3-amino-1,2,4-benzotriazine 4-oxide: Characterization of a new metabolite arising from bioreductive processing of the antitumor agent 3-amino-1,2,4-benzotriazine 1,4-dioxide (tirapazamine)

Article abstract of DOI:10.1021/jo001232j

Tirapazamine (1) is a promising antitumor agent that selectively causes DNA damage in hypoxic tumor cells, following one-electron bioreductive activation. Surprisingly, after more than 10 years of study, the products arising from bioreductive metabolism of tirapazamine have not been completely characterized. The two previously characterized metabolites are 3-amino-1,2,4-benzotriazine 1-oxide (3) and 3-amino-1,2,4-benzotriazine (5). In this work, 3-amino-1,2,4-benzotriazine 4-oxide (4) is identified for the first time as a product resulting from one-electron activation of the antitumor agent tirapazamine by the enzymes xanthine/xanthine oxidase and NADPH:cytochrome P450 oxidoreductase. As part of this work, the novel N-oxide (4) was unambiguously synthesized and characterized using NMR spectroscopy, UV-vis spectroscopy, LC/MS, and X-ray crystallography. Under conditions where the parent drug tirapazamine is enzymatically activated, the metabolite 4 is produced but readily undergoes further reduction to the benzotriazine (5). Thus, under circumstances where extensive reductive metabolism occurs, the yield of the 4-oxide (4) decreases. In contrast, the isomeric two-electron reduction product 3-amino-1,2,4-benzotriazine 1-oxide (3) does not readily undergo enzymatic reduction and, therefore, is found as a major bioreductive metabolite under all conditions. Finally, the ability of the 4-oxide metabolite (4) to participate in tirapazamine-mediated DNA damage is considered.

Full text of DOI:10.1021/jo001232j

J . Org. Chem. 2001, 66, 107-114
107
3-Am in o-1,2,4-ben zotr ia zin e 4-Oxid e: Ch a r a cter iza tion of a New
Meta bolite Ar isin g fr om Bior ed u ctive P r ocessin g of th e An titu m or
Agen t 3-Am in o-1,2,4-ben zotr ia zin e 1,4-Dioxid e (Tir a p a za m in e)
Tarra Fuchs, Goutam Chowdhury, Charles L. Barnes, and Kent S. Gates*
Departments of Chemistry and Biochemistry, University of MissourisColumbia,
Columbia, Missouri 65211
GatesK@missouri.edu
Received August 11, 2000
Tirapazamine (1) is a promising antitumor agent that selectively causes DNA damage in hypoxic
tumor cells, following one-electron bioreductive activation. Surprisingly, after more than 10 years
of study, the products arising from bioreductive metabolism of tirapazamine have not been
completely characterized. The two previously characterized metabolites are 3-amino-1,2,4-benzo-
triazine 1-oxide (3) and 3-amino-1,2,4-benzotriazine (5). In this work, 3-amino-1,2,4-benzotriazine
4-oxide (4) is identified for the first time as a product resulting from one-electron activation of the
antitumor agent tirapazamine by the enzymes xanthine/xanthine oxidase and NADPH:cytochrome
P450 oxidoreductase. As part of this work, the novel N-oxide (4) was unambiguously synthesized
and characterized using NMR spectroscopy, UV-vis spectroscopy, LC/MS, and X-ray crystal-
lography. Under conditions where the parent drug tirapazamine is enzymatically activated, the
metabolite 4 is produced but readily undergoes further reduction to the benzotriazine (5). Thus,
under circumstances where extensive reductive metabolism occurs, the yield of the 4-oxide (4)
decreases. In contrast, the isomeric two-electron reduction product 3-amino-1,2,4-benzotriazine
1-oxide (3) does not readily undergo enzymatic reduction and, therefore, is found as a major
bioreductive metabolite under all conditions. Finally, the ability of the 4-oxide metabolite (4) to
participate in tirapazamine-mediated DNA damage is considered.
In tr od u ction
tion of this drug have not been completely characterized.
In a recent review,15 it was noted that, in many experi-
ments, between 30 and 60% of the starting drug is left
unaccounted for by the known metabolites, 3-amino-1,2,4-
benzotriazine 1-oxide (3) and 3-amino-1,2,4-benzotriazine
(5).16,17 It is important to fully characterize the metabo-
lites stemming from bioreductive activation of tira-
pazamine because these products may provide insight
into the chemical events underlying drug action. In
addition, all metabolites should be examined for their
biological activity.
Tirapazamine (1) selectively kills the oxygen-poor
(hypoxic) cells found in solid tumors and is currently
undergoing phase I, II, and III clinical trials for the
treatment of various cancers.1 Tirapazamine undergoes
in vivo enzymatic reduction to its radical form (2)2-5 and,
in the absence of molecular oxygen, the resulting drug
radical, or its fragmentation product hydroxyl radical
(Scheme 1), causes cytotoxic DNA damage.5-12 Tira-
pazamine is relatively nontoxic to normally oxygenated
cells because the activated form of the drug (2) is rapidly
quenched by reaction with O2.13,14
Here we report the identification and characterization
of a new tirapazamine metabolite, 3-amino-1,2,4-benzo-
triazine 4-oxide (4). This compound is produced from in
vitro one-electron reductive activation of the antitumor
agent tirapazamine by the enzymes xanthine/xanthine
oxidase and NADPH:cytochrome P450 oxidoreductase.
Under conditions where the parent drug tirapazamine
is enzymatically activated, the metabolite 4 is produced
and then readily undergoes further reduction to the
benzotriazine 5. Thus, the yield of the 4-oxide metabolite
(4) varies depending upon reaction conditions and, under
circumstances where more extensive reductive metabo-
Although tirapazamine has been studied for over 10
years, the products stemming from bioreductive activa-
* Correspondence to Kent S. Gates. Phone number: (573) 882-6763;
Fax number: (573) 882-2754.
(1) Brown, J . M. Cancer Res. 1999, 59, 5863-5870.
(2) Brown, J . M. Br. J . Cancer 1993, 67, 1163-1170.
(3) Patterson, A. V.; Barham, H. M.; Chinje, E. C.; Adams, G. E.;
Harris, A. L.; Stratford, I. J . Br. J . Cancer 1995, 72, 1144-.
(4) Costa, A. K.; Baker, M. A.; Brown, J . M.; Trudell, J . R. Cancer
Res. 1989, 49, 925-.
(5) Biedermann, K. A.; Wang, J .; Graham, R. P. Br. J . Cancer 1991,
63, 358-362.
(6) Daniels, J . S.; Gates, K. S. J . Am. Chem. Soc. 1996, 118, 3380-
3385.
(7) Fitzsimmons, S. A.; Lewis, A. D.; Riley, R. J .; Workman, P.
Carcinogenesis 1994, 15, 1503-1510.
(13) Wardman, P.; Priyadarsini, K. I.; Dennis, M. F.; Everett, S. A.;
Naylor, M. A.; Patel, K. B.; Stratford, I. J .; Stratford, M. R. L.; Tracy,
M. Br. J . Cancer 1996, 74, S70-S74.
(8) Laderoute, K. L.; Wardman, P.; Rauth, M. Biochem. Pharmacol.
1988, 37, 1487-1495.
(14) Lloyd, R. V.; Duling, D. R.; Rumyantseva, G. V.; Mason, R. P.;
Bridson, P. K. Mol. Pharmacol. 1991, 40, 440-445.
(15) Patterson, A. V.; Saunders, M. P.; Chinje, E. C.; Patterson, L.
H.; Stratford, I. J . Anti-Cancer Drug Des. 1998, 13, 541-573.
(16) Laderoute, K.; Rauth, A. M. Biochem. Pharmacol. 1986, 35,
3417-3420.
(9) Siim, B. G.; van Zijl, P. L.; Brown, J . M. Br. J . Cancer 1996, 73,
952-960.
(10) Wang, J .; Biedermann, K. A.; Brown, J . M. Cancer Res. 1992,
52, 4473-4477.
(11) Peters, K. B.; Tung, J . J .; Brown, J . M. Proc. Am. Assoc. Cancer
Res. 2000, 41, 284.
(17) Baker, M. A.; Zeman, E. M.; Hirst, V. K.; Brown, J . M. Cancer
Res. 1988, 48, 5947-5952.
(12) Olive, P. L. Br. J . Cancer 1995, 71, 529.
10.1021/jo001232j CCC: $20.00 © 2001 American Chemical Society
Published on Web 12/08/2000

108 J . Org. Chem., Vol. 66, No. 1, 2001
Fuchs et al.
Sch em e 1
Syn th esis of 3-Am in o-1,2,4-ben zotr ia zin e 1,4-Dioxid e
(1). Compound 1 was prepared by a minor modification of
Mason and Tennant’s route.18 Compound 3 (250 mg, 1.49
mmol) was suspended in glacial acetic acid (12 mL) and heated
to 50 °C. To this suspension was added 30% hydrogen peroxide
(8 mL), and the resulting mixture was stirred in the dark for
10 h. Between two and four additional aliquots of acetic acid
(2 mL) and 30% hydrogen peroxide (2 mL) were added over
the next 24 h until starting material was largely consumed
(as judged by TLC). The reaction was lyophilized to dryness
(behind a blast shield), and the resulting burnt orange solid
purified by flash-column chromatography on silica gel (eluted
with 75:25 ethyl acetate:hexane, followed by 100% ethyl
acetate, followed by 25:75 2-propanol:ethyl acetate) to yield 1
(40-60%). The resulting red solid can be recrystallized from
2-propanol or ethanol to yield red plates: Rf ) 0.3 (15:85
2-propanol:ethyl acetate); 1H NMR (CDCl3) δ 8.45 (d, J ) 9.6
Hz, 1H), 8.10 (t, J ) 7.9 Hz, 2H), 7.82 (d, J ) 8.4 Hz, 1H),
7.68 (t, J ) 7.8 Hz, 1H), 6.84 (bs, 2H); 1H NMR (DMSO- d6) δ
8.21 (dd, J ) 8.3, 0.6 Hz, 1H), 8.15 (dd, J ) 8.3, 0.7 Hz, 1H),
8.03 (bs, 2H), 7.94 (ddd, J ) 7.1, 7.1, 1.3 Hz, 1H), 7.57 (ddd, J
) 8.6, 7.1, 1.3 Hz, 1H); 13C NMR (DMSO-d6) δ 151.4, 138.4,
135.4, 130.6, 126.9, 121.1, 117.0. MS (EI) m/z 178 (100), 162
(14), 136 (34), 78 (16), 77 (10), 76 (17).
lism occurs, the yield of the 4-oxide (4) decreases. As part
of this work we have performed the first unambiguous
synthesis and characterization of the heterocyclic N-oxide
4. Finally, the ability of this metabolite to contribute to
DNA damage by the antitumor agent tirapazamine is
considered.
Exp er im en ta l Section
Chemicals were purchased from the following suppliers and
were of the highest purity available: potassium acetate and
glycerol, Sigma Chemical Co.; tris(hydroxy)aminomethane,
xanthine, cyanamide, 2-nitroaniline, tert-butyl peracetate (in
mineral spirits), acetic anhydride, sodium hydrogen phosphate,
sodium dihydrogen phosphate, sodium hydrodisulfite (sodium
dithionite), thin-layer chromatography plates (silica gel, 250
µm thickness, 60 Å pore size), Aldrich Chemical Co.; hexane,
ethyl acetate, HPLC grade acetonitrile, methanol and 2-pro-
panol, hydrochloric acid, glacial acetic acid, 30% hydrogen
peroxide, sodium hydroxide, chromatographic silica gel-G type
60A, grade 633, Fisher; ethidium bromide, xanthine oxidase
(cat. no. 110442), catalase (cat. no. 106828), superoxide dis-
mutase (cat. no. 567680), pBR322 and pUC19, Roche Molec-
ular Biochemicals; SeaKem LE agarose, FMC bioproducts;
ethylenediaminetetraacetic acid, sodium dodecyl sulfate, United
States Biochemical Corp.; chloroform-d (w/0.05% v/v TMS),
dimethyl-d6 sulfoxide, Cambridge Isotope Laboratories Inc. and
desferal was a generous gift from CIBA. Ethyl acetate and
hexane were distilled before use. Pyrex brand standard boro-
silicate glass tubing was purchased from Ace Glass. Water for
HPLC and reactions was distilled, deionized, and glass redis-
tilled.
Syn th esis of 3-Am in o-1,2,4-ben zotr ia zin e 1-Oxid e (3).
Compound 3 was prepared by a modification of previous
methods.18,19 2-Nitroaniline (2.48 g; 18 mmol) and cyanamide
(1.51 g; 36 mmol) were heated with stirring at 100 °C until a
deep red melt formed. The reaction was cooled to room
temperature, and cold concentrated hydrochloric acid (6.5 mL,
78 mmol) was added dropwise. The reaction was then warmed
to 100 °C and stirred for 20 min, then cooled again to room
temperature. To the cooled reaction mixture NaOH was added
6.5 mL of a 16 M aqueous solution in 0.5 mL portions over
approximately 15 min. The resulting cloudy yellow-orange
mixture was warmed to 100 °C and stirred for 2 h. Addition
of water (∼25 mL) produced a yellow precipitate. The suspen-
sion was allowed to cool for 10 min, collected by vacuum
filtration, and then washed with water followed by ethyl
acetate to remove unreacted 2-nitroaniline (until the wash
changes from orange to light yellow). The bright yellow solid
collected from the filtration (1.07 g, 39% yield) was recrystal-
lized from 2-propanol to yield 3: (Rf 0.6 in 100% ethyl acetate;
a fluorescent yellow-green spot under long wave UV). 1H NMR
(CDCl3) δ 8.31 (dd, J ) 8.7, 0.9 Hz, 1H), 7.75 (ddd, J ) 7.7,
6.9, 1.4 Hz, 1H), 7.61 (dd, J ) 7.6, 1.0 Hz, 1H), 7.37 (ddd, J )
8.5, 7.0, 1.3 Hz, 1H), 5.13 (bs, 2H); 1H NMR (DMSO-d6) δ 8.14
(dd, J ) 8.6, 1.3 Hz, 1H), 7.77 (ddd, J ) 8.4, 6.9, 1.4 Hz, 1H),
7.52 (dd, J ) 7.8, 0.5 Hz, 1H), 7.33 (ddd, J ) 8.5, 7.0, 1.3 Hz,
2H), 7.31 (bs, 2H); 13C NMR (DMSO-d6) δ 160.2, 148.8, 135.7,
129.9, 125.8, 124.6, 119.8; MS (EI) m/z 162 (100), 118 (12), 91
(27), 90 (18); MS (ESI) m/z, 163 (100), 133 (12), 92 (20).
Syn th esis of 3-Am in o-1,2,4-ben zotr ia zin e (5). The ben-
zotriazine 5 was prepared via a modification of Mason and
Tennant’s procedure.18 To a solution of 3 (32 mg, 0.22 mmol)
in 70% ethanol-water (5 mL) was added sodium dithionite
(75 mg, 0.44 mmol). The resulting suspension was refluxed
for 30 min, an additional aliquot of sodium dithionite (38 mg,
0.22 mmol) added, and the suspension refluxed for another
30 min. One final addition of sodium dithionite (14 mg, 0.11
mmol) was added and the solution refluxed for another 30 min
at which time all starting material was consumed (as judged
by TLC). The solvent was removed by rotary evaporation with
warming and the resulting pale yellow residue purified by
column chromatography on silica gel (eluted with ethyl
acetate) to yield 16 mg (57%) of 5: Rf ) 0.77 (100% ethyl
acetate); 1H NMR (CDCl3) δ 8.32 (dd, J ) 8.4, 0.7 Hz, 1H),
7.79 (ddd, J ) 7.6, 7.7, 1.4 Hz, 1H), 7.62 (dd, J ) 1.2, 6.7 Hz,
1H), 7.51 (ddd, J ) 7.6, 6.9, 1.2 Hz, 1H), 5.61 (bs, 1H); 1H NMR
(DMSO-d6) δ 8.17 (dd, J ) 8.4, 1.4 Hz, 1H), 7.78 (ddd, J ) 7.4,
6.6, 1.5 Hz, 1H), 7.59 (sb, 2H), 7.52 (dd, J ) 8.6, 0.6 Hz, 1H),
7.45 (ddd, J ) 8.3, 6.9, 1.2 Hz, 1H); 13C NMR (DMSO-d6) δ
160.5, 142.1, 141.9, 135.6, 129.4, 125.8, 124.7; MS (EI) m/z 146
(41), 118 (100), 91 (46), 76 (20).
Syn th esis of 3-Am in o-1,2,4-ben zotr ia zin e 2-Oxid e (6).
Hydrogen peroxide (30%, 3.5 mL) was added to a solution of 5
(100 mg, 0.68 mmol) in glacial acetic acid (10 mL). Upon
addition of hydrogen peroxide, compound 5 dissolves and the
resulting solution was stirred for 24 h. The solvent was then
removed by lyophilization (behind a blast shield) and the
resulting brownish-yellow residue purified by flash column
chromatography on silica gel (eluted with 1:1 ethyl acetate:
hexane) to give 69 mg of 6 (63%): Rf ) 0.87 (100% ethyl
acetate); 1H NMR (CDCl3) 7.78 (d, J ) 8.2 Hz, 1H), 7.66-7.58
(m, 2H), 7.51 (t, J ) 7.1 Hz, 1H), 6.23 (bs, 2H) 1H NMR
(DMSO-d6) δ 8.15 (bs, 2H), 7.65-7.41 (m, 4H); 13C NMR
(DMSO-d6) δ 149.1, 135.5, 133.8, 130.8, 125.8, 124.2, 123.9.
MS (EI) m/z, 162 (61), 133 (11), 132 (100), 118 (13), 117 (33),
105 (81), 91 (23), 90 (64), 78 (23). Crystals of 6 suitable for
X-ray diffraction were obtained by slow evaporation of a 1:1
ethyl acetate-hexane solution. Crystal data for 6: monoclinic,
space group P21/n, a ) 5.07771(6) Å, b ) 18.681(2) Å, c )
(18) Mason, J . C.; Tennant, G. J . Chem. Soc. B 1970, 911-916.
(19) J iu, J .; Mueller, G. P. J . Org. Chem. 1959, 24, 813-818.

3-Amino-1,2,4-benzotriazine 4-Oxide
J . Org. Chem., Vol. 66, No. 1, 2001 109
7.8231(10) Å, â ) 104.719(2)°, V ) 717.62(15) Å3, Fcalcd ) 1.501
mg‚m-3, Mo KR radiation (λ ) 0.71070 Å) for Z ) 4. Intensity
data were collected on a Bruker SMART system at 173 K.
Least squares refinement based on 1588 reflections with Inet
> 2.0σ(Inet) (out of 4447 total reflections) gave a final R )
0.0418. All crystallographic calculations were conducted using
SHELX 97.
high vacuum (∼1 mmHg) and the resulting residue purified
by flash column chromatography on silica gel (eluted with ethyl
acetate) to yield 9 mg (90%) of 7 as a yellow powder: Rf ) 0.5
1
(100% ethyl acetate); H NMR (CDCl3) δ 9.54 (bs, 1H), 8.49
(d, J ) 8.5 Hz, 1H), 8.41 (d, J ) 8.4 Hz, 1H), 7.96 (ddd, J )
8.6, 7.6, 1.3 Hz, 1H), 7.83 (ddd, J ) 8.5, 7.7, 1.4 Hz, 1H), 2.74
(s, 3H); 1H NMR (DMSO-d6) δ 10.96 (bs, 1H), 8.50 (d, J ) 7.7
Hz, 1H), 8.35 (d, J ) 8.4 Hz, 1H), 8.08 (d, J ) 7.5 Hz, 1H),
7.97 (t, J ) 7.5 Hz, 1H).13C NMR (DMSO-d6) δ 169.1, 148.1,
146.0, 135.9, 134.8, 131.5, 129.9, 117.6, 24.4. Crystals of 7
suitable for X-ray diffraction were obtained by slow evapora-
tion of 2-propanol. Crystal data for 7: monoclinic, space group
P21/n, a ) 4.0178(4) Å, b ) 15.5538(14) Å, c )14.0228(13) Å,
â ) 93.004(2)°, V ) 875.11(14) Å3, Fcalcd ) 1.542 mg‚m-3, θrange
) 1.96-27.11°, Mo KR radiation (λ ) 0.71070 Å) for Z ) 4.
Intensity data were collected on a Bruker SMART system at
173 K. Least squares refinement based on 1916 reflections
with Inet > 2.0σ(Inet) (out of 5427 total reflections) gave a final
R ) 0.0496. All crystallographic calculations were conducted
using SHELX 97.
Ch a r a cter iza tion of P r od u cts Ar isin g fr om In Vitr o
Meta bolism of N-Oxid es (1). In a typical assay, a solution
of Tris buffer (pH 7, 25 mM, final volume 150 µL) containing
the appropriate N-oxide (250 µM) and xanthine in a Pyrex tube
(i.d. 3.0 mm, o.d. 5.0 mm, sealed at one end) was degassed by
three freeze-pump-thaw cycles and then sealed under vacuum.
The stock solution of xanthine (5 mM) was prepared by
dissolving the compound in 40% NaOH. Sealed tubes were
transferred to an inert atmosphere glovebox, scored, opened,
and transferred to eppendorf tubes. To these solutions xan-
thine oxidase (degassed in the glovebox by bubbling with
argon) was added, and the assays were incubated in the
glovebox protected from the light for 1 h at room temperature.
Xanthine oxidase was removed from the assay mixtures by
centrifugation (Brinkmann tabletop centrifuge, @10000 rpm)
through Amicon Microcon (YM3) filters. The filtrate was
analyzed by HPLC, employing a C-18 reverse-phase Rainin
Microsorb-MV column (100 Å sphere size, 5 µm pore size, 25
cm length, 4.6 mm i.d.) eluted with an isocratic mobile phase
composed of 1% acetic acid, 25% methanol, and 74% water at
a flow rate of 0.9 mL/min. The products were monitored by
UV absorbance (254 nm). The 4-oxide produced by in vitro
metabolism of tirapazamine was shown to comigrate with
authentic synthetic 4-oxide under three different HPLC condi-
tions: the conditions described above, as well as eluction of a
C-18 reverse-phase Rainin Microsorb-MV column column or
a Rainin Microsorb-MV Phenyl column (100 Å sphere size, 5
µm pore size, 25 cm length, 4.6 mm i.d.) with an isocratic
mobile phase of 1% acetic acid, 5% acetonitrile, 94% water.
Concentrations of various components of the reaction mixtures
were determined from peak area or peak height using calibra-
tion curves prepared by injection of known amounts of
authentic products.
Syn th esis of 3-Am in o-1,2,4-ben zotr ia zin e 4-Oxid e (4)
via Oxid a tion of th e Cor r esp on d in g Ben zotr ia zin e (5)
w ith Hyd r ogen P er oxid e. To a stirred solution of the
benzotriazine (5, 500 mg, 3.1 mmol) in glacial acetic acid (25
mL) was added 30% hydrogen peroxide (8 mL), and the
resulting solution then stirred, protected from the light at 50
°C for 2 h. The reaction mixture (containing starting material,
2-oxide 6, 1-oxide 3, 4-oxide 4, di-N-oxide 1, and an additional
uncharacterized product with an Rf ) 0.36 in 100% ethyl
acetate, blue fluorescent TLC spot under long wave UV) was
lyophilized (behind a blast shield) to produce an orange-yellow
solid. This solid was dissolved in ethyl acetate, mixed with
silica gel (3.5 g), and then evaporated under reduced pressure.
This silica gel mixture was then placed on the top of a silica
gel column (88 g) packed in ethyl acetate. Elution of the column
with ethyl acetate followed by 15:85 2-propanol:ethyl acetate
provided ∼50 mg of impure 4. A second step of column
chromatography of this material on silica gel (eluted with 75:
25 ethyl acetate:hexane, followed by 100% ethyl acetate)
afforded 40 mg (7% overall yield) of 4 as a mustard yellow
powder: Rf ) 0.49 (100% ethyl acetate, green fluorescent TLC
spot under long wave UV366); 1H NMR (CDCl3) δ 8.38 (dd, J )
5.9, 1.2 Hz, 1H), 8.35 (dd, J ) 8.3, 1.2 Hz, 1H), 7.88 (ddd, J )
7.9, 7.8, 1.2 Hz, 1H), 7.65 (ddd, J ) 7.8, 7.7, 1.2 Hz, 1H), 6.5
1
(bs, H2); H NMR (DMSO-d6) δ 8.30 (d, J ) 8.6 Hz, 1H), 8.2
(bs, 2H), 8.16 (d, J ) 8.4 Hz, 1H), 7.91 (ddd, J ) 8.6, 7.1, 1.2
Hz, 1H), 7.63 (ddd, J ) 8.4, 7.0, 1.2 Hz, 1H) 13C NMR (DMSO-
d6) δ 154.0, 142.7, 135.01, 129.6, 127.2, 127.1, 115.9; MS (EI)
m/z 162 (100), 146 (3), 117 (13), 116 (12), 90 (33), 77 (12); MS
(ESI) m/z 163.3 (20), 162.7 (33), 118.4 (27), 117.9 (100), 91.4
(32), 90.6 (38); UV-vis λmax 208 (ꢀ ) 9200), 256 (ꢀ ) 27000),
326 (ꢀ ) 3400), and 462 (ꢀ ) 3600) nm.
Syn th esis of 3-Am in o-1,2,4-ben zotr ia zin e 4-Oxid e (4)
by Rea ction of 3-Am in o-1,2,4-ben zotr ia zin e 1,4-Dioxid e
(1) w ith ter t-Bu tyl P er a ceta te. To a solution of 1 (50 mg,
0.28 mmol) in water (5 mL) was added tert-butyl peracetate
(268 µL, 50 wt % in mineral spirits, 0.84 mmol), and the stirred
mixture was refluxed in a 105-112 °C oil bath for 21 h. An
additional aliquot (50 µL, 0.15 mmol) of tert-butyl peracetate
was added and the reaction refluxed for 29 h. A final aliquot
of tert-butyl peracetate (50 µL, 0.15 mmol) was added to the
reddish-orange mixture and refluxing continued for 16 h. The
reaction was cooled and transferred to a test tube. The flask
was rinsed with 2-propanol and the combined solution centri-
fuged to remove a small amount (6 mg) of reddish-black solid
which had formed on the sides of the reaction flask. The
orange-red supernatant was lyophilized (behind a blast shield)
to produce a reddish-orange solid. The final reaction mixture
contained significant amounts of 4, 3, 3-amino-7-methyl-1,2,4-
benzotriazine 1-oxide, and starting material (1). Repeated flash
column chromatography as described above gave pure 4 (1.7
mg, 3.7%).
Syn th esis of 3-Am in o-1,2,4-ben zotr ia zin e 4-Oxid e (4)
by Rea ction of 3-Am in o-1,2,4-ben zotr ia zin e (5) w ith
m -Ch lor op er ben zoic Acid . The benzotriazine (5, 50 mg, 0.3
mmol) and m-chloroperbenzoic acid (196 mg, Aldrich 57-86%
pure) were dissolved in methylene chloride (5 mL) with stirring
at room temperature. After 2.5 h, the wine red solution was
evaporated under reduced pressure. Column chromatography
on silica gel (eluted with 75:25 ethyl acetate:hexane, followed
by 100% ethyl acetate) afforded 2 mg (4%) of 4. The reaction
also affords significant amounts of 1, 3, 6, and an additional
uncharacterized product (Rf ) 0.36 in 100% ethyl acetate).
Syn th esis of 3-Aceta m id o-1,2,4-ben zotr ia zin e 4-Oxid e
(7). The 4-oxide (4, 8 mg, 0.05 mmol) was dissolved in neat
acetic anhydride (0.7 mL, 7.4 mmol). After stirring at 45-50
°C for 3 h nearly all starting material was consumed as judged
by TLC. The acetic anhydride was removed by warming under
LC/MS a n d LC/MS/MS An a lysis of th e P r od u cts Re-
su ltin g F r om In Vitr o Meta bolism of Tir a p a za m in e (1).
Samples were prepared as described above, but on 1 mL scale
employing larger Pyrex tubes (i.d. 6.0 mm) for freeze-pump-
thaw-degassing. Samples were passed through Amicon filters
prior to analysis. LC/MS and LC/MS/MS experiments were
carried out on a Finnigan TSQ 7000 triple quadrupole instru-
ment interfaced to a ThermoSeparations liquid chromatograph
(TSP4000). HPLC was performed on a C-18 reverse-phase
Rainin Microsorb-MV column (100 Å sphere size, 5 µm pore
size, 25 cm length, 4.6 mm i.d.) at a flow rate of 1 mL/min
eluted with an isocratic mobile phase composed of 20%
methanol, 10% acetonitrile, 70% water. Positive ion electro-
spray was used as the means of ionization and collision-
induced dissociation (CID) involved argon gas (∼2 mTorr).
Other instrument settings included a capillary voltage of 4.5
kV, a capillary temperature of 350 °C, and a source temper-
ature of 75 °C.
In vestiga tion of th e DNA-Clea vin g P r op er ties of th e
3-Am in o-1,2,4-ben zotr ia zin e 4-Oxid e Meta bolite (4). All
assays were performed under anaerobic conditions and were
carried out in an inert atmosphere glovebox. In a typical assay,

110 J . Org. Chem., Vol. 66, No. 1, 2001
Fuchs et al.
a solution of sodium phosphate buffer (pH 7, 25 mM, final
volume 70 µL) containing the appropriate N-oxide (250 µM),
xanthine (250-500 µM), DNA (supercoiled pBR322 or pUC19,
0.6 µg), and desferal (1 mM) in a Pyrex tube (i.d. 3.0 mm, o.d.
5.0 mm, sealed at one end) was degassed by three freeze-
pump-thaw cycles and then sealed under vacuum. Sealed
tubes were transferred to an inert atmosphere glovebox,
scored, opened, and transferred to eppendorf tubes. To these
solutions were added superoxide dismutase (700 ng), catalase
(7 µg), and xanthine oxidase (0.04 units, each degassed in the
glovebox by bubbling with argon) and the assays incubated in
the glovebox protected from the light for 1 h at room temper-
ature. Following incubation, loading buffer (5 µL of 50%
glycerol loading buffer containing 0.1% bromophenol blue, 150
mM EDTA, 1% SDS in 2 M Tris, 1 M acetate, pH 8) was added
to the reactions and the resulting mixture loaded immediately
onto a 0.9% agarose gel containing 0.87 µg/mL ethidium
bromide. The gel was electrophoresed for approximately 4 h
at 80 V in TAE buffer (40 mM tris, 20 mM acetate, 1 mM
EDTA, pH 8). DNA in the gel was visualized by UV-transil-
lumination, the gel image recorded, and amounts of DNA in
each band quantitated using an Alpha Innotech IS-1000 digital
imaging system. The values reported are uncorrected for
differential ethidium staining of form I and II DNA.
Sch em e 2
Sch em e 3
existed for the putative tirapazamine metabolite 3-amino-
1,2,4-benzotriazine 4-oxide (4).
We prepared the suspected tirapazamine metabolite
3-amino-1,2,4-benzotriazine 4-oxide (4) by three inde-
pendent routes and fully characterized the compound by
NMR spectroscopy, UV-vis spectroscopy, LC/MS, and
X-ray crystallography. Treatment of tirapazamine with
methyl radical (H3C•) generated by thermolysis of tert-
butyl peracetate24,25 in water, followed by separation of
the resulting products by flash column chromatography,
affords the 4-oxide (4) in low yield (4%) (Scheme 2). An
alternate approach involving oxidation of 3-amino-1,2,4-
benzotriazine (5) with m-CPBA also provides a 4% yield
of the 4-oxide (Scheme 3). The most convenient route
involves treatment of the benzotriazine (5) with 30%
hydrogen peroxide in glacial acetic acid at 50 °C to afford
the 4-oxide in 7% yield (Scheme 3). The yields of 4 in
these reactions are low because the compound is pro-
duced as a relatively minor component of complex
mixtures that include the mono-N-oxides 3 and 6, the
benzotriazine 5, and tirapazamine (1).
The 1H and 13C NMR spectra of the 4-oxide (4) are
clearly distinct from that of the 1-oxide (3) and 2-oxide
(6) isomers. Consistent with the trends observed in the
proton NMR’s of tirapazamine (1) and its 1-oxide ana-
logue (3),18 the proton on the 5-position of the aromatic
ring in 4 (in the peri-position, adjacent to the N-oxide
oxygen) is shifted downfield relative to the same proton
in the benzotriazine derivative (5). Similarly, the mass
spectrum (EI) of the 4-oxide is distinct from that of the
isomeric mono-N-oxides 3 and 6. In addition to the
molecular ion peak at m/z 162, the 4-oxide shows
characteristic fragmentation products at m/z 134 (loss of
N2) and m/z 117 (loss of OH and N2). In contrast, the
1-oxide (3) shows a characteristic fragmentation product
at m/z 118 (loss of N2O) and the 2-oxide (6) fragments
with loss of NO to yield a peak at m/z 132. ESI-MS/MS
of the 4-oxide yields a characteristic fragment ion corre-
sponding to [M + H - N2 - OH]+ at m/z 118, whereas
the 1-oxide produces a major fragment corresponding to
[M + H - NO]+ at m/z 133. Initial confirmation of the
structure of the synthetic 4-oxide was provided by expos-
ing the compound to reaction conditions designed to yield
characteristic products. For example, this compound can
be oxidized to tirapazamine (1) by treatment with
Resu lts a n d Discu ssion
Syn th esis a n d Ch a r a cter iza tion of 3-Am in o-1,2,4-
ben zotr ia zin e 4-Oxid e (4). During the course of previ-
ous studies on the antitumor agent tirapazamine,6,20,21
we examined the products resulting from in vitro reduc-
tive metabolism of this drug. In these earlier studies, we
observed the expected15-17 two-electron reduction product
3-amino-1,2,4-benzotriazine 1-oxide (3) and the expected
four-electron reduction product 3-amino-1,2,4-benzotri-
azine (5). In addition, we were intrigued to observe
significant amounts of an uncharacterized product under
some conditions.21 We suspected that the unknown
product might be the “alternate” two-electron reduction
product, 3-amino-1,2,4-benzotriazine 4-oxide (4). Thus,
we prepared the putative 4-oxide metabolite (4) and
undertook experiments to determine whether this com-
pound is, in fact, produced by bioreductive metabolism
of the antitumor agent tirapazamine.
Interestingly, although tirapazamine (1) and its cor-
responding 2-oxide (6) have been known in the chemical
literature for over 40 years,22 and the corresponding
1-oxide (3) and benzotriazine (5) analogues known for
nearly 90 years,23 the 4-oxide analogue (4) has not
previously been characterized. In 1957, Robbins and co-
workers claimed22 to isolate the 4-oxide (4) from the
treatment of 3-amino-1,2,4-benzotriazine (5) with hydro-
gen peroxide at room temperature; however, based on
NMR data, the structure of this product was later called
into question and reassigned as the 2-oxide (6).18 Thus,
when we began this work, no unambiguous synthesis
(20) Hwang, J .-T.; Greenberg, M. M.; Fuchs, T.; Gates, K. S.
Biochemistry 1999, 38, 14248-14255.
(21) Daniels, J . S.; Gates, K. S.; Tronche, C.; Greenberg, M. M.
Chem. Res. Toxicol. 1998, 11, 1254-1257.
(24) Zady, M. F.; Wong, J . L. J . Org. Chem. 1980, 45, 2373-2377.
(25) Beckwith, A. L. J .; Zavitsas, A. A. J . Am. Chem. Soc. 1986, 108,
8230-8236.
(22) Robbins, R. F.; Schofield, K. J . Chem. Soc. 1957, 3186-3194.
(23) Arndt, F. Chem. Ber. 1913, 46, 3522-3530.

3-Amino-1,2,4-benzotriazine 4-Oxide
J . Org. Chem., Vol. 66, No. 1, 2001 111
Sch em e 4
thine/xanthine oxidase enzyme system under hypoxic
conditions reveals the presence of xanthine and uric acid
(eluting at ∼3 min), the well-known 1-oxide and benzo-
triazine metabolites (3 and 5), and two additional previ-
ously uncharacterized compounds eluting from the HPLC
at approximately 8 and 10 min (Figure 1). No significant
metabolism of tirapazamine is observed under aerobic
conditions, indicating that all of the products seen under
hypoxic conditions are formed via the oxygen-sensitive
radical 2 (Scheme 1).
hydrogen peroxide in acetic acid26 and is reduced to the
corresponding benzotriazine 5 by treatment with sodium
dithionite (data not shown).18
The identity of the unknown tirapazamine metabolite
eluting at ∼8 min in our HPLC analysis was established
by several experiments. Consistent with our initial
suspicion that the compound was the heterocyclic mono-
N-oxide 4, we found that the substance decomposes upon
treatment of the crude reaction mixture with sodium
dithionite, a reagent known to readily reduce N-oxides.
Evidence for the exact identity of the unknown metabolite
was provided by the observation that the HPLC retention
time of this compound matches that of authentic, syn-
thetically prepared 4-oxide (4) under several different
HPLC conditions. Coinjection of authentic material with
the reaction mixture obtained from in vitro metabolism
of 1 under anaerobic conditions conclusively showed that
the ∼8 min peak comigrates with the 4-oxide (4). LC/
MS analysis of the reaction mixture provided confirma-
tion that the new tirapazamine metabolite observed in
our HPLC traces is 3-amino-1,2,4-benzotriazine 4-oxide
(4). LC/ESI-MS of the 8 min peak yielded an ion corre-
sponding to [M + H + CH3CN]+ at m/z 204. Fragmenta-
tion of this ion was examined by LC/ESI-MS/MS using
CID. In this experiment, we observed the characteristic
fragment ion of the 4-oxide corresponding to [M + H -
N2 - OH]+ at m/z 118. The LC/MS properties of the 8
min metabolite are identical to those of authentic,
synthetically prepared 4-oxide (4). Additional studies
showed that the 4-oxide metabolite (4) is also produced
(along with the other expected metabolites 3 and 5) by
in vitro metabolism of tirapazamine by NADPH:cyto-
chrome P450 oxidoreductase under hypoxic conditions
(data not shown). The identity of the metabolite eluting
at ∼10 min remains under investigation.
The structure of the synthetically prepared 4-oxide (4)
was ultimately confirmed by X-ray crystallography. The
4-oxide readily crystallizes from 2-propanol, but the
resulting material did not yield a high-resolution X-ray
structure due to disorder in the crystals. The acetylated
derivative 3-acetamido-1,2,4-benzotriazine 4-oxide (7)
was prepared by treatment of 4 with neat acetic anhy-
dride (Scheme 4) in the hopes that this analogue would
yield a high quality X-ray structure. As an interesting
aside, we find that acetylation of the exocyclic amino
group in 4 is quite efficient relative to the same reaction
on the 1-oxide isomer (3). This finding is reminiscent of
the previous observation that 2-aminopyridine N-oxide
undergoes N-acetylation much more rapidly than 3-ami-
nopyridine N-oxide.27 In this earlier study, neighboring
group assistance involving initial acetylation of the
N-oxide oxygen, followed by intramolecular acetyl-group
transfer to the adjacent amino group was invoked to
explain the relatively rapid N-acetylation of 2-aminopy-
ridine N-oxide. Happily, 7 crystallizes from 2-propanol,
and the resulting crystals provided a high-resolution
X-ray structure which confirms that this compound is the
acetylated 4-oxide derivative (see Supporting Information
for X-ray data). The structure of the 2-oxide (6) was also
unambiguously confirmed by X-ray analysis.
Ch a r a cter iza tion of 3-Am in o-1,2,4-ben zotr ia zin e
4-Oxid e (4) P r od u ced by En zym a tic Activa tion of
Tir a p a za m in e (1). The identity of the enzymes respon-
sible for in vivo reductive activation of tirapazamine
remains under investigation.15 NADPH:cytochrome P450
reductase is thought to play a major role,15 but it has
also been suggested that DNA damage by the drug may
be primarily triggered by an as yet uncharacterized
reductase located in the cell nucleus.28
In the experiments described here, we have primarily
utilized the xanthine/xanthine oxidase enzyme system
to investigate reductive metabolism of tirapazamine.
Although xanthine oxidase may not play a central role
in the in vivo metabolism of tirapazamine, this enzyme
system provides a reasonable model system for studying
the reductive activation of tirapazamine. In support of
this notion, the xanthine/xanthine oxidase system is able
to initiate DNA cleavage by the drug6,8,29 and generates
the same primary tirapazamine metabolites (i.e., 3 and
5) that are produced by other in vitro and in vivo
systems.7,8,16,17,30
Yield of 3-Am in o-1,2,4-ben zotr ia zin e 4-Oxid e (4)
fr om in Vitr o Meta bolism of Tir a p a za m in e u n d er
Va r iou s Con d it ion s. In vitro metabolism of tira-
pazamine (1) by xanthine/xanthine oxidase produces
significant quantities of 3-amino-1,2,4-benzotriazine 4-ox-
ide (4). For example, incubation of tirapazamine (250 µM)
with xanthine oxidase (0.04 units) and xanthine (250 µM)
under hypoxic conditions in aqueous buffer (tris, pH 7)
yields a 1:1.5:6 ratio of 4, 5, and 3, respectively. Under
these conditions, significant amounts (15-30%) of tira-
pazamine remain unmetabolized. Interestingly, under
conditions where greater amounts of tirapazamine are
consumed (e.g., when higher concentrations of xanthine
are employed), yields of the 4-oxide metabolite decrease
(Figure 2). This finding can be explained by the observa-
tion that the 4-oxide (4) produced by metabolism of
tirapazamine readily undergoes further xanthine/xan-
thine oxidase-mediated reduction to 3-amino-1,2,4-ben-
zotriazine (5). In contrast, the 1-oxide (3) is relatively
resistant to further reduction by xanthine/xanthine oxi-
dase. For instance, treatment of the 1-oxide (3) and the
4-oxide (4) (250 µM) with xanthine oxidase (0.04 units)
and xanthine (250 µM) under anaerobic conditions in tris
buffer (pH 7) affords 5% and 74% yields of 3-amino-1,2,4-
Reverse-phase HPLC analysis of the products resulting
from in vitro metabolism of tirapazamine by the xan-
(26) This result does not provide unambiguous evidence that the
product is the 4-oxide (4) because, in the parent benzotriazine system,
the 2-oxide is also known22 to isomerize during oxidation to afford the
1,4-dioxide.
(27) Deady, L. W.; Stanborough, M. S. Aust. J . Chem. 1982, 35,
1841-1849.
(28) Evans, J . W.; Yudoh, K.; Delahoussaye, Y. M.; Brown, J . M.
Cancer Res. 1998, 58, 2098-2101.
(29) J ones, G. D. D.; Weinfeld, M. Cancer Res. 1996, 56, 1584-1590.

112 J . Org. Chem., Vol. 66, No. 1, 2001
Fuchs et al.
F igu r e 1. Reverse-phase HPLC analysis of the metabolites resulting from reductive activation of tirapazamine by the xanthine/
xanthine oxidase enzyme system.
aqueous buffered solution of the 4-oxide stored in a clear
glass container under ambient fluorescent lab lighting
is completely degraded after two weeks (data not shown).
It is perhaps worth noting, however, that the unidentified
compound eluting at ∼10 min in our experiments (Figure
1) is not a photolytic breakdown product of the 4-oxide,
but is actually a product resulting from bioreductive
metabolism.
In vestiga tion of th e DNA-Clea vin g P r op er ties of
th e 3-Am in o-1,2,4-ben zotr ia zin e 4-Oxid e Meta bolite
(4). Tirapazamine derives its biological activity from
redox-activated DNA damage; thus, it is important to
characterize the ability of the drug and its metabolites
to react with DNA. In vitro DNA cleavage by tira-
pazamine has previously been investigated,6,8,29 and it
was found that, in conjunction with the xanthine/xan-
F igu r e 2. Metabolism of tirapazamine (1) by xanthine/
thine oxidase enzyme system, under hypoxic conditions,
xanthine oxidase: yields of 1-oxide (3), 4-oxide (4), benzotri-
the drug yields significant amounts of single strand
breaks.6 The 1-oxide metabolite 3 causes no detectable
azine (5), and tirapazamine (1) as a function of xanthine
concentration. Assays were performed as described in the
Experimental Seciton.
DNA strand cleavage alone and only small amounts of
redox-activated strand scission.6 Similarly, previous stud-
ies have shown that the benzotriazine metabolite 5 does
not cleave DNA alone or in conjunction with enzymatic
activation.6
benzotriazine (5), respectively. Unlike the enzymatic
reduction of tirapazamine and its 1-oxide analogue (3),
xanthine/xanthine oxidase-mediated reduction of the
4-oxide (4) occurs with reasonable efficiency even under
aerobic conditions. Specifically, in analogous experiments
under conditions of ambient oxygenation, 8-14% of the
4-oxide (4) is consumed, whereas <1% of the 1-oxide (3)
is converted to benzotriazine (5).
In the absence of enzymatic reduction, the 4-oxide
metabolite is stable in solution. For example, a sample
of the 4-oxide stored in buffered aqueous solution (sodium
phosphate, pH 7) at room-temperature protected from
light remains unchanged (as judged by 1H NMR and
TLC) over the course of eight weeks. However, the
4-oxide, like many N-oxides, is photosensitive.31,32 An
The ability of the newly characterized 4-oxide metabo-
lite (4) to cleave double-stranded plasmid DNA was
investigated using an agarose gel electrophoresis assay.
This assay allows a quantitative assessment of spontane-
ous DNA strand cleavage by measuring the conversion
of uncut supercoiled (form I) plasmid DNA to the nicked
form (form II). As demonstrated previously,6,8,29 we find
that incubation of tirapazamine with the xanthine/
xanthine oxidase enzyme system under hypoxic condi-
tions yields significant amounts of single-strand DNA
cleavage (Table 1). In contrast, the 4-oxide (4) does not
produce significant DNA cleavage under these condi-
tions.33 Similarly, incubation of the 4-oxide with plasmid
(30) Walton, M. I.; Workman, P. Biochem. Pharmacol. 1990, 39,
1735-1742.
(31) Daniels, J . S.; Chatterji, T.; MacGillivray, L. R.; Gates, K. S. J .
Org. Chem. 1998, 63, 10027-10030.
(33) Similarly, the 4-oxide does not cause significant DNA cleavage
when incubated in the presence of the NADPH:cytochrome P450
oxidoreductase reducing system.
(32) Fuchs, T.; Gates, K. S.; Hwang, J .-T.; Greenberg, M. M. Chem.
Res. Toxicol. 1999, 12, 1190-1194.

3-Amino-1,2,4-benzotriazine 4-Oxide
J . Org. Chem., Vol. 66, No. 1, 2001 113
Ta ble 1. In vestiga tion of th e DNA-Clea vin g P r op er ties
of 3-Am in o-1,2,4-ben zotr ia zin e 4-Oxid e (4)
formed by enzymatic processing of tirapazamine and may
provide some insight regarding the identity of the missing
mass balance. Quantitative HPLC analysis of our reac-
tions indicates ∼90% of the starting material balance in
our reactions can be accounted for in unchanged tira-
pazamine (1), 1-oxide (3), 4-oxide (4), and benzotriazine
(5).
reactiona
% form I % form II
S valueb
anaerobic; DNA alone
anaerobic; DNA + X/XO
anaerobic; 250 µM 1 + X/XO
anaerobic; 250 µM 4 + X/XO
anaerobic; 250 µM 4 + XO
anaerobic; 250 µM 4 alone
a
72
71
16
67
77
70
28
29
84
33
23
30
0.33 ((0.15)
0.34 ((0.11)
1.80 ((0.49)
0.40 ((0.25)
0.26 ((0.11)
0.35 ((0.13)
There are several chemical pathways by which the
4-oxide metabolite (4) can be produced. First, one-electron
reduction of tirapazamine to the delocalized radical (2)
may yield the metabolite 4 via a mechanism analogous
to those proposed previously6,8,16 for the formation of the
metabolite 3. Alternatively, it is possible that the 4-oxide
metabolite (4) is formed by two sequential one-electron
enzymatic reductions of the drug (1). Finally, we have
previously shown that the 4-oxide (4) is formed by the
reaction of tirapazamine with carbon-centered radicals.20
Accordingly, a portion of the observed 4-oxide metabolite
in the current studies may arise from reaction of the
parent drug, tirapazamine, with carbon-centered radicals
that are generated by the action of the reductively
activated drug on organic substrates in the reaction
mixture. In the reactions reported here, carbon-centered
radicals may be derived from the organic buffer, Tris,
xanthine oxidase, xanthine, or uric acid. When DNA is
present, deoxyribose radicals resulting from the action
of the activated drug on the biopolymer will react with
the parent drug to produce the 4-oxide metabolite.20,21
The 4-oxide metabolite characterized here does not
cause efficient DNA strand cleavage either alone or in
conjunction with reductive activation by the xanthine/
xanthine oxidase enzyme system. The metabolite (4)
readily undergoes reduction by xanthine/xanthine oxi-
dase; however, we suggest that enzymatic processing of
4 does not produce DNA damage due to the fact that the
compound is converted to the benzotriazine (5) either via
direct two-electron enzymatic reduction or two sequential
enzymatic one-electron transfers, rather than by one-
electron reduction followed by N-O bond fragmentation
to produce hydroxyl radical (analogous to the reaction of
2 shown in Scheme 1). Although 4 does not cleave DNA,
this compound may nonetheless play a significant indi-
rect role in facilitating strand cleavage by tirapazamine.
We previously demonstrated that the 4-oxide (4) ef-
ficiently transfers the oxygen from its N-oxide functional
group to a deoxyribose radical in double-stranded DNA
with a rate constant of 3.1 × 107 M-1 s-1, thus resulting
in net conversion of the DNA radical into a base-labile
strand break.20 Typically, DNA strand break formation
by radicals is inefficient under hypoxic conditions because
molecular oxygen is required for the efficient conversion
of DNA radicals into strand breaks.36-4136-42 However, the
4-oxide (4) acts as an oxygen surrogate and, thus, may
serve to increase the overall efficiency of drug-mediated
DNA strand cleavage under hypoxic conditions. Further
Cleavage reactions contained supercoiled pBR322 or pUC19
DNA (37 µM bp) in a solution of sodium phosphate buffer (25 mM,
pH 7.0), xanthine (250-500 µM), desferral (1 mM), xanthine
oxidase (0.04 U), catalase (7 µg), and superoxide dismutase (700
ng). Reactions solutions were freeze-pump-thaw degassed three
times, enzymes were added, and the reactions were incubated for
1 h at 25 °C in an argon-filled glovebox. Reactions and densito-
metry were performed as described in the Experimental Section.
b
S is the mean number of strand breaks per plasmid molecule
and is calculated using the equation: S ) -ln(% form I DNA).42
DNA in the absence of the xanthine/xanthine oxidase
system produces no detectable cleavage.34
Con clu sion s
We have unambiguously shown that 3-amino-1,2,4-
benzotriazine 4-oxide (4) is produced by in vitro reductive
metabolism of the antitumor agent tirapazamine. Under
conditions where tirapazamine (1) is bioreductively ac-
tivated, the yield of this metabolite is typically moderate
because the 4-oxide that is produced readily undergoes
further reduction to the benzotriazine analogue (5,
Scheme 1). On the other hand, the isomeric two-electron
reduction product 3-amino-1,2,4-benzotriazine 1-oxide (3)
is not readily reduced by the xanthine/xanthine oxidase
enzyme and, therefore, is found as a major metabolite
resulting from bioreductive processing of tirapazamine
regardless of the reaction conditions. Together, these
results suggest that the 3-amino-1,2,4-benzotriazine (5)
product which has long been known as a metabolite of
tirapazamine may be formed predominantly via reduction
of the 4-oxide (4).
Although the 4-oxide metabolite (4) described here has
not previously been characterized, an inspection of the
literature indicates that this compound may have previ-
ously been seen as a product resulting from the metabo-
lism of tirapazamine under several different condi-
tions.8,17,30,35 For example, a compound with similar
relative retention time on C18-reverse phase HPLC was
produced by in vivo metabolism of tirapazamine in CHO
cells and in mice.17,35 Similarly, enzymatic and radiolytic
activation of the drug afforded an uncharacterized
productslikely the 4-oxidesthat migrated between the
drug (1) and the benzotriazine (5) on reverse-phase
HPLC.8,30 We find that the 4-oxide metabolite is produced
by hypoxic metabolism of tirapazamine by the NADPH:
cytochrome P450 reductase enzyme system as well as the
xanthine/xanthine oxidase enzyme system.
Previous researchers who have analyzed the products
resulting from metabolism of tirapazamine have noted
that significant amounts (30-60%) of the starting mass
balance cannot be accounted for in the remaining starting
material, the 1-oxide (3), and the benzotriazine (5)
metabolites.8,9,15,30 Our characterization of the metabolite
4 helps provide a more complete picture of the products
(36) Lennartz, M.; Coquerelle, T.; Hagen, U. Int. J . Radiat. Biol.
1973, 24, 621-625.
(37) Lucke-Huhle, C.; Braun, A.; Hagen, U. Z. Naturforsh. 1970,
25b, 1264-1268.
(38) Greenberg, M. M. Chem. Res. Toxicol. 1998, 11, 1235-1248.
(39) Breen, A. P.; Murphy, J . A. Free Rad. Biol. Med. 1995, 18,
1033-1077.
(40) Pratviel, G.; Bernadou, J .; Meunier, B. Angew. Chem., Int. Ed.
Engl. 1995, 34, 746-769.
(34) Enzyme-activated DNA cleavage by 4 is not above background
even when considered at the 50% confidence level.
(35) Zamalloa, E.; Aldana, I.; Dios, C.; Monge, A. Bol. Soc. Quim.
Peru 1998, 64, 109-125.
(41) Pogozelski, W. K.; Tullius, T. D. Chem. Rev. 1998, 98, 1089-
1107.
(42) Hertzberg, R. P.; Dervan, P. B. Biochemistry 1984, 23, 3934-
3945.

114 J . Org. Chem., Vol. 66, No. 1, 2001
Fuchs et al.
Su p p or tin g In for m a tion Ava ila ble: Complete X-ray
data for 6 and 7 and 1H NMR spectra for 1, 3, 4, 5, 6, and 7.
This material is available free of charge via the Internet at
http://pubs.acs.org.
investigations of the chemical and biological properties
of this newly characterized tirapazamine metabolite are
currently underway.
Ack n ow led gm en t. We thank Monsanto Corpora-
tion and the American Cancer Society (RPG-00-028-01)
for financial support of this work.
J O001232J