Radical Chemistry and Cytotoxicity of Bioreductive 3-Substituted Quinoxaline Di-N-Oxides

Article abstract of DOI:10.1021/acs.chemrestox.6b00133

The radical chemistry and cytotoxicity of a series of quinoxaline di-N-oxide (QDO) compounds has been investigated to explore the mechanism of action of this class of bioreductive drugs. A series of water-soluble 3-trifluoromethyl (4-10), 3-phenyl (11-19)

Full text of DOI:10.1021/acs.chemrestox.6b00133

Article
pubs.acs.org/crt
Radical Chemistry and Cytotoxicity of Bioreductive 3‑Substituted
Quinoxaline Di‑N‑Oxides
Robert F. Anderson,*,†,‡,§ Pooja Yadav,† Sujata S. Shinde,‡ Cho R. Hong,‡ Susan M. Pullen,‡
Johannes Reynisson,† William R. Wilson,‡,§ and Michael P. Hay‡,§
́
‡
§
†School of Chemical Sciences, Auckland Cancer Society Research Centre, and Maurice Wilkins Centre, University of Auckland,
Private Bag 92019, Auckland 1142, New Zealand
S
* Supporting Information
ABSTRACT: The radical chemistry and cytotoxicity of a series
of quinoxaline di-N-oxide (QDO) compounds has been investi-
gated to explore the mechanism of action of this class of biore-
ductive drugs. A series of water-soluble 3-trifluoromethyl (4−10),
3-phenyl (11−19), and 3-methyl (20-21) substituted QDO com-
pounds were designed to span a range of electron affinities
consistent with bioreduction. The stoichiometry of loss of QDOs
by steady-state radiolysis of anaerobic aqueous formate buffer
indicated that one-electron reduction of QDOs generates radicals
able to initiate chain reactions by oxidation of formate. The
3-trifluoromethyl analogues exhibited long chain reactions consistent with the release of the HO•, as identified in EPR spin
trapping experiments. Several carbon-centered radical intermediates, produced by anaerobic incubation of the QDO compounds
with N-terminal truncated cytochrome P450 reductase (POR), were characterized using N-tert-butyl-α-phenylnitrone (PBN) and
5-(diethoxyphosphoryl)-5-methyl-1-pyrroline-N-oxide (DEPMPO) spin traps and were observed by EPR. Experimental data
were well simulated for the production of strongly oxidizing radicals, capable of H atom abstraction from methyl groups.
The kinetics of formation and decay of the radicals produced following one-electron reduction of the parent compounds, both in
oxic and anoxic solutions, were determined using pulse radiolysis. Back oxidation of the initially formed radical anions by
molecular oxygen did not compete effectively with the breakdown of the radical anions to form oxidizing radicals. The QDO
compounds displayed low hypoxic selectivity when tested against oxic and hypoxic cancer cell lines in vitro. The results from this
study form a kinetic description and explanation of the low hypoxia-selective cytotoxicity of QDOs against cancer cells compared
to the related benzotriazine 1,4-dioxide (BTO) class of compounds.
the kinetics of formation and reactivity of the oxidizing radicals
arising from the breakdown of the initial radical anion formed
upon one-electron reduction of the compounds. Formation of
the putative oxidizing radicals is thought to be inhibited by the
back oxidation of the intermediary radical anion by its reaction
with molecular oxygen, (kO2), and that this is the basis for
hypoxia-selective cytotoxicity (as shown in Scheme 1). The
HO• had been suggested to be released following reduction of
QDO compounds bearing a 3-CF3 substituent, on the basis of
DNA strand cleavage and product analysis,15 and this has been
confirmed by EPR studies using the spin-trap 5-(diethoxyphos-
phoryl)-5-methyl-1-pyrroline-N-oxide (DEPMPO, Chart 1) with
two closely related analogues,17 as depicted for the 7-chloro-3-
trifluoromethyl QDO analogue in Scheme 1. Supporting DFT
calculations indicate that release of the HO• is more ener-
getically favorable from the protonated N4−O radical anion
(pathway 3a) than from the N1−O radical anion, even though
formation of both are approximately equi-energetic.17 A carbon-
centered (C-centered) radical was also trapped with hyperfine
INTRODUCTION
■
Heteroaryl N-oxides have been explored widely as bioreductive
agents for their anticancer1−3 and antimicrobial properties.4,5
One class is the benzotriazine 1,4-dioxides (BTO), with the
best studied examples, tirapazamine (TPZ (1)) and SN30000
(2; Chart 1), exhibiting high hypoxia-selective cytotoxicity
against tumor cells in vitro6−9 and in vivo.1,9,10 TPZ has also
demonstrated activity in clinical trials.11,12 The isosteric
quinoxaline di-N-oxides (QDO) have also been investigated
as bioreductive prodrugs in vitro, with the discovery that the
3-aminoquinoxaline-2-carbonitrile QDO (3) and its analogues
also exhibit hypoxia-selectivity.3,13 However, while many
compounds based on the QDO scaffold exhibit considerable
cytotoxicity against hypoxic cells, the majority of the QDO
compounds described in the literature have been found to be
highly cytotoxic under oxic conditions as well, compromising
hypoxic selectivity.14−16 This low selectivity is in contrast to
most BTO compounds; understanding the reasons for the
increased aerobic toxicity of QDO compounds is crucial to the
further development of QDO compounds as hypoxia-selective
cytotoxins. Our hypothesis is that the difference in hypoxic
selectivity between BTO and QDO compounds is the result of
Received: April 22, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/acs.chemrestox.6b00133
Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology
Article
Chart 1. Structures of BTO, QDO, and Spin Trap Compounds
Scheme 1. Reductive Reaction Pathways of 7-Chloro-3-
trifluoromethyl QDO
Scheme 2. Preparation of QDOs 4−23
coupling constants (hfc) that are distinctly different from a
general C-centered radical but did not closely match those of an
aryl radical. An aryl radical has been identified as being formed
using N-tert-butyl-α-phenylnitrone (PBN, Chart 1), as the spin
trap following the one-electron reduction of the BTO com-
pound SN30000 (2)37 and an analogous pathway of formation
(path 4a) are depicted in Scheme 1, although direct evidence
for it was not found. Differences in the efficiencies and sta-
bilities between the two spin traps as well as the high reactivity
of aryl radicals causing the formation of secondary radicals may
confound a precise interpretation of these EPR data. It was also
noted in the previous study that the HO• was not released from
a 3-CH3 QDO analogue or from a 3-Ph QDO analogue upon
their one-electron reduction.
In the present study, EPR spectroscopy and pulse radiolysis
are used to identify and characterize the redox and kinetic
parameters of the radicals formed following the one-electron
reduction of a range of new QDO compounds. We hybridized
the known hypoxia-selective 2-cyano-3-amino3,13 and
2-thienoyl QDO15 cores using an electron-withdrawing amide
bearing a methylpiperazine side chain at the 2-position to
increase aqueous solubility and to thus facilitate evaluation
of in vitro cytotoxicity. We placed CF3, Ph, and Me substituents
at the 3-position to directly influence the electron affinity of the
QDO core and further modulated electron affinity through
substituents on the benzo ring of the QDO core. These com-
pounds were designed to explore the substituent effects on
radical properties and cytotoxicity by covering the electronic
space for efficient bioreduction.3,13 These chemical studies are
combined with in vitro cytotoxicity data to define the radical
chemistry underlying the cytotoxicity of QDO compounds.
Synthesis of QDO Compounds, 4−21. QDO compounds were
prepared using an established synthetic approach (Scheme 2).
Oxadiazoles 25b (R1 = Cl and R2 = H) and 25c (R1 = R2 = H)
were commercially available, while 25a (R1 = R2 = Cl) and 25d−i
(see Table 1) were prepared from the corresponding nitroanilines 24.
Condensation of 1-methylpiperazine with β-ketoesters 26 gave the
corresponding ketoamides 27. Coupling of β-ketoamides 27 with
oxadiazoles 25 generally proceeded in good yields, but these were
compromised by difficult chromatographic purifications. Isomeric
assignment was based on previous precedents.3,15 All compounds with
a piperazine substituent were formulated as HCl salts. Compounds 4
and 10 were synthesized as previously described.17
Analyses were carried out in the Campbell Microanalytical
Laboratory, University of Otago, Dunedin, NZ. All final products
were analyzed by reverse-phase HPLC, (Altima C18 5 μm column,
150 × 3.2 mm; Alltech Associated, Inc., Deerfield, IL) using an Agilent
HP1100 equipped with a diode-array detector. Mobile phases were
gradients of 80% acetonitrile/20% H2O (v/v) in 45 mM ammonium
formate at pH 3.5 and 0.5 mL/min. Final compound purity was
determined by monitoring at 330 100 nM and was >95%. Melting
points were determined on an Electrothermal 2300 Melting Point
Apparatus. NMR spectra were obtained on a Bruker Avance 400
spectrometer at 400 MHz for 1H and 100 MHz for 13C spectra.
Spectra were obtained in CDCl3 unless otherwise specified and were
referenced to Me4Si. Chemical shifts and coupling constants were
recorded in units of ppm and Hz, respectively. Low resolution mass
spectra were gathered by direct injection of methanolic solutions into a
Surveyor MSQ mass spectrometer using an atmospheric pressure
chemical ionization (APCI) mode with a corona voltage of 50 V and a
source temperature of 400 °C. High resolution mass spectra (HRMS)
were measured on a Bruker microTOF-QII hybrid quadrupole time of
flight (Q-TOF) mass spectrometer interfaced with either an elec-
trospray ionization (ESI) or atmospheric pressure chemical ionization
(APCI) probe allowing positive or negative ion detection. EtOAc
refers to ethyl acetate, EtOH refers to ethanol, and MeOH refers to
methanol (pet −400 mesh). DCM refers to dichloromethane, and
ether refers to petroleum ether boiling fraction 40−60 °C. The spin
MATERIALS AND METHODS
Caution: Azides should be considered hazardous and should not be
dried at elevated temperatures.
■
B
DOI: 10.1021/acs.chemrestox.6b00133
Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology
Article
Table 1. Structures and Physical Chemistry Properties of 3-Substituted QDO Compounds
X
R1
Cl
R2
Cl
Eo′ (mV)
103·k (s−1) (pH 7) 520 nm 104·kacid (s−1) 520 nm pKr 520 nm 106·kO2 (M−1 s−1
)
G-loss (μM·Gy−1
)
4
CF3
CF3
CF3
CF3
CF3
CF3
CF3
Ph
−254
−306
−356
−399
−415
−428
−430
−439
−491
−493
−509
−543
−566
−617
−620
−630
−493
−579
8
0.17 0.05
0.28 0.04
0.32 0.06
1.43 0.10
1.64 0.11
2.24 0.28
1.90 0.20
2.05 0.45
0.87 0.12
0.64 0.06
0.63 0.06
2.21 0.35
6.09 1.15
8.22 1.32
13.9 1.66
15.7 0.40
3.19 0.55
4.45 0.13
1.20 0.32
1.75 0.20
1.81 0.18
3.31 0.26
4.41 0.57
3.16 0.78
0.20 0.04
1.39 0.20
2.51 0.72
1.58 0.29
2.65 0.69
3.13 0.74
3.22 0.76
3.43 0.22
7.39 0.40
9.50 1.34
1.12 0.04
2.47 0.33
3.69 0.08
4.02 0.51
4.13 0.10
4.80 0.06
5.11 0.13
5.08 0.18
5.27 0.10
4.87 0.09
5.05 0.51
5.09 0.14
5.18 0.14
5.11 0.13
5.72 0.09
5.86 0.18
5.93 0.11
6.12 0.33
5.04 0.08
5.51 0.06
0.68 0.04
3.10 0.78
3.71 0.41
6.77 1.00
6.73 0.65
8.78 0.48
10.5 0.44
6.10 0.42
9.29 0.92
12.4 0.44
14.2 0.31
25.9 0.41
17.6 2.0
38.1 3.1
73.7 8.3
63.4 3.0
14.6 0.3
50.7 7.5
3.48 0.42
3.61 0.16
8.79 0.11
6.57 0.32
3.04 0.07
2.69 0.09
4.41 0.10
1.17 0.04
1.80 0.02
1.90 0.07
1.77 0.09
1.44 0.03
1.47 0.04
1.05 0.02
1.08 0.09
0.93 0.19
1.46 0.08
-
5
Cl
H
8
8
8
9
7
9
8
9
8
8
8
8
6
5
9
8
9
6
H
H
7
CH3
OCH3
CH3
-(CH2)3-
Cl
H
8
H
9
CH3
10
11
12
13
14
15
16
17
18
19
20
21
Cl
F
Ph
F
Ph
Cl
H
Ph
F
H
Ph
H
H
Ph
CH3
OCH3
CH3
-(CH2)3-
Cl
H
Ph
H
Ph
CH3
Ph
CH3
CH3
Cl
H
H
trap PBN was obtained from Fluka and freshly distilled, and DEPMPO
was obtained from Alexis.
of 1-oxide 25a (337 mg, 1.65 mmol), ketoamide 27a (392 mg,
1.65 mmol), CaCl2 (333 mg, 3.30 mmol), and K2CO3 (20 mg) in
EtOH (10 mL) was stirred at 50 °C for 16 h. The reaction was cooled
to 20 °C and the solvent evaporated. The residue was purified by
column chromatography, eluting with a gradient (0−5%) of MeOH/
DCM, to give quinoxaline dioxide 4 (237 mg, 34%) as a yellow
5,6-Dichlorobenzo[c][1,2,5]oxadiazole 1-Oxide (25a). Method A.
A solution of NaNO2 (0.52 g, 7.46 mmol) in water (10 mL) was
added dropwise to a stirred suspension of 4,5-dichloro-2-nitroaniline
24a (1.03 g, 4.98 mmol) in concentrated HCl (10 mL) and water
(30 mL) at 5 °C. The mixture was stirred at 5 °C for 1 h. This mixture
was added to a stirred solution of NaN3 (0.81 g, 12.5 mmol) and
NaOAc (8.17 g, 100 mmol) in water (70 mL), and the mixture was
stirred and allowed to warm to 20 °C over 3 h. The resulting pre-
cipitate was filtered, washed with water (2 × 10 mL), and air-dried to
give the crude azide. A solution of the azide in toluene (30 mL) was
added dropwise to toluene (50 mL) at reflux temperature and the
solution stirred at reflux temperature for 4 h. The solvent was
evaporated and the residue purified by chromatography, eluting with
5% EtOAc/pet. ether, to give 1-oxide 25a (0.77 g, 75%) as a yellow
powder: mp (EtOAc/pet. ether) 131−133 °C [lit.3 mp (MeOH)
130.8−131.2 °C]; 1H NMR (400 MHz, CDCl3) δ 7.87 (br s, 2H, H-4,
H-7); MS m/z 202.1 (MH+, 100%). Anal. Calcd for C6H2Cl2N2O2: C,
35.15; H, 0.98; N, 13.67. Found: C, 35.44; H, 1.09; N, 13.66%.
4,4,4-Trifluoro-1-(4-methylpiperazin-1-yl)butane-1,3-dione (27a).
Method B. A mixture of ethyl 4,4,4-trifluoro-3-oxobutanoate 26a
(1.0 g, 5.44 mmol) and 1-methylpiperazine (0.54 g, 5.44 mmol) in dry
toluene (5 mL) was stirred in a sealed tube at 110 °C for 12 h. The
mixture was cooled to 20 °C and partitioned between EtOAc (50 mL)
and water (50 mL). The aqueous layer was extracted with EtOAc
(50 mL). The combined organic fraction was extracted with aqueous
HCl (1 M, 2 × 25 mL). The combined aqueous extracts were cooled
and the pH adjusted to pH 9 with aqueous NH3 solution and then
extracted with EtOAc (3 × 75 mL). The combined organic fraction
was washed with brine (10 mL), dried, and the solvent evaporated.
The crude oil was purified by column chromatography, eluting with
5% MeOH/DCM, to give ketoamide 27a (811 mg, 63%) as a yellow
oil: 1H NMR (400 MHz, CDCl3) δ (mixture of keto- and enol forms)
15.26 (s, 0.5H), 5.81 (s, 0.5H), 4.12 (s, 1H), 3.68−3.72 (m, 2H),
3.51−3.54 (m, 2H), 2.42−2.48 (m, 4H), 2.33 and 2.31 (2 s, 3H); MS
m/z 239.2 (MH+, 100%). HRMS calcd for C9H14N2O2 (MH+) m/z
239.1002, found 239.1006 (1.5 ppm).
1
powder: mp (MeOH) 190−193 °C; H NMR (400 MHz, CDCl3) δ
8.71 (s 1H, H-5), 8.69 (s, 1H, H-8), 3.78−3.89 (m, 2H, CH2N), 3.37
(ddd, 1H, J = 13.1, 7.2, 3.4 Hz, CH2N), 3.28 (ddd, 1H, J = 13.1, 6.5,
3.4 Hz, CH2N), 2.61 (dt, 1H, J = 11.2, 5.0 Hz, CH2N), 2.55 (ddd, 1H,
J = 11.4, 6.5, 3.4 Hz, CH2N), 2.47 (dt, 1H, J = 11.2, 5.1 Hz, CH2N),
2.33−2.41 (m, 4H, CH2N, NCH3); 13C NMR (100 MHz, CDCl3) δ
155.5, 139.9, 139.0, 136.4, 136.3 (2), 130.2 (q, J = 34.5 Hz), 121.6,
121.5, 118.3 (q, J = 275.5 Hz), 54.0, 53.3, 45.5, 45.4, 41.5; MS 426.2
(MH+, 100%). Anal. Calcd for C15H13Cl2F3N4O3: C, 42.37; H, 3.08;
N, 13.18. Found: C, 42.42; H, 3.07; N, 12.89%. HPLC purity: 98.0%.
6-Chloro-3-(4-methylpiperazine-1-carbonyl)-2-(trifluoromethyl)-
quinoxaline 1,4-Dioxide (5). Quinoxaline dioxide 5 was prepared
from 5-chlorobenzo[c][1,2,5]oxadiazole 1-oxide (25b) and ketoamide
27a using Method C as previously described.17
2-(4-Methylpiperazine-1-carbonyl)-3-(trifluoromethyl)-
quinoxaline 1,4-Dioxide (6). Method C. Reaction of benzo[c][1,2,5]-
oxadiazole 1-oxide (25c) (150 mg, 1.10 mmol), ketoamide 27a (263 mg,
1.10 mmol), CaCl2 (244 mg, 2.20 mmol), and K2CO3 (10 mg) in EtOH
(10 mL) gave quinoxaline dioxide 6 (81 mg, 21%) as a yellow powder:
1
mp (MeOH) 172−175 °C; H NMR (400 MHz, CDCl3) δ 8.58−8.66
(m, 2H, H-5, H-8), 7.93−8.01 (m, 2H, H-6, H-7), 3.83−3.90 (m, 2H,
CH2N), 3.42 (ddd, 1H, J = 13.1, 7.4, 3.4 Hz, CH2N), 3.32 (ddd, 1H,
J = 13.1, 6.4, 3.4 Hz, CH2N), 2.64 (dt, 1H, J = 11.1, 5.0 Hz, CH2N), 2.58
(ddd, 1H, J = 11.4, 6.4, 3.4 Hz, CH2N), 2.47 (dt, 1H, J = 11.1, 5.3 Hz,
CH2N), 2.32−2.41 (m, 4H, CH2N, NCH3); 13C NMR (100 MHz,
CDCl3) δ 156.7, 138.5, 138.4, 136.2, 134.1, 133.3, 130.1 (q, J = 34.5 Hz),
120.9, 120.8, 119.3 (q, J = 275.5 Hz), 54.6, 53.9, 46.0 (2), 42.0; MS m/z
341.2 (MH+, 100%). Anal. Calcd for C15H15F3N4O3·1/2H2O: C, 50.00;
H, 4.60; N, 15.05. Found: C, 49.70; H, 4.25; N, 14.89%. HPLC purity:
98.0%.
6-Methylbenzo[c][1,2,5]oxadiazole 1-Oxide (25d). Method A.
Reaction of NaNO2 (2.76 g, 39.9 mmol) and 4-methyl-2-nitroaniline
(24d) (4.05 g, 26.6 mmol) and subsequently reaction with NaN3
(4.33 g, 66.5 mmol) and NaOAc (43.7 g, 532 mmol) in water (300 mL)
6,7-Dichloro-2-(4-methylpiperazine-1-carbonyl)-3-
(trifluoromethyl)quinoxaline 1,4-Dioxide (4). Method C. A mixture
C
DOI: 10.1021/acs.chemrestox.6b00133
Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology
Article
gave 1-oxide 25d (3.84 g, 96%) as a tan powder: mp (EtOAc/pet.
2-(4-Methylpiperazine-1-carbonyl)-3-(trifluoromethyl)-7,8-dihy-
dro-6H-cyclopenta[g]quinoxaline 1,4-Dioxide (10). Quinoxaline
dioxide 10 was prepared from 6,7-dihydro-5H-indeno[5,6-c][1,2,5]-
oxadiazole 1-oxide (25g) and ketoamide 27a using Method C as
previously described.17
ether) 94−95 °C (lit.18 mp 98 °C); H NMR (400 MHz, CDCl3)
1
δ 7.36 (br s, 1H, H-4), 7.09 (br s, 2H, H-7, H-5), 2.40 (s, 3H, CH3);
MS m/z 151.2 (MH+, 100%).
6-Methyl-3-(4-methylpiperazine-1-carbonyl)-2-(trifluoromethyl)-
quinoxaline 1,4-Dioxide (7). Method C. Reaction of 1-oxide 25d
(255 mg, 1.70 mmol), ketoamide 27a (405 mg, 1.70 mmol), CaCl2
(377 mg, 3.40 mmol), and K2CO3 (20 mg) in EtOH (10 mL) gave
quinoxaline dioxide 7 (52 mg, 8%) as a yellow powder: mp (EtOAc/
pet. ether) 175−178 °C; 1H NMR (400 MHz, CDCl3) δ 8.50 (d, 1H,
J = 8.9 Hz, H-8), 8.37 (d, 1H, J = 1.6 Hz, H-5), 7.75 (dd, 1H, J = 8.9,
1.6 Hz, H-7), 3.80−3.90 (m, 2H, CH2N), 3.40 (ddd, 1H, J = 13.1, 7.4,
3.4 Hz, CH2N), 3.30 (ddd, 1H, J = 13.1, 6.3, 3.4 Hz, CH2N), 2.60−
2.67 (m, 4H, 6-CH3, CH2N), 2.57 (ddd, 1H, J = 11.4, 6.3, 3.4 Hz,
CH2N), 2.46 (dt, 1H, J = 11.2, 5.1 Hz, CH2N), 2.32−2.39 (m, 4H,
CH2N, NCH3); 13C NMR (100 MHz, CDCl3) δ 156.8, 146.2, 138.2,
136.8, 136.2, 135.1, 128.3 (q, J = 34.0 Hz), 120.6, 119.7, 119.2 (q, J =
275.0 Hz), 54.6, 54.0, 46.1(2), 42.1, 22.1; MS m/z 371.2 (MH+,
100%). Anal. Calcd for C16H17F3N4O3: C, 51.89; H, 4.63; N, 15.13.
Found: C, 52.22; H, 4.75; N, 14.73%. HPLC purity: 96.2%.
6-Methoxybenzo[c][1,2,5]oxadiazole 1-Oxide (25e). Method A.
Reaction of NaNO2 (3.49 g, 50.6 mmol) and 4-methoxy-2-nitroaniline
(24e) (5.00 g, 29.7 mmol) with subsequent reaction with NaN3
(4.83 g, 74.3 mmol) and NaOAc (48.7 g, 594 mmol) gave 1-oxide 25e
(3.90 g, 79%) as a yellow powder: mp (EtOAc/pet. ether) 115−
117 °C [lit.18 mp 118−119 °C]; 1H NMR (400 MHz, CDCl3) δ (two
regioisomers) 7.52 (br s, 0.4H, H-7), 7.31 (br s, 0.6H, H-7), 7.05 (br s,
0.4H, H-5), 6.89 (br s, 0.6H, H-5), 6.62 (br s, 0.6H, H-4), 6.39 (br s,
0.4H, H-4), 3.90 (s, 3H, OCH3); MS m/z 167.2 (MH+, 100%).
6-Methoxy-3-(4-methylpiperazine-1-carbonyl)-2-
(trifluoromethyl)quinoxaline 1,4-Dioxide (8). Method C. Reaction of
1-oxide 25e (304 mg, 1.83 mmol), ketoamide 27a (436 mg, 1.83 mmol),
CaCl2 (406 mg, 3.66 mmol), and K2CO3 (40 mg) in EtOH (10 mL)
gave quinoxaline dioxide 8 (220 mg, 31%) as a yellow powder: mp
(EtOAc/pet. ether) 191−193 °C; 1H NMR (400 MHz, CDCl3) δ 8.52
(d, 1H, J = 9.5 Hz, H-8), 7.87 (d, 1H, J = 2.7 Hz, H-5), 7.50 (dd, 1H,
J = 9.5, 2.7 Hz, H-7), 4.03 (s, 3H, OCH3), 3.80−3.91 (m, 2H, CH2N),
3.40 (ddd, 1H, J = 13.1, 7.4, 3.4 Hz, CH2N), 3.30 (ddd, 1H, J = 13.1,
6.4, 3.4 Hz, CH2N), 2.63 (dt, 1H, J = 10.9, 5.0 Hz, CH2N), 2.57
(ddd, 1H, J = 10.9, 7.4, 3.4 Hz, CH2N), 2.47 (dt, 1H, J = 11.1,
5.0 Hz, CH2N), 2.32−2.41 (m, 4H, CH2N, NCH3); 13C NMR
(100 MHz, CDCl3) δ 164.6, 156.8, 140.1, 136.6, 133.7, 128.1 (q, J =
34.1 Hz), 125.5, 122.4, 119.3 (q, J = 274.8 Hz), 99.1, 56.9, 54.6, 54.0,
46.1 (2), 42.0; MS m/z 387.2 (MH+, 100%). Anal. Calcd for
C16H17F3N4O4: C, 49.74; H, 4.44; N, 14.50. Found: C, 49.79; H, 4.31;
N, 14.52%. HPLC purity: 95.2%.
1-(1-Methylpiperidin-4-yl)-3-phenylpropane-1,3-dione (27b).
Method B. Reaction of ethyl 3-oxo-3-phenylpropanoate (26b) (2.3 g,
12.0 mmol) and 1-methylpiperazine (3.60 g, 36.0 mmol) gave
1
ketoamide 27b (1.72 g, 58%) as a yellow oil: H NMR (400 MHz,
CDCl3) δ (mixture of keto- and enol amides) 15.26 (s, 0.4H, OH),
8.04 and 7.78 (2 × t, J = 8.3 Hz, 2H, H-2, H-6), 7.60 (br tt, J = 7.4,
1.3 Hz, 1H, H-4), 7.47−7.52 and 7.38−7.44 (2 × m, 2H, H-3, H-5),
5.81 (s, 0.4H, COCH), 4.12 (s, 1.2H, COCH2), 3.68 and 3.53
(2 × t, J = 5.0 Hz, 4H, 2 × CH2N), 2.45 and 2.42 (2 × br t, 4H, J =
5.0 Hz, 2 × CH2N), 2.33 and 2.30 (2 × s, 3H, NCH3); MS m/z 247.2
(MH+, 100%). HRMS calcd for C14H19N2O2 (MH+) m/z 247.1441,
found 247.1445 (−1.6 ppm).
6,7-Dichloro-2-(4-methylpiperazine-1-carbonyl)-3-phenylqui-
noxaline 1,4-Dioxide (11). Method C. Reaction of 1-oxide 24a
(210 mg, 1.03 mmol), ketoamide 27b (278 mg, 1.13 mmol), CaCl2
(11 mg, 0.10 mmol), and ethanolamine (cat., 2 drops) in dry MeOH
(5 mL) gave quinoxaline dioxide 11 (74 mg, 16%) as a yellow powder:
mp (MeOH) 194−196 °C; 1H NMR (400 MHz, CDCl3) δ 8.77
(s, 1H, H-8), 8.73 (s, 1H, H-5), 7.61−7.64 (m, 2H, H-3′, H-5′), 7.52−
7.56 (m, 3H, H-2′, H-4′, H-6′), 3.51−3.63 (m, 2H, CH2N), 3.61 (ddd,
1H, J = 13.2, 7.3, 3.3 Hz, CH2N), 3.52 (ddd, 1H, J = 13.2, 6.4, 3.3 Hz,
CH2N), 3.32 (ddd, 1H, J = 13.0, 7.5, 3.3 Hz, CH2N), 3.08 (ddd, 1H,
J = 13.0, 6.4, 3.2 Hz, CH2N), 2.37−2.48 (m, 2H, CH2N), 2.19 (s, 3H,
NCH3), 1.96−2.03 (m, 1H, CH2N), 1.83−1.89 (m, 1H, CH2N); MS
m/z 434.2 (MH+, 100%). Anal. Calcd for C20H18Cl2N4O3: C, 55.44;
H, 4.19; N, 12.93. Found: C, 55.66; H, 4.20; N, 12.73%. HPLC purity:
97.4%.
5,6-Difluorobenzo[c][1,2,5]oxadiazole 1-Oxide (25h). Method A.
Reaction of NaNO2 (0.59 g, 8.62 mmol) in water (5 mL) and 4,5-
difluoro-2-nitroaniline (24h) (1.00 g, 5.74 mmol) with subsequent re-
action with NaN3 (0.93 g, 14.4 mmol) and NaOAc (9.41 g, 115 mmol)
gave 1-oxide 25h (0.77 g, 75%) as a yellow solid: mp (EtOAc/pet.
ether) 135−138 °C [lit.20 mp 140−147 °C]; H NMR δ (400 MHz,
1
CDCl3) 7.24 (br s, 2H, H-4, H-7); MS m/z 172.1 (M−, 100%).
6,7-Difluoro-2-(4-methylpiperazine-1-carbonyl)-3-phenylqui-
noxaline 1,4-Dioxide (12). Method C. Reaction of 1-oxide 25h
(362 mg, 2.10 mmol), ketoamide 27b (570 mg, 2.30 mmol), CaCl2
(23 mg, 0.21 mmol), and ethanolamine (1 drop) in dry EtOH
(10 mL) gave quinoxaline dioxide 12 (361 mg, 43%) as a yellow
1
powder: mp (MeOH/EtOAc) 169−172 °C; H NMR δ (400 MHz,
CDCl3) 8.41−8.49 (s, 1H, H-5, H-8), 7.61−7.65 (m, 2H, H-2′, H-6′),
7.51−7.68 (m, 3H, H-3′, H-4′, H-5′), 3.50−3.62 (m, 2H, CH2N), 3.32
(ddd, 1H, J = 13.0, 7.6, 3.3 Hz, CH2N), 3.09 (ddd, 1H, J = 13.0, 6.3,
3.3 Hz, CH2N), 2.36−2.47 (m, 2H, CH2N), 2.18 (s, 3H, NCH3),
1.95−2.03 (m, 1H, CH2N), 1.80−1.88 (m, 1H, CH2N); 13C NMR
(100 MHz, CDCl3) δ 157.6, 155.2 (dd, J = 259.1, 12.6 Hz), 153.7 (dd,
J = 261.3, 14.6 Hz), 140.4, 138.1, 135.5 (d, J = 7.5 Hz), 134.6 (d, J =
7.4 Hz), 131.3, 129.8 (2), 128.9 (2), 127.0, 109.2 (dd, J = 24.0, 4.5 Hz,
2), 54.6, 53.9, 46.0, 45.9, 41.6; MS m/z 401.3 (MH+, 100%). Anal.
Calcd for C20H18F2N4O3: C, 60.00; H, 4.53; N, 13.99. Found: C,
59.74; H, 4.64; N, 13.83%. HPLC purity: 97.2%.
5,6-Dimethylbenzo[c][1,2,5]oxadiazole 1-Oxide (25f). Method A.
Reaction of NaNO2 (2.82 g, 40.9 mmol) and 4,5-dimethyl-2-nitro-
aniline (24f) (4.00 g, 24.1 mmol) with subsequent reaction with NaN3
(3.91 g, 60.2 mmol) and NaOAc (39.5 g, 480 mmol) gave 1-oxide 25f
(2.29 g, 58%) as a yellow powder: mp (toluene) 139−140 °C [lit.19
1
mp (EtOH) 139.4−139.7 °C]; H NMR (400 MHz, CDCl3) δ 7.44
(br s, 2H, H-4, H-7), 2.32 (s, 3H, CH3), 2.31 (s, 3H, CH3); MS m/z
165.2 (MH+, 100%).
6,7-Dimethyl-2-(4-methylpiperazine-1-carbonyl)-3-
(trifluoromethyl)quinoxaline 1,4-Dioxide (9). Method C. Reaction
of 5,6-dimethylbenzo[c][1,2,5]oxadiazole 1-oxide (25f) (528 mg,
3.22 mmol), ketoamide 27a (766 mg, 3.22 mmol), CaCl2 (715 mg,
6.44 mmol), and K2CO3 (20 mg) in EtOH (10 mL) gave quinoxaline
dioxide 9 (492 mg, 40%) as a yellow powder: mp (MeOH) 197−
6-Chloro-3-(4-methylpiperazine-1-carbonyl)-2-phenylquinoxa-
line 1,4-Dioxide (13). Method C. A mixture of 5-chlorobenzo[c]-
[1,2,5]oxadiazole 1-oxide (25b) (714 mg, 4.19 mmol), ketoamide 27b
(1.13 g, 4.60 mmol), CaCl2 (0.42 mmol, 47 mg), and ethanolamine
(cat., 1 drop) in dry MeOH (5 mL) was stirred at 20 °C for 60 h. The
solvent was evaporated, and the residue was purified by column
chromatography, eluting with a gradient (0−6%) of (MeOH/DCM),
to give quinoxaline dioxide 13 (238 mg, 19%) as a yellow powder: mp
1
199 °C; H NMR δ 8.35 (s 1H, H-8), 8.33 (s, 1H, H-5), 3.80−3.90
(m, 2H, CH2N), 3.39 (ddd, 1H, J = 13.1, 7.4, 3.4 Hz, CH2N), 3.29
(ddd, 1H, J = 13.1, 6.4, 3.4 Hz, CH2N), 2.63 (dt, 1H, J = 11.1, 5.0 Hz,
CH2N), 2.52−2.57 (m, 7H, 2 × CH3, CH2N), 2.46 (dt, 1H, J = 11.1,
5.2 Hz, CH2N), 2.31−2.40 (m, 4H, CH2N, NCH3); 13C NMR
(100 MHz, CDCl3) δ 157.0, 146.1, 145.0, 136.8, 136.7, 135.6, 129.4
(q, J = 34.5 Hz), 120.1, 120.0, 119.4 (q, J = 275.2 Hz), 54.7, 54.1, 46.2,
46.1, 38.9, 20.8, 20.7; MS m/z 385.2 (MH+, 100%). Anal. Calcd for
C17H19F3N4O3: C, 53.12; H, 4.98; N, 14.58. Found: C, 53.47; H, 5.21;
N, 14.19%. HPLC purity: 95.9%.
1
(MeOH/EtOAc) 205−208 °C; H NMR [400 MHz, (CD3)2SO] δ
8.51 (d, 1H, J = 9.2 Hz, H-8), 8.48 (d, 1H, J = 2.2 Hz, H-5), 8.04 (dd,
1H, J = 9.2, 2.2 Hz, H-7), 7.50−7.60 (m, 5H, H-2′, H-4′, H-3′, H-5′,
H-6′), 3.40−3.50 (m, 2H, CH2N), 3.25−3.30 (m, 2H, CH2N), 3.17−
3.23 (m, 2H, CH2N), 2.30−2.40 (m, 2H, CH2N), 2.18 (s, 3H, NCH3),
1.68−1.80 (m, 2H, CH2N); 13C NMR (100 MHz, CDCl3) δ 158.0,
D
DOI: 10.1021/acs.chemrestox.6b00133
Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology
Article
139.4, 138.5, 137.9, 137.1, 133.5, 131.2, 129.9 (2), 129.0 (2), 127.3,
122.7, 120.1, 54.7, 54.0, 46.1, 46.0, 41.7; MS m/z 399.2 (MH+, 100%).
HRMS Calcd for C20H19ClN4NaO3 (M+Na+) m/z 421.1038, found
421.1022 (3.9 ppm). HPLC purity: 94.7%.
6-Fluorobenzo[c][1,2,5]oxadiazole 1-Oxide (25i). Method A.
Reaction of NaNO2 (3.76 g, 54.5 mmol) and 4-methoxy-2-nitroaniline
(24i) (5.00 g, 32.1 mmol) with subsequent reaction with NaN3
(5.21 g, 80.1 mmol) and NaOAc (52.5 g, 641 mmol) gave 1-oxide
25i (3.44 g, 70%) as an orange solid: mp (EtOAc/pet. ether)
48−49 °C (lit.20 mp 40−42 °C); 1H NMR (400 MHz, CDCl3) δ 7.40
(br s, 1H, H-4), 7.14 (br s, 1H, H-5), 7.07 (br s, 1H, H-7); MS m/z
154.0 (M−, 100%).
J = 9.5 Hz, H-8), 7.92 (d, 1H, J = 2.7 Hz, H-5), 7.61−7.65 (m, 2H,
H-3′, H-5′), 7.50−7.54 (m, 3H, H-2′, H-4′, H-6′), 7.47 (dd, 1H, J =
9.5, 2.7 Hz, H-7), 4.04 (s, 3H, OCH3), 3.52−3.64 (m, 2H, CH2N),
3.35 (ddd, 1H, J = 13.1, 7.7, 3.3 Hz, CH2N), 3.11 (ddd, 1H, J = 13.1,
6.3, 3.2 Hz, CH2N), 2.38−2.49 (m, 2H, CH2N), 2.18 (s, 3H, NCH3),
1.97 (ddd, 1H, J = 11.2, 7.1, 3.3 Hz, CH2N), 1.85 (ddd, 1H, J = 11.2,
7.7, 3.2 Hz, CH2N); 13C NMR (100 MHz, CDCl3) δ 163.0, 158.3,
138.9, 138.0, 137.9, 133.6, 130.8, 130.1 (2), 128.8 (2), 127.6, 124.8,
122.4, 99.0, 56.7, 54.7, 54.0, 46.1 (2), 41.6; MS m/z 395.2 (MH+,
100%). Anal. Calcd for C21H22N4O4: C, 63.95; H, 5.62; N, 14.20.
Found: C, 64.19; H, 5.65; N, 14.33%. HPLC purity: 98.7%.
6,7-Dimethyl-2-(4-methylpiperazine-1-carbonyl)-3-phenylqui-
noxaline 1,4-Dioxide (18). Method C. Reaction of 1-oxide 25f (570 mg,
3.47 mmol), ketoamide 27b (940 mg, 3.82 mmol), CaCl2 (39 mg,
0.35 mmol), and ethanolamine (cat., 2 drops) in dry MeOH (40 mL)
gave quinoxaline dioxide 18 (1.11 g, 81%) as a yellow solid: 1H NMR
(400 MHz, CDCl3) δ 8.39 (s, 1H, H-8), 8.36 (s, 1H, H-5), 7.61−7.66
(m, 2H, H-3′, H-5′), 7.49−7.53 (m, 3H, H-2′, H-4′, H-6′), 3.51−3.63
(m, 2H, CH2N), 3.33 (ddd, 1H, J = 13.1, 7.6, 3.3 Hz, CH2N), 3.10
(ddd, 1H, J = 13.1, 6.3, 3.3 Hz, CH2N), 2.54 (s, 6H, 2 × CH3), 2.36−
2.47 (m, 2H, CH2N), 2.18 (s, 3H, NCH3), 1.94−2.00 (m, 1H, CH2N),
1.82−1.87 (m, 1H, CH2N); 13C NMR (100 MHz, CDCl3) δ 158.4,
145.0, 143.5, 139.0, 136.9, 136.6, 135.7, 130.7, 130.0 (2), 128.7 (2),
127.8, 120.1, 119.7, 54.7, 54.0, 46.0 (2), 41.6, 20.6, 20.5; MS 393.2
(MH+, 100%). HRMS calcd for C22H25N4O3 (MH+) m/z 393.1921,
found 393.1911 (3.8 ppm). HPLC purity: 98.5%.
6-Fluoro-3-(4-methylpiperazine-1-carbonyl)-2-phenylquinoxa-
line 1,4-Dioxide (14). Method C. Reaction of 1-oxide 25i (285 mg,
1.85 mmol), ketoamide 27b (501 mg, 2.04 mmol), CaCl2 (0.19 mmol,
21 mg), and ethanolamine (cat., 2 drops) in dry MeOH (20 mL) gave
quinoxaline dioxide 14 (263 mg, 37%) as a yellow powder: mp
(MeOH) 215−217 °C; 1H NMR (400 MHz, CDCl3) δ 8.70 (dd, 1H,
J = 9.5, 5.0 Hz, H-5), 8.29 (dd, 1H, J = 8.4, 2.7 Hz, H-5), 7.61−7.67
(m, 3H, H-7, H-3′, H-5′), 7.51−7.55 (m, 3H, H-2′, H-4′, H-6′), 3.50−
3.63 (m, 2H, CH2N), 3.33 (ddd, 1H, J = 13.0, 7.6, 3.2 Hz, CH2N),
3.09 (ddd, 1H, J = 13.0, 6.3, 3.2 Hz, CH2N), 2.33−2.49 (m, 2H,
CH2N), 2.19 (s, 3H, NCH3), 1.98 (ddd, 1H, J = 11.0, 7.2, 3.4 Hz,
CH2N), 1.85 (ddd, 1H, J = 11.0, 7.4, 3.2 Hz, CH2N); 13C NMR
(100 MHz, CDCl3) δ 165.8, 163.2, 157.9, 139.4, 138.7 (d, J = 81 Hz),
135.4, 131.1, 130.0 (2), 129.0 (2), 127.3, 124.1 (d, J = 9.2 Hz), 122.3
(d, J = 25.7 Hz), 106.3 (d, J = 28.0 Hz), 54.7, 54.0, 46.1, 46.0, 41.7;
MS m/z 382.2 (MH+, 100%). Anal. Calcd for C20H19FN4O3: C, 62.82;
H, 5.01; N, 14.65. Found: C, 62.90; H, 5.05; N, 14.71%. HPLC purity:
95.0%.
2-[(4-Methyl-1-piperazinyl)carbonyl]-3-phenylquinoxaline 1,4-
Dioxide (15). Method C. Reaction of benzo[c][1,2,5]oxadiazole
1-oxide (25c) (166 mg, 1.22 mmol), ketoamide 27b (300 mg, 1.22
mmol), CaCl2 (271 mg, 2.44 mmol), and K2CO3 (20 mg) in EtOH
(15 mL) gave quinoxaline dioxide 15 (270 mg, 61%) as a yellow
powder: mp (MeOH/H2O) 178−182 °C; 1H NMR (400 MHz,
CDCl3) δ 8.61−8.67 (m, 2H, H-5, H-8), 7.87−7.94 (m, 2H, H-6,
H-7), 7.64−7.68 (m, 2H, H-3′, H-5′), 7.50−7.57 (m, 3H, H-2′, H-4′,
H-6′), 3.53−3.64 (m, 2H, CH2N), 3.35 (ddd, 1H, J = 13.1, 7.7, 3.3 Hz,
CH2N), 3.12 (ddd, 1H, J = 13.1, 6.3, 3.3 Hz, CH2N), 2.38−2.50
(m, 2H, CH2N), 2.18 (s, 3H, NCH3), 1.95−2.05 (m, 1H, CH2N), 1.85
(ddd, J = 11.1, 7.7, 3.3 Hz, 1H, CH2N); 13C NMR (100 MHz, CDCl3)
δ 157.5, 139.1, 137.8, 137.0, 136.9, 132.0, 131.5, 130.4, 129.3 (2),
128.2 (2), 126.9, 120.3, 120.0, 54.0, 53.4, 45.4, 45.3, 41.0; MS 365.2
(MH+, 100%). Anal. Calcd for C20H20N4O3·3/4H2O: C, 63.56; H,
5.73; N, 14.83. Found: C, 63.41; H, 5.51; N, 14.72%. HPLC purity:
99.0%.
6,7-Dihydro-5H-indeno[5,6-c][1,2,5]oxadiazole 1-Oxide (25g).
Method A. Reaction of NaNO2 (0.95 g, 13.7 mmol) and 6-nitro-
2,3-dihydro-1H-inden-5-amine21 (24g) with subsequent reaction with
NaN3 (1.21 g, 20.2 mmol) and NaOAc (13.3 g, 162 mmol) gave
1-oxide 25g (1.24 g, 95%) as a yellow powder: mp (EtOAc/pet. ether)
1
102−104 °C; H NMR (400 MHz, CDCl3) δ 7.28 (br s, 1H, H-8),
7.10 (br s, 1H, H-4), 2.89−2.94 (m, 4H, H-5, H-7), 2.12 (pent, 2H,
J = 7.4 Hz, H-6); 13C NMR (100 MHz, CDCl3) δ 154.1, 152.7, 149.3,
115.2, 110.8, 105.9, 32.6, 26.1 (2). Anal. Calc. for C9H8N2O2: C,
61.36; H, 4.58; N, 15.90. Found: C, 61.53; H, 4.62; N, 15.65%.
2-(4-Methylpiperazine-1-carbonyl)-3-phenyl-7,8-dihydro-6H-
cyclopenta[g]quinoxaline 1,4-Dioxide (19). Method C. A mixture
of 1-oxide 25g (398 mg, 2.26 mmol), ketoamide 27b (612 mg,
2.48 mmol), CaCl2 (25 mg, 0.23 mmol), and ethanolamine (cat.,
3 drops) in dry MeOH (30 mL) gave quinoxaline dioxide 19 (744 mg,
1
81%) as a yellow solid: H NMR (400 MHz, CDCl3) δ 8.46 (s, 1H,
H-9), 8.42 (s, 1H, H-5), 7.60−7.66 (m, 2H, H-3′, H-5′), 7.48−7.54
(m, 3H, H-2′, H-4′, H-6′), 3.52−3.62 (m, 2H, CH2N), 3.33 (ddd, 1H,
J = 13.1, 7.8, 3.3 Hz, CH2N), 3.17 (t, 4H, J = 7.6 Hz, H-6, H-8), 3.10
(ddd, 1H, J = 13.1, 6.2, 3.3 Hz, CH2N), 2.37−2.48 (m, 2H, CH2N),
2.27 (dpent, 2H, J = 7.6, 4.0 Hz, H-7), 2.18 (s, 3H, NCH3), 1.95 (dt,
1H, J = 11.3, 5.6 Hz, CH2N), 1.84 (ddd, 1H, J = 11.3, 7.7, 3.3 Hz,
CH2N); 13C NMR (100 MHz, CDCl3) δ 158.4, 151.6, 151.1, 138.8,
137.7, 136.9, 136.8, 130.7, 130.0, 128.8 (2), 127.8 (2), 115.3, 114.9,
54.7, 54.0, 46.0 (2), 41.6, 33.2, 33.1, 25.8; MS 405.2 (MH+, 100%).
HRMS calcd for C23H25N4NaO3 (M + Na+) m/z 427.1741, found
427.1729 (2.8 ppm). HPLC purity: 99.4%.
6-Methyl-3-(4-methylpiperazine-1-carbonyl)-2-phenylquinoxa-
line 1,4-Dioxide (16). Method C. Reaction of 1-oxide 25d (217 mg,
1.45 mmol), ketoamide 27b (392 mg, 1.59 mmol), CaCl2 (0.14 mmol,
16 mg), and ethanolamine (cat., 1 drop) in dry MeOH (5 mL) gave
quinoxaline dioxide 16 (65 mg, 12%) as a yellow powder: mp
1
(MeOH) 264−267 °C; H NMR (400 MHz, CDCl3) δ 8.54 (d, 1H,
J = 8.8 Hz, H-8), 8.42 (br s, 1H, H-5), 7.71 (dd, 1H, J = 8.8, 1.6 Hz,
H-7), 7.62−7.64 (m, 2H, H-3′, H-5′), 7.50−7.54 (m, 3H, H-2′,
H-4′, H-6′), 3.50−3.50 (m, 2H, CH2N), 3.34 (ddd, 1H, J = 13.1, 7.6,
3.3 Hz, CH2N), 3.11 (ddd, 1H, J = 13.1, 6.3, 3.2 Hz, CH2N), 2.64
(s, 3H, CH3), 2.37−2.50 (m, 2H, CH2N), 2.20 (s, 3H, NCH3), 1.98
(ddd, 1H, J = 11.6, 7.1, 3.6 Hz, CH2N), 1.86 (ddd, 1H, J = 11.3, 7.6,
3.1 Hz, CH2N); 13C NMR (100 MHz, CDCl3) δ 158.3, 143.8, 139.0,
137.7, 137.3, 136.7, 134.5, 130.9, 130.0 (2), 128.8 (2), 127.6, 120.7,
119.5, 54.7, 54.0, 46.0 (2), 41.6, 22.0; MS m/z 379.2 (MH+, 100%).
Anal. Calcd for C21H22N4O3: C, 66.65; H, 5.86; N, 14.81. Found: C,
66.71; H, 5.84; N, 14.93%. HPLC purity: 98.8%.
1-(4-Methylpiperazin-1-yl)butane-1,3-dione (27c). Method B.
Reaction of ethyl 3-oxobutanoate (26c) (2.32 g, 20.0 mmol) and
1-methylpiperazine (2.00 g, 20.0 mmol) in dry toluene (5 mL) gave
the ketoamide22 27c (0.74 g, 20%) as a yellow oil: 1H NMR
(400 MHz, CDCl3) δ (mixture of keto- and enol amides) 14.71
(s, 0.2H, OH), 5.13 (s, 0.2H, CH), 3.66 (br t, 2H, J = 5.1 Hz,
CH2N), 3.55 (s, 1.6H, CH2), 3.44 (br t, 2H, J = 5.1 Hz, CH2N), 2.38
(br t, 4H, J = 5.1 Hz, 2 × CH2N), 2.31 (s, 3H, NCH3), 2.28 (s, 2.4H,
CH3), 1.96 (s, 0.6H, CH3); MS m/z 185.1 (MH+, 100%).
6,7-Dichloro-2-(4-methylpiperazine-1-carbonyl)-3-methylqui-
noxaline 1,4-Dioxide (20). Method C. Reaction of 1-oxide 25a
(343 mg, 1.67 mmol), ketoamide 27c (308 mg, 1.67 mmol), CaCl2
(370 mg, 3.34 mmol), and K2CO3 (20 mg) in EtOH (10 mL) gave
quinoxaline dioxide 20 (301 mg, 49%) as a yellow powder: mp
6-Methoxy-3-(4-methylpiperazine-1-carbonyl)-2-phenylquinoxa-
line 1,4-Dioxide (17). Method C. Reaction of 1-oxide 25e (388 mg,
2.33 mmol), ketoamide 27b (632 mg, 2.56 mmol), CaCl2 (0.23 mmol,
25 mg), and ethanolamine (cat., 2 drops) in dry MeOH (40 mL) gave
quinoxaline dioxide 17 (409 mg, 45%) as a yellow powder: mp
1
(MeOH/EtOAc) 190−191 °C; H NMR (400 MHz, CDCl3) δ 8.73
(s, 1H, H-8), 8.67 (s, 1H, H-5), 3.85 (ddd, 1H, J = 12.9, 6.8, 3.4 Hz,
CH2N), 3.37 (ddd, 1H, J = 13.1, 7.0, 3.3 Hz, CH2N), 3.25 (ddd, 1H,
1
(MeOH) 234−237 °C; H NMR (400 MHz, CDCl3) δ 8.55 (d, 1H,
E
DOI: 10.1021/acs.chemrestox.6b00133
Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology
Article
J = 13.1, 6.8, 3.3 Hz, CH2N), 2.55−2.68 (m, 5H, CH2N, NCH3), 2.48
(ddd, 1H, J = 11.1, 7.0, 3.5 Hz, CH2N), 2.28−2.36 (m, 4H, CH2,
3-CH3); 13C NMR (100 MHz, CDCl3) δ 158.0, 140.8, 138.4, 137.9,
137.6, 136.3, 135.5, 121.9, 121.8, 55.3, 54.4, 46.4, 46.1, 42.1, 14.8; MS
m/z 371.1 (MH+, 100%). The compound was formulated as the
hydrochloride salt: mp 176−180 °C (dec.). Anal. Calcd for
C15H17Cl3N4O3·11/4H2O: C, 41.88; H, 4.20; N, 13.74. Found: C,
41.90; H, 4.65; N, 12.89%. HPLC purity: 97.1%.
previously described26 in N2O-saturated solutions of sodium formate
(0.1 M) at pH 7 (5 mM phosphate). G-loss values in μM·Gy−1
•−
(μmol J−1) were calculated using a G value for CO2 production
under these conditions of 0.68 μM·Gy−1.27
Methodology for EPR Experiments. EPR experiments were
carried out within a TE011 cavity on a JEOL (JES-FA-200) EPR spec-
trometer, equipped with a variable temperature controller (ES-DVT4)
and operated at a 9.1 GHz and 100 kHz field modulation. All solutions
were prepared in Milli-Q water pretreated with Chelex-100 resin to
remove trace amounts of polyvalent metal ions, and diethylene-
triaminepentaacetic acid (dtpa) was added to the solutions to sup-
press any Fenton-type chemistry. Aqueous solutions of the QDO
compounds (in phosphate buffer, 50 mM at pH 7.0), NADPH, and
the spin traps (PBN or DEPMPO), were predegassed separately with
N2 gas. Highly purified human NADPH/cytochrome P450 oxidor-
eductase (POR) was prepared as previously described.37 Superoxide
dismutase (SOD) from bovine erythrocytes, glucose-6-phosphate
(G-6-P), glucose-6-phosphate dehydrogenase (G-6-P-D), POR, and
the spin traps were added into the sample vial under N2 atmosphere.
All additions were carried out on ice and the enzyme reaction started
by the addition of NADPH. An EPR flat quartz cell, designed for the
variable temperature Dewar, was used for all measurements. Solutions
for EPR measurements were transferred under N2 from the sample vial
to the EPR cell, which had been flushed previously with N2 and
inserted into the EPR cavity. Extreme care was taken to keep the
system deaerated at all times. The temperature of the cavity was raised
to 37 °C to activate the enzyme incubation. EPR spectra, recorded at a
power of 20 mW, were averaged (50−100 scans) to improve the
signal-to-noise ratio over a scan range of 200 G (Guass) at a modulation
width of 0.5 G, scan time of 2 min, and time constant of 0.03 s.
Computer simulation of spectra were carried out using the WINSIM
EPR program available in the public domain of the NIEHS EPR
database. The correlation coefficient, r, for all the spectral simulations
was >0.9.
2-(4-Methylpiperazine-1-carbonyl)-3-methylquinoxaline 1,4-Di-
oxide (21). Method C. Reaction of benzo[c][1,2,5]oxadiazole
1-oxide (25c) (542 mg, 3.98 mmol), ketoamide 27c (733 mg,
3.98 mmol), CaCl2 (883 mg, 7.96 mmol), and K2CO3 (20 mg) in
EtOH (10 mL) gave quinoxaline dioxide 21 (725 mg, 60%) as a yellow
1
powder: mp (MeOH/EtOAc) 195−197 °C; H NMR (400 MHz,
CDCl3) δ 8.64 (dd, 1H, J = 8.3, 1.5 Hz, H-8), 8.57 (dd, 1H, J = 8.2,
1.6 Hz, H-5), 7.82−7.91 (m, 2H, H-6, H-7), 3.84−3.95 (m, 2H,
CH2N), 3.40 (ddd, 1H, J = 13.1, 7.3, 3.3 Hz, CH2N), 3.26 (ddd, 1H,
J = 13.1, 6.6, 3.3 Hz, CH2N), 2.57−2.69 (m, 5H, CH2N, NCH3),
2.44−2.50 (m, 1H, CH2N), 2.27−2.36 (m, 4H, CH2N, NCH3);
13C NMR (100 MHz, CDCl3) δ 158.6, 140.0, 137.9, 137.3, 137.0,
132.5, 131.6, 120.5, 120.3, 55.4, 54.5, 46.3, 46.2, 38.9, 14.8; MS m/z
365.2 (MH+, 100%). Anal. Calcd for C15H18N4O3: C, 59.59; H, 6.00;
N, 18.53. Found: C, 59.49; H, 5.84; N, 18.35%. HPLC purity: 97.7%.
Radiation Chemistry. Time-resolved optical absorption and
kinetic studies were carried out using the University of Auckland’s 4
MeV linear accelerator which delivers 200 ns electron pulses of
typically 3 Gy dose. The optical detection system and method of
dosimetry have been described.23 The radiolysis of water produces
three well-characterized reactive radical species used to initiate radical
reactions, as well as molecular products (G values in micromolar pro-
duced per absorbed dose of 1 Gy (J·kg−1) are given in parentheses).
H2O ⇝ eaq−(0.28) + HO·(0.28) + ·H(0.055) + H2(0.04)
+ H2O2(0.07) + H3O+(0.28)
Density Functional Theory Calculations. The geometry
optimization and energy calculations were performed with the
Gaussian 09 software suite using unrestricted DFT.28 The nonlocal
B3LYP functional hybrid method was employed29−31 and the standard
6-31+G(d, p)32,33 basis set was used for the geometry optimization
and frequency analysis. The zero-point vibrational energies (ZPE)
were scaled according to Wong (0.9804).34 In all cases, the normal
modes revealed no imaginary frequencies indicating that they
represent minima on the potential energy surface. The subsequent
energy calculations were performed with the larger 6-311+G(2df, p)
basis set. All of the structural optimization and energy calculations
were carried out with a water simulation based on the polarized
continuum model (IEFPCM).35 The piperazine moiety is protonated
on the distal nitrogen. The electron affinities and proton affinities
were calculated as described in Forseman and Frisch.36 In general,
the reaction steps were calculated by subtracting the ZPE corrected
energies of the reactants from the products. The results for the re-
actions designated in parentheses in the schemes are given in Table 3
and Table S1.
One electron reductions of the QDO analogues were carried out by
(i) the eaq−, while at the same time scavenging the oxidizing radicals
with 2-methylpropan-2-ol to form an inert radical, and (ii) electron
24
•−
transfer from the CO2 species (E(CO2/CO2•−) = −1.90 V ) in
N2O-saturated solutions (to quantitatively convert the e− to HO•)
aq
containing 0.1 M sodium formate, which in turn is oxidized by HO•
•
•−
and H atoms to give the reducing CO2 radical. Care was taken to
avoid UV photolysis of the test solutions by deploying optical filters
with a cutoff at or above 400 nm.
−
•−
eaq + A → A (AH·)
HO·(·H) + (CH3)3COH → ·CH2(CH3)2COH + H2O(H2)
−
N2O + eaq → HO· + OH− + N2
•−
HO·(·H) + HCOO− → H2O(H2) + CO2
•−
•−
CO2 + A → A + CO2
In Vitro Cytotoxicity. Human HT29 colorectal carcinoma cells
and SiHa cervical carcinoma cells were authenticated in-house by short
tandem repeat profiling. Cells were grown in α-MEM medium with
5% fetal bovine serum, without antibiotics, and were confirmed
mycoplasma-free using a PCR-ELISA method (Roche Diagnostics).
A SiHa cell line stably expressing POR (SiHa-POR)38 was maintained
in 1 μM puromycin. Log phase cells were seeded into 96 well plates at
1100 cells/well for HT29 and at 1500 cells/well for SiHa and SiHa-
POR and allowed to attach for 2 h, then test compounds were added
by dilution from DMSO stocks to give a top concentration of <1%
DMSO before serial 3-fold dilution in the plates. After 4 h, cultures
were washed 3 times with fresh medium and grown for a further 5 days
before staining with sulforhodamine B to determine IC50 values
(concentration to inhibit 50% growth) as previously described.39 For
anoxic exposure to compounds, cells were pelleted by centrifugation,
transferred to a Pd-catalyst anaerobic chamber (Bactron-II, Shell Lab),
resuspended in anoxic medium, and exposed to drugs as above but
Radical spectra are presented as the change in absorbance per Gy
dose (Abs·Gy−1) or change in extinction coefficient (Δε) using the
radiolytic yield of the reducing species per Gy. The one-electron
reduction potentials of the compounds, E(A/A•−), vs NHE were
determined at pH 7.0 (2.5 mM phosphate buffer) by establishing
redox equilibria within 20 μs between three mixtures of the one-
electron reduced compounds and the reference compounds methy-
lviologen (E(MV2+/MV•+) = −447
7 mV) or triquat (E(TQ2+/
TQ•+) = −548 7 mV) and calculating ΔE values from the equili-
brium constants, Ke, using the Nernst equation, as described in the
literature.25
Steady-state radiolysis experiments to determine G-loss values were
performed using a 60Co γ-source delivering a dose rate of ca. 5 Gy
min−1. The loss of the QDO compounds upon reduction by the
•−
CO2 radical species was followed by monitoring the sequential
changes in their UV−vis absorption spectra with radiation dose as
F
DOI: 10.1021/acs.chemrestox.6b00133
Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology
Article
BTO compounds.26 A secondary intermediate absorbing
maximally in the 400−430 nm region was formed at ca. twice
the rate of decay in absorption observed in the 500−520 nm
region followed by a loss in absorption at all accessible wave-
lengths, i.e., above the deployed 400 nm cutoff filter used to
avoid rapid photolysis of the compounds at shorter wave-
lengths. Both kinetic phases were found to be independent of
the initiating radiation dose (3−12 Gy), i.e., independent of the
initial radical concentration, confirming first-order kinetics. For
all compounds, the rate of decay of the initial 500−520 nm
absorption and the rate of buildup of the 400−430 nm absorp-
tion were independent of the initial concentration of com-
pounds in solution. An example of kinetic data for the 3-Ph-
substituted compound 19 at pH 7 are presented in Figure 2,
using medium and plates that had been equilibrated in the chamber
for at least 3 days. After drug washout, cells were grown and stained
as for the aerobic IC50 assays. The absorbance of aerobic and hypoxic
controls (8 wells/plate) had a mean of 0.19 (CV 14.7%) and 0.13
(CV 18.7%) for aerobic and hypoxic exposures, respectively, across all
assays.
RESULTS
■
Pulse Radiolysis Studies. Pulse radiolysis was used to
rapidly reduce QDO compounds to their radical anions, and sub-
sequent reactions were followed using time-resolved UV−vis
spectrophotometry. The one-electron reduction potentials,
Eo′, of the compounds at pH 7 were measured against appro-
priate redox indicators, and the values are presented in Table 1.
The central unsubstituted quinoxaline 1,4-dioxide is less
electron affinic than benzotriazine 1,4-dioxide with Eo′ values
of −575 mV40 compared to −318 mV.41 The piperazine
carboxamide side chain gave adequate aqueous solubility for the
analogues to be used in in vitro experiments. The strongly
electron withdrawing 3-CF3 group provided increased electron
affinity (Eo′of compound 6 = −356 mV) compared to TPZ
(Eo′= −456 mV), while the nearly electronically neutral 3-Ph
(e.g., 15, Eo′= −543 mV) and electron donating 3-CH3
substituents (e.g., 21 Eo′= −579 mV) reduced the electron
affinity. The electron affinity was further modulated by electron
withdrawing and donating groups at the 6- and 7-positions
providing a range in Eo′ values of ca. 200 mV for 3-CF3 (4−10)
and 3-Ph (11−19) compound sets. Dichloro-6,7-substitution of
the 3-CH3 compound 21 raised the electron affinity into the
region for bioreduction by POR. These Eo′ values span a range
over which the rate of enzymatic one-electron reduction is
expected to vary widely. One-electron reduction of all
Figure 2. Dependence of the rate of decay of the radical anion of 19
on substrate concentration in N2O-saturated solutions containing
sodium formate (0.1 M) at pH 7 (5 mM phosphate) as monitored
○
by the loss in absorption at 520 nm ( ) and gain in absorption at
compounds by either scavenging the e− or electron transfer
aq
●
420 nm ( ).
•−
from the CO2 species produced intermediates with broad
absorption spectra centered at ca. 500−520 nm, as illustrated
where the first-order buildup of the 400−420 nm absorption
occurs with k = 2.06 0.13 × 104 s−1 in contrast to the first-
order decay in the absorption at 520 nm of k = 1.57 0.04 ×
104 s−1. (See Supporting Information for exemplary data for
3-CF3 and 3-CH3-substituted compounds 10 and 23.) Proto-
tropic properties of the initially formed radical anions were
determined as pKr values from the absorption changes in the
500−520 nm region plotted against pH (Table 1). Represen-
tative absorption data for compound 18, obtained on formation
of the initial radical anion at 520 and 420 nm as well as fol-
lowing the buildup of the absorption at 420 nm, are presented
in Figure 3. The rate of decay of the radical anions of all
compounds absorbing in the 500−520 nm region at different
pH values extrapolated to zero substrate concentration (to
compensate for any radical−radical decay) were found to track
rates given by the expression (kacid × [H+])/([H+] + Kr), as
previously found for the BTO compounds,26 indicating that the
observed rate constants arise from a decay of the protonated
radical anions (data not shown). The observed wavelength-
dependent kinetics are interpreted as two different radicals
being formed upon one-electron reduction, one of which
decays to a species which absorbs minimally in the visible
region. The first order buildup of the 400−420 nm absorption
following one-electron reduction of the 3-CH3 substituted
compound 23 occurred with k = 5.8 0.7 × 103 s−1, while
the first-order decay in the absorption at 520 nm was with
for the 3-Ph substituted compound 18 (Figure 1). This initial
Figure 1. Time-resolved difference spectra between the radical
intermediates and parent compound 18 following one-electron
•−
reduction by the CO2 radical produced on pulse radiolysis (3 Gy
in 200 ns). The solution was saturated with N2O gas and contained 18
(0.1 mM), sodium formate (0.1 M), and phosphate buffer (5 mM)
adjusted to pH 7. The changes in extinction coefficient, measured at
□
○
●
10 μs ( ), 0.1 ms ( ), and 5 ms ( ) after the pulse, were calculated
•−
using a radical yield of 0.68 μM·Gy−1 for the CO2 radical.
absorption band decreased in intensity with increasing acidity,
while the first-order rate of decay of the intermediate (k)
increased with acidity, asymptotically approaching a plateau
value at low pH (ca. pH 3−4), (kacid), as seen previously for
k = 3.09
0.15 × 103 s−1. The built-up species absorbed
minimally above 440 nm in contrast to the transients observed
for the 3-CF3 and 3-Ph analogues.
G
DOI: 10.1021/acs.chemrestox.6b00133
Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology
Article
Figure 4. Dependence of the rate constant for the oxidation of the
radical anions of the QDO compounds by molecular oxygen on the Eo′
values of the compounds.
Figure 3. Changes in absorbance with pH following the pulse
radiolysis of the solution of 18 described in Figure 1 at different pH
■
values measured (i) 10 μs after the pulse at 430 nm ( ) and 520 nm
□
●
( ) and (ii) 40−400 μs after the pulse at 430 nm ( ).
The radical anion of a 3-CH3 substituted BTO compound
has been shown by EPR to undergo transformation to form
predominantly the 3-H2C• radical.41 This may also be the case
for the 3-CH3 substituted QDO compounds which give rise to
an absorption spectrum that differs from those arising from
3-Ph and 3-CF3 QDO analogues. The decay rate constants
obtained for all radical anions from observations at 520 nm,
both at pH 7 and at low pH (kacid), as well as their pKr values,
are presented in Table 1. Both the kinetic parameters and pKr
values increase with decreasing Eo′ values for the compounds,
as has been reported for a series of BTO compounds.37 The
second-order rate constants for the oxidation of the radical
anions of the compounds by molecular oxygen, kO2, obtained
from plots of the measured rates at 520 nm against increasing
O2 concentration (see Supporting Information for exemplary
data), are presented in Table 1. The determined kO2 values
increase with decreasing Eo′ values, as expected for an electron
transfer process. Plotting log kO2 vs Eo′ (Figure 4) gives the
expression:
Figure 5. Time-resolved difference spectra between the radical inter-
mediates and parent compound 19 following one-electron reduction
by the CO2 radical produced on pulse radiolysis (3 Gy in 200 ns).
The solution was saturated with N2O gas and contained 19 (0.5 mM),
sodium formate (0.1 M), and phosphate buffer (5 mM; adjusted to
•−
○
●
pH 4). The changes in absorbance, measured at 1 μs ( ), 10 μs ( ),
□
and 2 ms after the pulse ( ), were normalized to the radiation dose
(Gy). Inserts A and B: changes in transmittance at 420 nm for both
short and long time bases.
log kO2 = (4.91 0.14) − (4.54 0.28)Eo′/V
(n = 20, r = 0.956)
which closely matches that obtained for BTO compounds
where26
log kO2 = (4.57 0.09) − (4.99 0.21)Eo′/V
Further information on the properties of the radicals formed
from the radical anions of the QDO compounds was obtained
in solutions at low pH in the absence and presence of oxygen
for compound 19. The loss of the initial radical absorption at
500 nm and the formation of the transient absorbing maximally
in the 400−420 nm region are completed on the 10 μs time
scale under highly acidic conditions (Figure 5). This means that
kacid is greater than the rate of back oxidation of the radical
anion and that the reaction between the species responsible for
the transient at 400−420 nm and oxygen can be studied. The
radical was found to react with oxygen, forming a new transient
adduct, presumably a peroxyl radical (Figure 6). Such a peroxyl
radical is likely to be an aryl-peroxyl which absorbs in the visible
range,42 whereas alkyl-peroxyl radicals do not. The rate con-
stant for oxygen reacting with the transient at 400−420 nm was
determined as 3.0 0.1 × 108 M−1 s−1 (Figure 7).
Figure 6. Time-resolved difference spectra between the radical inter-
mediates and parent compound 19 following pulse radiolysis (3 Gy in
200 ns) of the solution as described in Figure 1 saturated with a gas
mixture of 4:1 N2O/air. The changes in absorbance, measured
at 1 μs ( ), 10 μs ( ), 100 μs ( ), and 2 ms ( ), after the pulse,
were normalized to the radiation dose (Gy). Insert: Observed biphasic
change in transmittance at 420 nm fitted to two consecutive first-order
reactions (red line).
■
□
○
●
H
DOI: 10.1021/acs.chemrestox.6b00133
Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology
Article
HO• (half-life ca. 1 min at pH 744), which is rarely detected in
biological systems. The PBN−OH adduct, which has hyperfine
coupling constants (hfc) of aN 15.6 G and aH 2.8 G,45 was
detected by reduction of 5. This unusual observation most
likely arises from the high electron affinity of 5 (Eo′ value of
−306 mV), driving its reduction to produce a relatively high
steady-state concentration of the HO•. Production of the HO•
has been previously confirmed upon one-electron reduction of
compounds 5 and 10 in the presence of DEPMPO, which
produces a more stable and hence measurable HO•-trapped
species. The previous study indicated that release of the HO•
radical requires a strongly electron-withdrawing substituent,
such as CF3, at the 3-position on the QDO scaffold.17 A com-
plex, multispecies spectrum of PBN-trapped radicals sub-
sequently built up over 2 h, most likely due reaction of the
relatively high flux of HO• with a range of components in the
reducing system. Two other 3-CF3-substituted compounds
investigated, 9 and 10, gave single PBN-trapped C-centered
radical spectra with hfc of aN 16.2 G and aH 3.5 G. The
simulated spectrum for 8 consisted of two species with hfc’s of
(i) aN 16.1 G, aH 3.77 G, and aHγ 1.10 G and (ii) aN 16.4 G
and aH 3.9 G (Figure 8). The hfc of species i are the same as
Figure 7. Dependence of the 1st component of the fitted transients in
Figure 6 on the concentration of O2, yielding a 2nd-order rate constant
of k = 3.0 0.1 × 108 M−1 s−1.
Steady-State Radiolysis Studies. The concentrations of
•−
the QDO compounds consumed upon reduction by the CO2
radical species per Gy (G-loss) are presented in Table 1. All of
the values obtained exceed the G value (radiolytic yield) of the
CO2•− radical species (0.68 μM·Gy−1), with compounds 4−10
exhibiting large G-loss values compared with those of the other
compounds. The 3-CF3 substitution is associated with the HO•
being released from the QDO compounds upon their one-
electron reduction,17 which would be scavenged by the high
concentration of formate ions in solution producing further
•−
CO2 radical species in a chain reaction. The fact that G-loss
values are not even higher for compounds 4−10 suggests that
bifurcation to other radicals on the QDO scaffold occurs along
with the HO• being released. The G-loss values measured for
all of the compounds increased with increasing concentrations
of formate ions (data not shown), also indicating that multiple
radicals are produced and competitive decay pathways are
present. G-loss values are not affected by second-order,
radical−radical reactions under steady-state radiolysis condi-
tions (∼0.08 Gy s−1); however, they were decreased when
samples are irradiated by a train of short pulses of radiation of
1 Gy in 200 ns (∼5 × 106 Gy s−1), i.e., to produce a high radical
concentration instantaneously (data not shown). As the initial
decay kinetics are found to be independent of radiation dose
under steady-state radiolysis conditions, the observed effects on
G-loss upon pulse irradiation must arise from secondary
radical−radical reactions competing with chain propagation.
The relatively smaller G-loss values for the 3-Ph-and 3-CH3-
substituted analogues, which do not eliminate the HO• radical,
could possibly arise from aryl radical undergoing H atom
abstraction from formate ions or radical−substrate reactions
forming transient adducts which break down into products.
EPR Experiments. The one-electron reduction of the QDO
compounds by POR was carried out anaerobically at 37 °C,
in situ in an EPR spectrometer at pH 7.0 in the presence of the
nitrone spin traps PBN or DEPMPO to identify radicals formed
upon the reduction of the QDO compounds. (Nitroso spin
traps, such as 2-methyl-2-nitrosopropane (MNP), were found
to be unstable in the presence of POR and could not be used.
The commonly used DMPO is easily oxidized giving false
conclusions43.) A representative selection of the compounds
having 3-CF3 (5, 9, and 10), 3-phenyl (11−13), or 3-CH3 (20)
substituents, spanning the Eo′ range of values between
−500 mV and −300 mV, were chosen for reduction by POR.
Acyclic PBN is known to form an unstable adduct with the
Figure 8. EPR spectra obtained by reduction of 8 by sPOR
protein (14 ng mL−1) at 37 °C containing dtpa (100 μM), SOD
(300 units mL−1), catalase (1500 units mL−1), glucose-6-phosphate-
dehydrogenase (13 units mL−1), glucose-6-phosphate (10 mM), and
NADPH (1 mM) at pH 7 in the presence of (a) PBN (50 mM) and
(b) PBN (50 mM) and DMSO (2 M); (c) simulated spectrum a
corresponding to a general C-centered radical and the •CH2O- radical
in a ratio of 0.43:0.57, r = 0.926. (d) Control, no 8 added.
that for the oxymethyl radical (•CH2O−) substituent trapped
on the BTO scaffold,41 while the large aH value of species
ii is close to the value reported for a trapped phenyl radical
(aH 3.9 G c.f. 4.0−4.2 G);46,47 however, the hfc’s of a trapped
quinoxalinyl-type radical for QDO compounds have not been
reported and may well be different from that of the aryl radical
on BTO scaffold. The three 3-Ph substituted compounds,
11−13, and the 3-CH3 substituted compound, 20, did not
form detectable, spin-trapped species with PBN after reduc-
tion. Analogous experiments were carried out with a high
concentration of added DMSO (2 M) as a HO• scavenger to
intercept any possible released HO• before it could react with
the QDO compounds or POR or other constituents of the
biochemical system used. The HO• is known to add to DMSO
to release the methyl radical, which is subsequently trapped as
the PBN−CH3 species with well-defined hfc’s.48,49 In contrast
I
DOI: 10.1021/acs.chemrestox.6b00133
Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology
Article
to the previous observation14 of DEPMPO−CH3 species
derived from 3-CF3 analogues 5 and 10, the PBN−CH3 radical
species was not observed for 8 in the presence of DMSO, but
a C-centered radical with hfc of aN 16.5 G and aH 3.64 G
was observed (Figure 8b). Such a species is thought to arise
from the spin trapping of the radical formed upon H atom
abstraction from a methyl group on DMSO.50 The same
C-centered radical is trapped by PBN for the 3-Ph and 3-CH3
QDO in the presence of added DMSO (Table 2). Reduction of
DMSO (2 M) was present throughout the incubation. Together,
these observations imply that the observed C-centered radicals
arise from reaction of the HO• and that the high concentration
of DMSO scavenges all of the released HO•.
Although two other 3-CF3-substituted compounds inves-
tigated by EPR, 8 and 9, did not give rise to a DEPMPO−OH
spectrum (presumably because any released HO• reacted faster
to give the observed C-centered radicals), the DEPMPO−CH3
radical spectrum was again exclusively obtained when DMSO
was present (Table 2). Of the three 3-Ph-substituted com-
pounds examined, 11 and 13 did not form detectable radical
adducts with DEPMPO after bioreduction, while compound 12
gave a general C-centered radical, which was unchanged in the
presence of added DMSO. In contrast to the combination of
PBN and DMSO trapping, a different radical species to PBN
alone, it is likely that the DEPMPO trapped C-centered radical
is on the phenyl moiety of 3-Ph compounds and does not
possess the oxidizing power to abstract an H atom from the
methyl group of DMSO. These observations give support to
the formation of two types of C-centered radicals for the 3-Ph
substituted compounds. The 3-CH3-substituted compound, 20,
exhibited a 1:1 mixture of radicals as seen above for compounds
5 and 10 (Figure 9). The commonality in the formation of the
Table 2. Hyperfine Coupling Constants in Gauss for Spin-
Trapped Radical Species
compd
PBN
PBN + DMSO
DEPMPO
DEPMPO + DMSO
b
a
5
aN15.6
aN 16.5
aH 3.67
aN 14.0 aN 14.6
aH 13.2 aH 21.7
aP 47.3 aP 47.0
30.7% 69.3%
aN 14.9 aN 14.5
aH 18.7 aH 20.8
aP 51.5 aP 49.9
6.1% 26.5%
aN 14.7
aN 15.2
aH 22.0
aP 48.2
aH 2.8
8
aN 16.4
aH 3.9 43%
aN 16.1
aH 3.8
aN 16.5
aH 3.64
aN 15.3
aH 22.2
aP 48.3
aHγ1.1
47%
aH 21.3
aP 47.6
67.4%
9
aN 16.3
aH 3.5
aN 16.5
aH 3.67
aN 15.0
aN 15.2
aH 22.2
aP 48.4
aN 15.2
aH 21.9
aP 47.7
aH 22.0
aP 48.7
b
10
aN 16.2
aH 3.45
aN 16.4
aH 3.67
aN 14.0 aN 14.7
aH 13.2 aH 21.3
aP 47.3 aP 47.6
16% 84%
11
12
c
c
aN 16.4
aH 3.67
aN 16.4
aH 3.67
c
aN 14.8
aH 21.5
aP 48.5
c
aN 14.9
aH 21.5
aP 48.4
13
20
c
c
aN 16.5
aH 3.67
aN 16.4
aH 3.67
Figure 9. EPR spectra obtained by reduction of 20 by sPOR pro-
tein (14 ng mL−1) at 37 °C containing dtpa (100 μM), SOD
(300 units mL−1), catalase (1500 units mL−1), glucose-6-phosphate-
dehydrogenase (13 units mL−1), glucose-6-phosphate (10 mM), and
NADPH (1 mM) at pH 7 in the presence of (a) DEPMPO (25 mM)
and (b) DEPMPO (25 mM) and DMSO (2 M DMSO). (c) Simulated
aN 14.9 aN 14.7
aH 21.5 aH 21.3
aP 46.2 aP 47.6
aHγ0.1
aN 14.4
aH 21.5
aP 47.1
50% 50%
a
b
c
•
Initial trapped radical. Ref 17. No EPR signal.
spectrum for the general C-centered and CH2 radical at a ratio of
0.50:0.50, r = 0.936. (d) Control, no 20 added.
these QDO produces radicals which most likely form unstable
adducts with PBN but which are also capable of H atom
abstraction reaction from a methyl group on DMSO. This
observation implies that the carbon−hydrogen bond strength of
such an oxidizing C-centered radical is greater than that of the
carbon−hydrogen bond of a methyl moiety. Cyclohexadienyl
radicals (possibly formed upon HO• addition to a benzenoid
moiety) are reducing radicals51 and cannot abstract an H atom
from a methyl group, implying that a strong sigma radical
species must be formed following the reduction of 9 and 10,
possibly a quinoxalinyl-type radical as can be postulated for 8.
The earlier study identified that following the release of HO•
from 5 and 10 upon bioreduction, similar C-centered radicals
were formed over a 2 h time period.17 Both the DEPMPO−OH
adduct and the C-centered, DEPMPO trapped radical spectra
were replaced by the DEPMPO−CH3 radical spectrum when
same radical species for compounds 5, 10, and 20 implies that
the C-centered radical resides on the benzene moiety of the
quinoxaline scaffold. The generalized scheme for the radical
pathways following one-electron reduction of the QDO com-
pounds (reaction 1a) is presented in Scheme 3, the pathways of
which are examined by DFT calculations in the next section.
DFT Calculations. Previous calculations for the radical
anions of 5 and 10 (3-CF3-substituted analogues) showed, within
error, equivalent changes in energy for protonation on N1−O
and N4−O through strongly exothermic reactions, with signi-
ficant spin located on C2 and C3, respectively.17 This is also
the case for the new 3-CF3 compound (4), the 3-Ph (11, 12,
and 19) and 3-CH3 (20) substituted compounds (Table 3).
The unimolecular elimination of the HO• from N4−OH (pathway
3a) is a more exothermic process than that from N1−OH anions
J
DOI: 10.1021/acs.chemrestox.6b00133
Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology
Article
Cytotoxicity Studies. The cytotoxic potencies of the QDO
compounds were evaluated under aerobic and anoxic condi-
tions, and the data were compared with that of TPZ and
SN30000. Table 4 gives IC50 values, for the full series of
compounds, for aerobic exposure of HT29 and SiHa cells and
for a cell line stably overexpressing POR (SiHa-POR), which is
known to be hypersensitive to BTOs.52 These determinations
used dilutions from DMSO stock solutions giving final DMSO
concentrations in culture of up to 1%, but DMSO concentra-
tions at the IC50 values of the test compounds were <0.5% in all
cases. Given the potential for DMSO to scavenge oxidizing
radicals, we confirmed that 0.5% DMSO did not protect against
aerobic cytotoxicity of SN30000 or three representative QDO
compounds (6,8,12) in each of the three cell lines (see
Supporting Information, Figure S4).
Scheme 3. Reaction Pathways of One-Electron Reduced
QDO Compounds
The aerobic cytotoxicities of the 3-CF3 QDOs (4−10) were
considerably greater than those of the reference BTOs TPZ and
SN30000 in both HT29 and SiHa cells. As expected, there was
a clear increase in potency with increasing Eo′ in both cell lines,
but QDO compounds with Eo′ values in the range −500 mV to
−400 mV (compounds 7−14) were much more cytotoxic than
TPZ (−456 mV) or SN30000 (−401 mV). The 3-Ph and
3-CH3 compounds were less cytotoxic than the corresponding
3-CF3 compounds, reflecting their lower Eo′ values, and again
showed a decrease in cytotoxic potency with decreasing Eo′.
Overexpression of POR markedly increased the aerobic cyto-
toxicity of the QDOs, with decreases in IC50 ranging from 3- to
17-fold for the 3-CF3 compounds confirming that aerobic
cytotoxicity is dependent on bioreduction. Notably, the QDOs
displayed little or no hypoxic selectivity in any of the cell lines
(HCR values in the range 0.26−15, with a median of only 1.8)
in marked contrast to TPZ (HCR 35−74) and SN30000 (HCR
47−173) which showed consistently greater HCRs than TPZ in
all cell lines as previously reported.52 The high aerobic cyto-
toxicities and lack of potentiation by hypoxia are consistent
with the rapid fragmentation of these compounds to cytotoxic
oxidizing radicals at rates too high for oxygen to competitively
reoxidize the initial one-electron reduced radicals.
(3b) of the CF3-substituted QDO compounds. However,
formation of the quinoxalinyl radical at the C5 position by the
elimination of water (4a) is more exothermic than that for the
elimination of the HO•, supporting the proposal for its concur-
rent formation. Formation of the quinoxalinyl radical at the C8
position, via the N1−OH anion (4c) is mainly less favorable.
Calculations for all three 3-Ph-substituted compounds show
that formation of an aryl radical on the Ph moiety (4b) is a
more strongly exothermic processes than for the quinoxalinyl
radical on C5. Additionally, calculations for compound 19
exhibited a near equi-exothermic outcome for the formation of
quinoxalinyl radicals on C5 and C8. As for 3-CH3-substituted
BTO compounds,41 formation of the 3-H2C• radical from the
radical anion of compound 20 (4b) is highly exothermic.
DISCUSSION
■
One-electron reduction of the QDO compounds, followed by
the first-order decay of their protonated radical anions, pro-
duces a variety of oxidizing radicals. Release of HO• from the
3-CF3-substituted QDOs, compounds 5 and 10, was demon-
strated previously by EPR using DEPMPO as the spin trap17
and now has been confirmed indirectly for these and other anal-
ogues in this study using DEPMPO combined with DMSO. This
approach allows for the observation of trapped CH3-DEPMPO
Table 3. DFT Calculations for the Change in Energy (kcal mol−1) for the Protonation Affinity and Formation of Radicals by the
a
Pathways Shown in Scheme 3
c
d
c
d
c
c
,
e
d
cmpd
(1a)
(2a)
(2b)
(3a)
(3b)
(4a)
(4b)
(4c)
f
4
−52.9−90.5
−44.9−88.2
−41.8−82.9
−44.9−83.9
−42.6−81.0
−34.2−75.9
−43.1−82.0
−311.0−273.9
−313.7−274.8
−319.7−278.0
−315.9−278.1
−317.5−278.8
−324.9−282.6
−314.6−278.4
−312.4−273.5
−315.5−274.9
−320.3−277.5
−318.7−277.8
−320.2−278.9
−321.6−281.7
−313.2−276.3
0.8 0.0
0.1−0.9
−0.8−2.6
1.0 1.4
5.7 2.4
5.2 2.0
0.1−2.1
−1.5−3.8
−3.0−6.5
−0.2−1.2
0.6−0.7
x x
x x
x x
4.8 0.6
4.8 0.1
b
5
b
10
−3.3 0.1
5.1 1.7
0.0 0.0
11
12
19
20
−3.2−4.5
−4.2−5.9
−4.0−6.5
−26.0−26.3
3.9−0.8
4.5−0.1
−4.3−5.1
−3.3−3.9
0.0 0.0
3.9 0.6
−2.9−0.6
0.5 2.1
−1.7−0.7
−2.0−1.3
−2.4−5.0
−0.7−0.6
a
b
c
d
First values are for the gas phase, and the second values are in water. Data are drawn from ref 17. Protonation at N4−O. Protonation at N1−O.
e
f
Radical formation on the 3-substituent. No pathway. See Scheme S1 in Supporting Information for a full listing of pathways.
K
DOI: 10.1021/acs.chemrestox.6b00133
Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology
Article
Table 4. Antiproliferative Potency of QDOs and the BTOs Tirapazamine (TPZ) and SN30000
b
HT29
SiHa
SiHa-POR
POR ratio
a
c
c
c
compd
TPZ
aerobic IC50 (μM)
HCR
aerobic IC50 (μM)
HCR
aerobic IC50 (μM)
HCR
oxic
hypoxic
424 18 (15)
375 89 (5)
0.28 0.09 (5)
0.88 0.18 (4)
12.6 1.1 (2)
28.7 6.4 (3)
40.1 5.8 (3)
107 1 (2)
>100 (2)
74.5
159
0.26
0.38
2.5
1.5
1.2
5.2
nd
144 7 (16)
245 27 (5)
0.25 0.01(2)
0.62 0.18 (2)
4.93 1.39 (3)
15.4 2.6 (3)
25.4 6.9 (3)
26.8 5.6 (2)
12.8 5.6 (3)
5.78 1.98 (4)
0.69 0.29 (2)
255 35 (2)
50.0 (1)
42.3
173
0.27
0.33
2.1
2.4
6.3
3.1
1.9
1.3
0.6
5.4
1.2
2.2
1.5
2.6
nd
20.5 1.26 (14)
18.4 3.23 (3)
0.08 0.03 (3)
0.12 0.03 (3)
0.60 0.08 (2)
1.45 0.35 (2)
1.47 0.39 (3)
2.15 0.04 (2)
1.35 0.37 (2)
2.63 0.73 (2)
0.53 0.02 (2)
48.1 20.1 (2)
25.0 (1)
35.4
47.2
0.40
0.22
1.0
1.1
1.2
3.5
1.8
3.0
1.7
15.0
0.9
2.4
1.5
2.2
nd
7.0
13.3
3.0
5.2
8.2
10.6
17.3
12.5
9.5
2.2
1.3
5.3
2.0
6.7
3.9
5.2
nd
5.9
3.6
4.5
3.5
3.9
5.1
3.2
14.1
9.2
5.1
3.8
14.7
1.5
7.2
4.0
4.5
nd
SN30000
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
7.75 1.83 (4)
0.81 0.21 (4)
335 20 (2)
140 (1)
1.4
0.9
2.5
2.9
1.1
0.6
1.5
1.1
1.5
1.1
nd
842 268 (2)
506 (1)
1120 204 (2)
1223 775 (2)
3135 531 (2)
>75 (3)
168 22 (2)
314 30 (2)
608 104 (2)
>75 (2)
1784 371 (2)
60.3 5.0 (3)
544 53 (2)
71.0 9 (2)
>3000 (2)
>800 (2)
nd
>800 (2)
nd
nd
2.8
5.2
40 3 (2)
1.7
nd
22.8 6.1 (4)
1694 143 (2)
5.0
3.8
1.8
nd
>3000 (2)
nd
a
b
Cells were exposed to compounds for 4 h. Values are means, and errors are SEMs, with the number of determinations given in parentheses. SiHa
c
IC50/SiHa-POR IC50. Hypoxic cytotoxicity ratio (aerobic IC50/hypoxic IC50). nd, not determined.
species, formed by HO• addition to DMSO and elimination of
the methyl radical. Unusually, we were able to observe HO•
spin-trapped by PBN directly, most likely because rapid
metabolism of the highly electron-affinic compound 5, giving
rise to a substantial steady-state concentration of HO•. PBN
trapped a C-centered radical formed from both 9 and 10, which
may be formed by HO• addition to the benzene ring of the
QDO core. However, the same hfc’s were obtained upon spin-
trapping one of the radicals formed upon one-electron reduc-
tion of tirapazamine, which does not release the HO• radical
upon one-electron reduction, and is assigned as a C-centered
radical formed on both the BTO and QDO rings (e.g., the
intermediate formed by pathway 2a, Scheme 1). An alternative
explanation based on theoretical calculations for BTO ring
opening following one-electron reduction to yield C-centered
radicals53 seems unlikely as no ring-opened products have been
reported for either ring systems. EPR data obtained on the
reduction of 8 indicates that a C-centered radical with a large
hfc for aH was formed, suggestive of an aryl-type radical, i.e.,
possibly a quinoxalinyl radical (Scheme 3) with slightly smaller
hfc for aH than for a phenyl-type radical. In addition, the
oxymethyl radical (•CH2O−) was detected (large hfc for aP),
which must arise from H atom abstraction from the CH3O-
substituent, either by a released HO• or possibly the suggested
quinoxalinyl radical. PBN did not produce the CH3−PBN
species in the presence of DMSO but did trap a different
C-centered radical consistent with H atom abstraction having
taken place from a methyl group on DMSO. The same spin-
trapped radical derived from DMSO was also observed for
the 3-Ph and 3-CH3 substituted QDO compounds, implying a
general phenomenon for these QDO compounds. Given that
HO• adds to DMSO, the strongly oxidizing radical which
abstracts the H atom from DMSO must have been formed
on the QDO core. Such a scenario implies that DMSO reacts
quickly with this oxidizing radical, forming a C-centered radical
on DMSO, which is more efficiently spin trapped by PBN
than by DEPMPO. It also implies that PBN spin-traps the
C-centered radical on DMSO more efficiently than the methyl
radical.
The pulse radiolysis data provide insight into the time scale
for the formation of the range of oxidizing radicals identified by
EPR. The formation of two different radical anion species can
be rationalized by protonation occurring at both the N1−O and
N4−O. Whereas water elimination involving the protonated
N1−OH moiety could only lead to the formation of a
quinoxalinyl-type radical (path 4c, Scheme 3), elimination
involving the N4−OH moiety could in addition involve the
substituent at the 3-position, i.e., the formation of the HO• for
3-CF3 substitution (path 3a), a generic formation of an aryl
radical at C5 (path 4a), formation of the −H2C• radical for
3-CH3 substitution (path 4b), and an exonuclear aryl-type
radical for 3-Ph substitution (path 4b). The rate constants for
the loss of absorbance at 520 nm at pH 7, k, increase as the Eo′
of the compounds decrease (Figure 10):
log k = −(1.05 0.32) − (4.59 0.65)Eo′ /V
This is understood from the associated increase in pKr
values, leading to a greater proportion of protonated radical
anion. A comparison between the rate constant data obtained
for QDO at pH 7 and published data for the BTO compounds
(Figure 10) shows that kelim is ca. an order of magnitude greater
for the QDO compounds than that for BTO compounds of
similar Eo′ between ca. −550 and −350 mV. For example, the
measured kelim values for the BTO compound TPZ, 1 (Eo′ =
−456
10 mV), and a QDO compound of comparable
electron-affinity, compound 11 (Eo′ = −439
8 mV), are
26
83 6 s−1 and 2050 450 s−1, respectively. While the back
oxidation of both BTO and QDO radical anions exhibit similar
dependency on the Eo′ of the compounds (i.e., for TPZ, 1 and
11, a kO2, value of ∼6 × 106 M−1 s−1), and the kinetics of back-
oxidation by O2 is similar in both series, the rapid decay of
the QDO radicals to cytotoxic intermediates out competes this
L
DOI: 10.1021/acs.chemrestox.6b00133
Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology
Article
nature of the oxidizing radicals formed from the one-electron
reduced compounds needs to be considered. The two major
series of QDO compounds in this work, −CF3 substituted and
−phenyl substituted QDO compounds, can only form aryl
radicals through the proposed dehydration reactions, as well
as in the case of the CF3 substituted QDO compounds, reac-
tion with the HO• radical. BTO compounds exhibiting good
hypoxia-selectivity generally have 3-NH2, 3-HNalkyl, or 3-alkyl
substituents26,54,55 which would form N-centered radicals
through the putative dehydration reaction. Such N-centered
radicals are not as reactive as the HO• radical or aryl radicals
and may well be able to diffuse greater distances to critical
cellular targets. A tentative conclusion can be made on the basis
of these observations is that hypoxia selectivity depends to a
large extent on the substituent adjacent to the N−OH moiety
having a labile hydrogen. This conclusion is supported by the
significant hypoxic selectivity demonstrated by QDO 3 which
has an adjacent H atom on the 3-NH2 group,3 whereas the
3-phenyl analogue of tirapazamine does not exhibit hypoxic
selectivity.55
Figure 10. Dependency of k on Eo′ for QDO compounds at pH 7
■
○
( ); the “plateau” level at low pH ( ). The dashed line represents
data for a series of BTO compounds at pH 7 from ref 26.
back oxidation. These kinetic differences between the two
classes of bioreductive compounds could result in higher levels
of cytotoxic radicals for the QDO compounds leading to
increased cytotoxicity under oxic conditions for QDO com-
pounds. The dependence of the aerobic cytotoxicity of the
QDOs for which full data were obtained (Table 4) on their Eo′
is illustrated for HT29 cells in Figure 11. While there is a good
This study has identified the formation of a range of reactive
oxygen-sensitive and oxygen-insensitive radicals produced upon
the one-electron reduction of a series of QDO compounds
and puts forward a kinetic explanation for the observed increase
in aerobic toxicity of these compounds compared to BTO
compounds. These studies underline the importance of both
redox and radical kinetic studies in the development of anti-
cancer bioreductive drugs.
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the ACS
Publications website at DOI: 10.1021/acs.chemrestox.6b00133.
Pulse radiolysis spectral and wavelength-dependent
kinetic data for 10, exemplary data for the oxidation of
the radical anions by O2 (for 19), cytotoxicity study in
the presence of DMSO, and DFT information (PDF)
Figure 11. Dependence of the aerobic cytotoxicity of HT29 cells on
the Eo′ values of QDO compounds. Black circles (4−9) are 3-CF3
substituted, and open circles (11−19) are 3-Ph substituted
compounds. Data were taken from Tables 1 and 4.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: r.anderson@auckland.ac.nz.
Funding
correlation between IC50 in HT29 cells and Eo′ for the 3-CF3
analogues, the data are scattered for the 3-Ph analogues. The
fact that similar IC50 values for all 3-Ph analogues are at lower
Eo′ values than those for the 3-CF3 analogues implies that more
effective cytotoxic radicals are formed than the HO• radical,
which is known to be released from the 3-CF3 analogues. This
may reflect the known short diffusion range of the HO• radical
in the biological milieu compared to the suggested aryl radical
and/or its peroxy form if the radicals are produced some
distance away from a critical target. However, differences in cell
uptake and/or bioreductive metabolism between the two series
of analogues may also play a role.
The similar aerobic and hypoxic cytotoxicities obtained for
the QDOs in this study, which result in low hypoxic selectivity,
are in marked contrast to similar BTO compounds for which
the aerobic cytotoxic potency is considerably lower, leading to
hypoxic selectivity. Apart from kinetic differences underlying
the differences between the two classes of compounds, the
This work was financially supported by the Health Research
Council of New Zealand, grant 11/323.
Notes
The authors declare no competing financial interest.
ABBREVIATIONS
■
QDO, 1,4-quinoline 1,4-N,N-dioxide; BTO, 1,2,4-benzotriazine
1,4-N,N-dioxide; PBN, N-tert-butyl-α-phenylnitrone; DEPM-
PO, 5-(diethoxyphosphoryl)-5-methyl-1-pyrroline-N-oxide;
DMSO, dimethysulfoxide; POR, N-terminal truncated cyto-
chrome P450 reductase; TPZ, tirapazamine; SN30000, (3-[3-
(4-morpholinyl)propyl]-7,8-dihydro-6H-indeno[5,6-e][1,2,4]-
triazine 1,4-dioxide; EPR, electron paramagnetic resonance;
DFT, density functional theory; Eo′, one-electron reduction
potential; HT29, human colorectal carcinoma cells; SiHa,
human cervical carcinoma cells; HCR, hypoxic cytotoxicity
ratio.
M
DOI: 10.1021/acs.chemrestox.6b00133
Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology
Article
(17) Yadav, P., Marshall, A. J., Reynisson, J., Denny, W. A., Hay, M.
P., and Anderson, R. F. (2014) Fragmentation of the quinoxaline N-
oxide bond to the.OH radical upon one-electron bioreduction. Chem.
Commun. 50, 13729−13731.
(18) Davis, S. M., Shaw, J. T., and Voorhies, J. D. (1965)
Depolarising Agents for Dry Cells, US Patent US3260621.
(19) Mallory, F. B., Manatt, S. L., and Wood, C. S. (1965) Rates of
some degenerate rearrangements as determined by nuclear magnetic
resonance spectroscopy. J. Am. Chem. Soc. 87, 5433−5438.
(20) Leyva, S., Castanedo, V., and Letva, E. (2003) Synthesis of novel
fluorobenzofuroxans by oxidation of anilines and thermal cyclization of
arylazides. J. Fluorine Chem. 121, 171−175.
(21) Hay, M. P., Hicks, K. O., Pchalek, K., Lee, H. H., Blaser, A.,
Pruijn, F. B., Anderson, R. F., Shinde, S. S., Wilson, W. R., and Denny,
W. A. (2008) Tricyclic [1,2,4]triazine 1,4-dioxides as hypoxia selective
cytotoxins. J. Med. Chem. 51, 6853−6865.
(22) Boido, A., Novelli, F., Savelli, F., Sparatore, F., Russo, S.,
Fillippelli, W., Susanna, V., and Marmo, E. (1989) Synthesis and
pharmacological activity of 3-methyl-3,4-dihydrobenzo- and pyrido-
1,2,4-triazine-3-acetamides. Farmaco 44, 279−301.
REFERENCES
■
(1) Brown, J. M., and Lemmon, M. J. (1990) Potentiation by the
hypoxic cytotoxin SR 4233 of cell killing produced by fractionated
irradiation of mouse tumors. Cancer Res. 50, 7745−7749.
(2) Dorie, M. J., and Brown, J. M. (1997) Modification of the
antitumor activity of chemotherapeutic drugs by the hypoxic cytotoxic
agent tirapazamine. Cancer Chemother. Pharmacol. 39, 361−366.
(3) Monge, A., Palop, J. A., Lopez de Cerain, A., Senador, V.,
Mertinez-Crespo, F. J., Sainz, Y., Narro, S., Garcia, E., de Miguel, C.,
Gonzalez, M., Hamilton, E., Barker, A. J., Clarke, E. D., and Greenhow,
D. T. (1995) Hypoxia-selective agents derived from quinoxalin 1,4-di-
N-oxides. J. Med. Chem. 38, 1786−1792.
(4) Crawford, P. W., Scamehorn, R. G., Hollstein, U., Ryan, M. D.,
and Kovacic, P. (1986) Cyclic voltametry of phenazines and
quinoxalines including mono- and di-N-oxides.mRelation to structure
and antimicrobial activity. Chem.-Biol. Interact. 60, 67−84.
(5) Suter, W., Rosselet, A., and Knusel, F. (1978) Mode of action of
quindoxin and substituted quinoxaline-di-N-oxides on Escherichia coli.
Antimicrob. Agents Chemother. 13, 770−783.
(6) Baker, M. A., Zeman, E. M., Hirst, V. K., and Brown, J. M. (1988)
Metabolism of SR 4233 by chinese hamster ovary cells: basis of
selective hypoxic cytotoxicity. Cancer Res. 48, 5947−5952.
(23) Anderson, R. F., Denny, W. A., Li, W., Packer, J. E., Tercel, M.,
and Wilson, W. R. (1997) Pulse radiolysis studies on the
fragmentation of arylmethyl quaternary nitrogen mustards by one-
electron reduction in aqueous solution. J. Phys. Chem. A 101, 9704−
9709.
(7) Biedermann, K. A., Wang, J., Graham, R. P., and Brown, J. M.
(1991) SR 4233 cytotoxicity and metabolism in DNA repair-
competent and repair-deficient cell cultures. Br. J. Cancer 63, 358−362.
(8) Brown, J. M., and Wang, L.-H. (1998) Tirapazamine: laboratory
data relevant to clinical activity. Anti-Cancer Drug Des. 13, 529−539.
(9) Hicks, K. O., Siim, B. G., Jaiswal, J. K., Pruijn, F. B., Fraser, A. M.,
Patel, R., Hogg, A., Liyanage, H. D. S., Dorie, M. J., Brown, J. M.,
Denny, W. A., Hay, M. P., and Wilson, W. R. (2010) Parmacokinetic/
pharmacodynamic modeling identifies SN30000 and SN29751 as
tirapazamine analogues with improved tissue penetration and hypoxic
cell killing in tumors. Clin. Cancer Res. 16, 4946−4957.
(10) Zeman, E. M., Hirst, V. K., Lemmon, M. J., and Brown, J. M.
(1988) Enhancement of radiation-induced tumor cell killing by the
hypoxic cell toxin SR 4233. Radiother. Oncol. 12, 209−218.
(11) Rischin, D., Peters, L., Fisher, R., Macann, A., Denham, J.,
Poulsen, M., Jackson, M., Kenny, L., Penniment, M., Corry, J., Lamb,
D., and McClure, B. (2005) Tirapazamine, cisplatin, and radiation
versus fluorouracil, cisplatin, and radiation in patients with locally
advanced head and neck cancer: A randomized phase II trial of the
trans-tasman radiation oncology group (TROG 98.02). J. Clin. Oncol.
23, 79−87.
(12) von Pawel, J., von Roemeling, R., Gatzemeier, U., Boyer, M.,
Elisson, L. O., Clark, P. T. D., Rey, A., Butler, T. W., Hirsh, V., Olver,
I., Bergman, B., Ayoub, J., Richardson, G., Dunlop, D., Arcenas, A.,
Vescio, R., Viallet, J., and Treat, J. (2000) Tirapazamine plus cisplatin
versus cisplatin in advanced non-small-cell lung cancer: A report of the
international CATAPULT I study group. Cisplatin and tirapazamine in
Subjects with Advanced Previously Untreated Non-Small-Cell Lung
Tumors. J. Clin. Oncol. 18, 1351−1359.
(13) Monge, A., Martinez-Crespo, F. J., Decerain, A. L., Palop, J. A.,
Narro, S., Senador, V., Marin, A., Sainz, Y., Gonalez, M., Hamilton, E.,
and Baker, A. J. (1995) Hypoxia-selective agents derived from 2-
quinoxalinecarbonitrile-1,4-di-N-oxides. J. Med. Chem. 38, 4488−4494.
(14) Ismail, M. M. F., Amin, K. M., Noaman, E., Soliman, D. H., and
Ammar, Y. A. (2010) New quinoxaline 1,4-di-N-oxides: Anticancer
and hypoxia-selective therapeutic agents. Eur. J. Med. Chem. 45, 2733−
2738.
(24) Schwarz, H. A., and Dodson, R. W. (1989) Reduction potentials
.‑
of CO2 and the alcohol radicals. J. Phys. Chem. 93, 409−414.
(25) Wardman, P. (1989) Reduction potentials of one-electron
couples involving free radicals in aqueous solution. J. Phys. Chem. Ref.
Data 18, 1637−1755.
(26) Anderson, R. F., Shinde, S. S., Hay, M. P., Gamage, S. A., and
Denny, W. A. (2005) Radical properties governing the hypoxia-
selective cytotoxicity of antitumor 3-amino-1,2,4-benzotriazine 1,4-
dioxides. Org. Biomol. Chem. 3, 2167−2174.
(27) Mulazzani, Q. G., Venturi, M., Hoffman, M. Z., and Rodgers, M.
A. J. (1986) Interaction of formate and oxalate ions with radiation-
generated radicals in aqueous solution. Methylviologen as a
mechanistic probe. J. Phys. Chem. 90, 5347−5352.
(28) Frisch, M. J., Trucks, G. W., Schlegel, H. B., Scuseria, G. E.,
Robb, M. A. J., Cheeseman, J. R, Scalmani, G., Barone, V., Mennucci,
B., Petersson, G. A., Nakatsuji, H., Caricato, M., Li, X., Hratchian, H.
P., Izmaylov, A. F., Bloino, J., Zheng, G., Sonnenberg, J. L., Hada, M.,
Ehara, M., Toyota, K., Fukuda, R., Hasegawa, J., Ishida, M., Nakajima,
T., Honda, Y., Kitao, O., Nakai, H., Vreven, V., Montgomery, J. J. A.,
Peralta, E., Ogliaro, F., Bearpark, M., Heyd, J. J., Brothers, E., Kudin, K.
N., Staroverov, V. N., Kobayashi, R., Normand, J., Raghavachari, K.,
Rendell, A., Burant, J. C., Iyengar, S. S., Tomasi, J., Cossi, M., Rega, N.,
Millam, J. M., Klene, M., Knox, J. E., Cross, J. B., Bakken, V., Adamo,
C., Jaramillo, J., Gomperts, R., Stratmann, R. E., Yazyev, O., Austin, A.
J., Cammi, R., Pomelli, C., Ochterski, J. W., Martin, R. L., Morokuma,
K., Zakrzewski, V. G., Voth, G. A., Salvador, P., Dannenberg, J. J.,
Dapprich, S., Daniels, A. D., Farkas, O., Foresman, J. B., Ortiz, J. V.,
Cioslowski, J., and Fox, D. J. Gaussian 09, revision D.02., Gaussian,
Inc., Wallingford, CT, 2009.
(29) Becke, A. D. (1988) Density-functional exchange-energy
approximation with correct asymptotic behavior. Phys. Rev. A: At.,
Mol., Opt. Phys. 38, 3098−3100.
(30) Becke, A. D. (1993) Density-functional thermochemistry. III.
The role of exact exchange. J. Chem. Phys. 98, 5648−5642.
(31) Lee, C. T., Yang, W. T., and Parr, R. G. (1988) Development of
the Cole-Salvetti correlation-energy formula into a functional of the
electron density. Phys. Rev. B: Condens. Matter Mater. Phys. 37, 785−
789.
́
(15) Junnotula, V., Rajapakse, A., Arbillaga, L., LOpez de Cerain, A.,
Solano, B., Villar, R., Monge, A., and Gates, K. S. (2010) DNA strand
cleaving properties and hypoxia-selective cytotoxicity of 7-chloro-2-
thienylcarbonyl-3-trifluoromethylquinoxaline 1,4-dioxide. Bioorg. Med.
Chem. 18, 3125−3132.
(16) Solano, B., Junnotula, V., Marin, A., Villar, R., Burguete, A.,
Vicente, E., Perez-Silanes, S., Aldana, I., Monge, A., Dutta, S., Sarkar,
U., and Gates, K. S. (2007) Synthesis and biological evaluation of new
2-arylcarbonyl-3-trifluoromethylquinoxaline 1,4,-dioxide derivatives
and their reduced analogues. J. Med. Chem. 50, 5485−5492.
(32) Frisch, M. J., Pople, J. A., and Binkley, J. S. (1984) J. Chem. Phys.
80, 3265−3269.
(33) Hariharan, P. C., and Pople, J. A. (1973) The influence of
polarization functions on molecular orbital hydrogenation energies.
Theoretica Chim. Acta 28, 213−222.
N
DOI: 10.1021/acs.chemrestox.6b00133
Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology
Article
(34) Wong, M. W. (1996) Vibrational frequency prediction using
density functional theory. Chem. Phys. Lett. 256, 391−399.
(35) Tomasi, J., Mennucci, B., and Cammi, R. (2005) Chem. Rev. 105,
2999−3093.
(36) Forseman, J. B., and Frisch, A. E. (1996) Exploring Chemistry
with Electronic Structure Methods, pp 142−143, Gaussian Inc.:
Wallingford, CT.
homolysis, stepwise dehydration, or triazene ring-opening. Chem. Res.
Toxicol. 25, 634−645.
(54) Hay, M. P., Gamage, S. A., Kovacs, M. S., Pruijn, F. B.,
Anderson, R. F., Patterson, A. V., Wilson, W. R., Brown, J. M., and
Denny, W. A. (2003) Structure-activity relationships of 1,2,4-
benzotriazine 1,4-dioxides as hypoxia-selective analalogues of tirapaz-
amine. J. Med. Chem. 46, 169−182.
(55) Hay, M. P., Pchalek, K., Pruijn, F. B., Hicks, K. O., Siim, B. G.,
Anderson, R. F., Shinde, S. S., Phillips, V., Denny, W. A., and Wilson,
W. R. (2007) Hypoxia-selective 3-alkyl 1,2,4-benzotriazine 1,4-
dioxides: The influence of hydrogen bond donors on extravascular
transport and antitumor activity. J. Med. Chem. 50, 6654−6664.
(37) Anderson, R. F., Yadav, P., Patel, D., Reynisson, J., Tipparaju, S.
R., Guise, C. P., Patterson, A. V., Denny, W. A., Maroz, A., Shinde, S.
S., and Hay, M. P. (2014) Characterisation of radicals formed by the
triazine 1,4-dioxide hypoxia-activated prodrug, SN30000. Org. Biomol.
Chem. 12, 3386−3392.
(38) Guise, C. P., Wang, A. T., Theil, A., Bridewell, D. J., Wilson, W.
R., and Patterson, A. V. (2007) Identification of human reductases that
activate the dinitrobenzamide mustard prodrug PR-104A: A role for
the NADPH:cytochrome P450 oxidoreductase under hypoxia.
Biochem. Pharmacol. 74, 810−820.
(39) Hunter, F. W., Wang, J., Patel, R., Hsu, H.-H., Hickey, A. J. R.,
Hay, M. P., and Wilson, W. R. (2012) Homologous recombination
repair-dependent cytotoxicity of the benzotriazine di-N-oxide CEN-
209: Comparison with other hypoxia-activated prodrugs. Biochem.
Pharmacol. 83, 574−585.
(40) Anderson, R. F., Shinde, S. S., Hay, M. P., and Denny, W. A.
(2006) Potentiation of the cytotoxicity of the anticancer agent
tirapazamine by benzotriazine N-oxides:The role of redox equilibria. J.
Am. Chem. Soc. 128, 245−249.
(41) Shinde, S. S., Maroz, A., Hay, M. P., Patterson, A. V., Denny, W.
A., and Anderson, R. F. (2010) Characterization of radicals formed
following enzymatic reduction of 3-substituted analogues of the
hypoxia-activated cytotoxin 3-amino-1,2,4-benzotriazine 1,4-dioxide
(tirapazamine). J. Am. Chem. Soc. 132, 2591−2599.
(42) Alfassi, Z. B., Marguet, S., and Neta, P. (1994) Formation and
reactivity of phenylperoxyl radicals in aqueous solution. J. Phys. Chem.
98, 8019−8023.
(43) Ranguelova, K., and Mason, R. P. (2011) The fidelity of spin
trapping with DMPO in biological systems. Magn. Reson. Chem. 49,
152−158.
(44) Kotake, Y., and Janzen, E. G. (1991) Decay and fate of the
hydroxyl radical adduct od a-phenyl-N-tert-butylnitrone in aqueous
media. J. Am. Chem. Soc. 113, 9503−9506.
(45) Buettner, G. R. (1987) Spin trapping:ESR parameters of spin
adducts. Free Radical Biol. Med. 3, 259−303.
(46) Hill, H. A. O., and Thornalley, P. J. (1983) The effect of spin
traps on phenylhydrazine-induced haemolysis. Biochim. Biophys. Acta,
Mol. Cell Res. 762, 44−51.
(47) Janzen, E. G., Coulter, G. A., Oehler, U. M., and Bergsma, J. P.
(1982) Solvent effects on the nitrogen and β-hydrogen hyperfine
splitting constants of aminoxyl radicals obtained in spin trapping
experiments. Can. J. Chem. 60, 2725−2733.
(48) Dixon, W. T., Norman, R. O. C., and Buley, A. L. (1964)
Electron spin resonance studies of oxidation. Part II. Aliphatic acids
and substituted acids. J. Chem. Soc., 3625−3634.
(49) Karoui, H., Chalier, F., Finet, J.-P., and Tordo, P. (2011)
DEPMPO: an efficient tool for the coupled ESR-spin trapping of
alkylperoxyl radicals in water. Org. Biomol. Chem. 9, 2473−2480.
(50) Saprin, A., and Piette, L. H. (1977) Spin trapping and its
application in the study of lipid peroxidation and free radical
production with liver microsomes. Arch. Biochem. Biophys. 180, 480−
492.
(51) Anderson, R. F. (1979) Oxidation of the cyclohexadienyl radical
by metal ions: A pulse radiolysis study. Radiat. Phys. Chem. 13, 155−
157.
(52) Wang, J., Guise, C. P., Dachs, G. U., Phung, Y., Hsu, A. U.,
Lambie, N. K., Patterson, A. V., and Wilson, W. R. (2014)
Identification of one-electron reductases that activate both the hypoxia
prodrug SN30000 and diagnostic probe EF5. Biochem. Pharmacol. 91,
436−446.
(53) Yin, J., Glaser, R., and Gates, K. S. (2012) On the reaction
mechanism of tirapazamine reduction chemistry: Unimolecular N-OH
O
DOI: 10.1021/acs.chemrestox.6b00133
Chem. Res. Toxicol. XXXX, XXX, XXX−XXX