 psi of H2 for 12 h. The
reaction mixture was filtered through sintered funnel to
remove the catalyst. To the filtrate was added 100 mL of acetic
anhydride, and the solution was stirred at room temperature
for 1.5 h. The solvent was reduced in vacuo to afford an oil, to
which 300 mL of diethyl ether was added resulting in crystal-
lization of the product as a white solid: 897 mg (52%) yield;
mp 223-226 °C; TLC (5% methanol in chloroform) Rf ) 0.2;
The coordinates for rat DT-diaphorase included bound FAD,
duraquinone (substrate), and Cibacron blue (inhibitor). The
human DT-diaphorase coordinates included FAD as the only
bound molecule. For modeling purposes, INSIGHT II from
Molecular Simulations, Inc. (San Diego) was used. Rat DT-
diaphorase exists as a dimer with two catalytic sites, and
human DT-diaphorase exists as a tetramer with four catalytic
active sites. All the superimpositions and minimizations were
performed on a single monomer unit bearing a single catalytic

Benzimidazolediones as DT-Diaphorase Substrates
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 6 1219
active site for rat (1QRD:A) as well as for human DT-
diaphorase (1D4A:A). After downloading the PDB coordinates
for rat and human DT-diaphorase, hydrogens were added, and
the potentials and charges of atoms for the entire monomer
were fixed. The FAD cofactor for rat DT-diaphorase had to
have the atom potentials assigned. The resulting structure was
minimized for 3000 iterations (RMSD ) 0.03) using the
consistent valence force field (CVFF). Alternatively, the entire
protein backbone was fixed after adding hydrogens, and the
structure was subjected to steepest descent minimization to
an RMS derivative value of less than 0.001 using consistent
valence force field. Either minimization method afforded a
protein structure suitable for modeling.
and the flow rate was adjusted at 0.7 mL/min with the mobile
phase consisting of hexane, 1,2-dichloroethane, and ethanol
[57:30:13]. Chirex column (00B-3014-RO) was used for the
separation of compound 6. Measured retention time in minutes
was as follows: 4.15 for S-6 and 6.26 for R-6.
Ack n ow led gm en t. We wish to thank Professor
David Ross for the generous supply of NSC 460 recom-
binant DT-diaphorase. This work was supported by the
National Science Foundation, the National Institutes
of Health, and the Arizona Disease Control Commission
Refer en ces
Mod elin g in to th e DT-Dia p h or a se Active Site. The
substrates for rat and human DT-diaphorase were constructed
utilizing fragments available in fragment libraries of BUILDER
module. The molecules were constructed and then energy
minimized for 500-1000 iterations before superimposing on
the enzyme bound substrate (duraquinone) of rat DT-diapho-
rase. These substrates are mainly heterocycles with short side
chains, and conformational searching was not necessary. There
are two to four possible ways for the substituted quinone
substrates 1-5 to be superimposed on the enzyme-bound
duroquinone. The optimal binding orientation of the substrate
was obtained by superimposing the quinones on duraquinone
in all two to four possible orientations. After each superimposi-
tion, duraquinone was deleted from the enzyme active site.
The potentials for the native structures were fixed, and the
superimposed quinone, along with the rat DT-diaphorase and
bound FAD, was associated as an assembly and then energy
minimized using DISCOVER module. This process was re-
peated for all two to four possible orientations. The orientation
with the lowest energy (van der Waals and electrostatic) was
chosen as the optimal enzyme-bound structure. This quinone-
rat DT-diaphorase assembly was then utilized as a template
for generating models for human enzyme. This was done by
superimposing the isoalloxazine ring of the FAD bound cofac-
tor from this assembly unto that of the FAD bound cofactor
in the human enzyme. After the superimposition, rat DT-
diaphorase enzyme and the rat FAD were removed from the
assembly, and the human DT-diaphorase bearing the quinone
substrate was made into an assembly and the energy mini-
mized for either 5000 iterations or until an RMSD of 0.05 Å
was reached (15 000 to 33 000 iterations).
(1) Rauth, A. M.; Goldberg, Z.; Misra, V. DT-diaphorase: Possible
roles in cancer chemotherapy and carcinogenesis. Oncol. Res.
1997, 9, 339-349.
(2) Rauth, A. M.; Melo, T.; Misra, V. Bioreductive therapies: An
overview of drugs and their mechanisms of action. Int. J . Radiat.
Oncol. Biol. Phys. 1998, 42, 755-762.
(3) Stratford, I. J .; Workman, P. Bioreductive drugs into the next
millennium. Anti. Cancer Drug. Des. 1998, 13, 519-528.
(4) Fitzsimmons, S. A.; Workman, P.; Grever, M.; Paull, K.; Ca-
malier, R.; Lewis, A. D. Reductase enzyme expression across the
National Cancer Institute Tumor cell line panel: correlation with
sensitivity to mitomycin C and EO9. J . Natl. Cancer Inst. 1996,
88, 259-69.
(5) Siegel, D.; Beall, H.; Senekowitsch, C.; Kasai, M.; Arai, H.;
Gibson, N. W.; Ross, D. Bioreductive Activation of Mitomycin C
by DT-Diaphorase. Biochemistry 1992, 31, 7879-7889.
(6) Cummings, J .; Spanswick, V. J .; Tomasz, M.; Smyth, J . F.
Enzymology of mitomycin C metabolic activation in tumour
tissue - Implications for enzyme-directed bioreductive drug
development. Biochem. Pharmacol. 1998, 56, 405-414.
(7) Naylor, M. A.; Swann, E.; Everett, S. A.; J affar, M.; Nolan, J .;
Robertson, N.; Lockyer, S. D.; Patel, K. B.; Dennis, M. F.;
Stratford, M. R. L.; Wardman, P.; Adams, G. E.; Moody, C. J .;
Stratford, I. J . Indolequinone antitumor agents: Reductive
activation and elimination from (5-methoxy-1-methyl-4,7-diox-
oindol-3-yl)methyl derivatives and hypoxia-selective cytotoxicity
in vitro. J . Med. Chem. 1998, 41, 2720-2731.
(8) J affar, M.; Naylor, M. A.; Robertson, N.; Stratford, I. J . Targeting
hypoxia with a new generation of indolequinones. Anti. Cancer
Drug. Des. 1998, 13, 593-609.
(9) Skibo, E. S.; Gordon, S.; Bess, L.; Boruah, R.; Heileman, J .
Studies of Pyrrolo[1,2-a]benzimidazole Quinone DT-Diaphorase
Substrate Activity, Topoisomerase II Inhibition Activity, and
DNA Reductive Alkylation. J . Med. Chem. 1997, 40, 1327-1339.
(10) Skibo, E. B. Pyrrolobenzimidazoles in cancer treatment. Expert
Opin. Ther. Pat. 1998, 8, 673-701.
(11) Moran, J . L.; Siegel, D.; Ross, D. A potential mechanism
underlying the increased susceptibility of individuals with a
polymorphism in NAD(P)H: quinone oxidoreductase 1 (NQO1)
to benzene toxicity. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 8150-
8155.
(12) J arabak, J .; Harvey, R. G. Studies on 3 Reductases Which Have
Polycyclic Aromatic Hydrocarbon Quinones as Substrates. Arch.
Biochem. Biophys. 1993, 303, 394-401.
Optimization of DT-diaphorase bound quinones was ob-
tained using the DOCKING module of INSIGHT II. The
consistent valence force field (CVFF) was employed for all
DOCKING purposes. The GRID was generated around the
active site residues (W-105, F-106, Y-155, H-161, H-194). The
rest of the residues of the monomeric enzyme were treated as
immovable or rigid area. The optimal binding orientation was
achieved by utilizing convergent gradient (Polak-Ribiere)
method for 300 iterations. Generally these docked structures
were very similar to the minimized structures obtained above.
(13) Swanson, M. S.; Haugen, D. A.; Reilly, C. A.; Stamoudis, V. C.
Protection by Uridine Diphosphoglucuronic Acid and Dt-Dia-
phorase Against the Cytotoxicity of Polycyclic Aromatic- Hydro-
carbons Isolated From a Complex Coal-Gasification Condensate.
Toxicol. Appl. Pharmacol. 1986, 84, 336-345.
(14) Schulz, W. G.; Skibo, E. B. Inhibitors of topoisomerase II based
on the benzodiimidazole and dipyrroloimidazobenzimidazole ring
systems: Controlling DT-diaphorase reductive inactivation with
steric bulk. J . Med. Chem. 2000, 43, 629-638.
(15) Xing, C.; Skibo, E. B.; Dorr, R. T. Aziridinyl Quinone Antitumor
Agents based on Indoles and Cyclopent[b]indoles: Structure
Activity Relationships for Cytotoxicity and Antitumor Activity.
J . Med. Chem. 2001, 44, 3545-3562.
(16) Craigo, W. A.; LeSueur, B. W.; Skibo, E. B. Design of Highly
Active Analogues of the Pyrrolo[1,2-a]benzimidazole Antitumor
Agents. J . Med. Chem. 1999, 42, 3324-3333.
(17) Skibo, E. B.; Islam, I.; Schulz, W. G.; Zhou, R.; Bess, L.; Boruah,
R. The Organic Chemistry of the Pyrrolo[1,2-a]benzimidazole
Antitumor Agents. An Example of Rational Drug Design. Synlett
1996, 297-309.
(18) Beall, H. D.; Mulcahy, R. T.; Siegel, D.; Traver, R. D.; Gibson,
N. W.; Ross, D. Metabolism of Bioreductive Antitumor Com-
pounds by Purified Rat and Human DT-Diaphorase. Cancer Res.
1994, 54, 3196-3201.
(19) Faig, M.; Bianchet, M. A.; Talalay, P.; Chen, S.; Winski, S.; Ross,
D.; Amzel, L. M. Structures of recombinant human and mouse
NAD(P)H: quinone oxidoreductases: species comparison and
structural changes with substrate binding and release. Proc.
Natl. Acad. Sci. U.S.A. 2000, 97, 3177-82.
Kin etics Stu d ies Usin g Recom bin a n t DT-Dia p h or a se.
Kinetics studies were carried out in 0.05 M pH 7.4 Tris buffer
under anaerobic conditions, employing Thunberg cuvettes. A
4.0 mM PBI stock solution was prepared in dimethyl sulfoxide
(DMSO). To the top port of the cuvette was added the PBI
stock solution, and to the bottom port was added the recom-
binant DT-diaphorase and NADH stock solutions in Tris
buffer. Both portions were purged with argon for 20 min and
equilibrated at 30 °C for 10 min. The ports were then mixed,
and the reaction was followed at 336 nm for 10 min to obtain
initial rates. The final concentrations of the mixture were 0.3
mM NADH, 1.3 × 10-6 to 6.7 × 10-5 M of PBI, and 1.4 × 10-9
M of recombinant DT-diaphorase. To calculate Vmax, the value
of ∆ꢀ must be obtained. ∆ꢀ was calculated from the initial and
final absorbances for complete PBI reduction; all values are
between 8000 and 9000 M-1 cm-1. All data were fitted to a
Lineweaver-Burke plot from which kcat and Km were obtained.
Ch ir a l High -P r essu r e Liqu id Ch r om a togr a p h y HPLC
separations of enantiomers were done on Chirex column (50
× 3.2 mm) from Phenomenex. The detector was set at 285 nm,

1220 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 6
Suleman and Skibo
(20) 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, 8846-880.
(21) Chen, S. A.; Wu, K. B.; Zhang, D.; Sherman, M.; Knox, R.; Yang,
C. S. Molecular characterization of binding of substrates and
inhibitors to DT-diaphorase: Combined approach involving site-
directed mutagenesis, inhibitor-binding analysis, and computer
modeling. Mol. Pharmacol. 1999, 56, 272-278.
(24) Huang, X.; Suleman, A.; Skibo, E. B. Rational Design of Pyrrolo-
[1,2-a]benzimidazole Based Antitumor Agents Targeting the
DNA Major Groove. Bioorg. Chem. 2000, 28, 324-337.
(25) Fersht, A. R. The Hydrogen Bond in Molecular Recognition.
Trends Biochem. Sci. 1987, 12, 301-304.
(26) Schlegel, B. P.; Ratnam, K.; Penning, T. M. Retention of NADPH-
linked quinone reductase activity in an aldo-keto reductase
following mutation of the catalytic tyrosine. Biochemistry 1998,
37, 11003-11011.
(22) Skelly, J . V.; Sanderson, M. R.; Suter, D. A.; Baumann, U.; Read,
M. A.; Gregory, D. S. J .; Bennett, M.; Hobbs, S. M.; Neidle, S.
Crystal structure of human DT-diaphorase: A model for interac-
tion with the cytotoxic prodrug 5-(Aziridin-1-yl)-2,4-dinitroben-
zamide (CB1954). J . Med. Chem. 1999, 42, 4325-4330.
(23) Phillips, R. M.; Naylor, M. A.; J affar, M.; Doughty, S. W.; Everett,
S. A.; Breen, A. G.; Choudry, G. A.; Stratford, I. J . Bioreductive
activation of a series of indolequinones by human DT- diapho-
rase: Structure-activity relationships. J . Med. Chem. 1999, 42,
4071-4080.
(27) Skibo, E. B.; Lee, C. H. Facile Oxidation of Methoxide to
Formaldehyde by a Heterocyclic Quinone. J . Am. Chem. Soc.
1985, 107, 4591-4593.
(28) Cavelier, G.; Amzel, L. M. Mechanism of NAD(P)H: quinone
reductase: Ab initio studies of reduced flavin. Protein Struct.
Funct. Genet. 2001, 43, (4), 420-432.
J M0104365