Sulfoxide-containing aromatic nitrogen mustards as hypoxia-directed bioreductive cytotoxins

Article abstract of DOI:10.1021/jm9904957

A series of diaryl and alkylaryl sulfoxide-containing nitrogen mustards were synthesized and evaluated for their hypoxia-selective cytotoxicity against V-79 cells in vitro as well as for their metabolism profiles with the rat S-9 fractions. In general, th

Full text of DOI:10.1021/jm9904957

4160
J . Med. Chem. 2000, 43, 4160-4168
Su lfoxid e-Con ta in in g Ar om a tic Nitr ogen Mu sta r d s a s Hyp oxia -Dir ected
Bior ed u ctive Cytotoxin s
Zhong-Yue Sun, Emad Botros, Ai-Duen Su, Yu Kim, Enju Wang,† Nesrine Z. Baturay, and Chul-Hoon Kwon*
Department of Pharmaceutical Sciences, College of Pharmacy and Allied Health Professions, and Department of Chemistry,
School of Arts and Sciences, St. J ohn’s University, J amaica, New York 11439
Received September 29, 1999
A series of diaryl and alkylaryl sulfoxide-containing nitrogen mustards were synthesized and
evaluated for their hypoxia-selective cytotoxicity against V-79 cells in vitro as well as for their
metabolism profiles with the rat S-9 fractions. In general, the diaryl sulfoxides (4, 5, and 7-9)
showed much greater hypoxia selectivity (11-27-fold) than the alkylaryl sulfoxides (∼3-fold)
(1 and 3). The fused diphenyl sulfoxides (10 and 11), on the other hand, showed very low hypoxia
selectivity (1.3-3-fold). Compound 10 was highly cytotoxic under both aerobic and anaerobic
conditions, while 11 showed low cytotoxicity under both conditions. The bioreduction of 8 by
the rat S-9 fraction under anaerobic conditions was inhibited by menadione and enhanced by
benzaldehyde, acetaldehyde, or 2-hydroxypyrimidine suggesting the involvement of aldehyde
oxidase in the reduction of the sulfoxides. Bioreductive metabolism studies of selected model
sulfoxides suggested that diaryl sulfoxides are better substrates for aldehyde oxidase than
alkylaryl sulfoxides.
In tr od u ction
matory activity of the parent drug.14 The reduction of
sulfoxides by mammalian systems can be catalyzed by
aldehyde oxidase15-18 and the thioredoxin enzyme sys-
tem.15,19 Under anaerobic conditions, aldehyde oxidase
plus electron donors, or a combination of aldehyde
oxidase and xanthine oxidase plus xanthine, catalyzes
the reduction of sulfoxides (e.g., sulindac, diphenyl
sulfoxide, phenothiazine sulfoxide, and dibenzyl sulfox-
ide) to their corresponding sulfides.16-18 The sulfoxide-
reducing activity of aldehyde oxidase is enhanced by the
addition of electron donors, such as aldehydes or 2-hy-
droxypyrimidine.18 The specificity of enzyme for sulfox-
ide reduction varies with sulfoxide structures. Both rat
and rabbit liver aldehyde oxidases or the renal thiore-
doxin system, for example, can reduce sulindac. Diphe-
nyl sulfoxide, however, is mainly reduced by aldehyde
oxidase.15
We previously reported that 4-[bis(2-chloroethyl)-
amino]-1-(methylsulfinyl)benzene (1), a sulfoxide-con-
taining nitrogen mustard, showed some hypoxia-selec-
tive cytotoxicity against the Chinese hamster ovary
(CHO) cell line.20 It has been proposed that the cyto-
toxicity be generated by the more reactive sulfide 2
formed from 1 upon bioreduction as shown in Scheme
1.21 An in vitro study with the rat S-9 fraction showed
that 1 was bioreduced to 2 more significantly under
anaerobic conditions in the presence of a NADPH-
generating system. This suggested a possible involve-
ment of the NADPH-dependent cytochrome P-450 re-
ductase, which has been shown to catalyze the reduction
of nitromin to nitrogen mustard,22 in the reduction of
the sulfoxide to the sulfide. These results provided the
first evidence supporting the concept of enhanced biore-
duction of the sulfoxide to the sulfide under hypoxic
conditions.
Solid tumors contain areas of low oxygen tension
(hypoxia) generally thought to arise in solid tumors due
to a poor and disorganized blood supply.1 Although these
microenvironmental properties induce tumor resistance
to radiation therapy and to many conventional chemo-
therapeutic agents, an opportunity exists to exploit the
hypoxic environment of solid tumors for targeting
cytotoxic anticancer agents. Approaches to the use of
bioreductive agents, which are selectively toxic to hy-
poxic cells upon enzymatic reduction, in the treatment
of solid tumors continue to develop steadily.1-5 Much
of the work has been centered on nitroheterocyclics (e.g.,
RSU 1069), quinones (e.g., mitomycin C), and di-N-
oxides (e.g., tirapazamine, SR4233). Bioreductive acti-
vation of these hypoxia-selective cytotoxins can be
catalyzed by a number of enzymes such as cytochrome
P-450, cytochrome P-450 reductase, DT-diaphorase
(quinone-oxidoreductase), xanthine oxidase, and alde-
hyde oxidase.6-10 It appears that several different
enzymes participate to different extents with the various
bioreductive agents.11 Workman11 suggests a distinct
possibility that one can design a successful bioreductive
drug possessing a promiscuous relationship with several
reductases or tailored for a particular enzyme that
might be hyperexpressed in a particular tumor.
Sulfoxides are known to be susceptible to bioreduc-
tion. Sulindac, a sulfoxide-containing nonsteroidal an-
tiinflammatory drug, can undergo reduction to the
sulfide or oxidation to the sulfone.12 The oxidation is
the dominant process under normal physiological condi-
tions; however, the reduction process becomes signifi-
cant under anaerobic conditions.13 The reductive me-
tabolite sulfide is largely responsible for the antiinflam-
In this report, a series of sulfoxide-containing alky-
lating agents were synthesized (Chart 1) and evaluated
for their hypoxia-selective cytotoxicity. The correspond-
* To whom correspondence should be addressed. Tel: 718-990-5214.
Fax: 718-990-6551. E-mail: kwonc@stjohns.edu.
† Department of Chemistry.
10.1021/jm9904957 CCC: $19.00 © 2000 American Chemical Society
Published on Web 09/23/2000

Sulfoxide-Containing Aromatic Nitrogen Mustards
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4161
Sch em e 1
Sch em e 2a
Ch a r t 1
ing sulfides and sulfones of some of these sulfoxides
were also synthesized for comparison purposes. In
addition, several model sulfoxides were studied to
understand the relationship between the structures and
their chemical/biochemical reduction properties.
a
Reagents: (a) ClCH2CH2OH, CaCO3 for 14, 17, and 20; (b)
POCl3; (c) H2O2, TFA; (d) ethylene oxide, AcOH for 25; (e) H2O2,
(NH4)2MoO4.
Sch em e 3a
Ch em istr y
The (2-chloroethyl)aminophenyl- (22) and [bis(2-chlo-
roethyl)]aminophenyl-substituted sulfides (14, 17, 25,
20) and [bis(2-chloroethyl)]amino-substituted diben-
zothiophenes (32, 34) were synthesized by hydroxyethy-
lation of the corresponding aminophenyl sulfides with
ethylene oxide23 or with 2-chloroethanol21 followed by
chlorination of the hydroxyl group (21, 13, 16, 24, 19,
31, 33) with phosphoryl chloride (Schemes 2, 4). The
corresponding sulfoxides (3-7, 10, 11) were prepared
by oxidizing the sulfides with hydrogen peroxide in
trifluoroacetic acid (TFA)24 or with m-chloroperbenzoic
acid (mCPBA)25 (Schemes 2, 4). The sulfone (28) was
prepared by oxidation of the corresponding sulfide or
sulfoxide with hydrogen peroxide using ammonium
molybdate as catalyst26 (Scheme 2).
a
Reagents: (a) Fe, HCl; (b) 9-chloroacridine, catalytic HCl.
Sch em e 4a
The amino compounds were synthesized from the
corresponding nitro compounds by reduction either with
Sn/HCl27 (12, 15) or with Fe/HCl (8, 26, 29) (Scheme
3). The (acridin-9-ylamino)phenyl [bis(2-chloroethyl)-
amino]phenyl sulfide (27), sulfoxide (9), and sulfone (30)
were synthesized by condensation of 9-chloroacridine
with 26, 8, and 29, respectively (Scheme 3).28
Resu lts a n d Discu ssion
A selected series of substituted sulfinyl aryl nitrogen
mustards (3-11) were synthesized as potential biore-
ductive cytotoxins. The cytotoxicity of aromatic nitrogen
mustards is very dependent on the electronic properties
of the substituents on the aromatic ring.21,29 Both the
parent sulfoxide nitrogen mustards and the oxidative
product sulfones should be relatively unreactive due to
electron-withdrawing effects of the substituents. On the
other hand, the bioreductive product sulfides should be
more reactive and provide higher cytotoxicity. To verify
this, the cytotoxicity profile of selected sulfoxide com-
a
Reagents: (a) ClCH2CH2OH, CaCO3; (b) POCl3; (c) H2O2, TFA
for 10; (d) mCPBA for 11.
pounds (7-9) was compared with that of the corre-
sponding sulfides and sulfones. The comparative cyto-
toxicity was determined by SRB assay against Chinese
hamster lung (V-79) cells under normal conditions
(Table 1). The sulfide compounds (25-27), which do not
require bioactivation for their cytotoxicity, were 10-14-
fold more cytotoxic than their corresponding sulfoxides;

4162 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Sun et al.
Ta ble 1. Relative Aerobic Cytotoxicities of
4-({4-[Bis(2-chloroethyl)amino]phenyl}sulfinyl)-1-nitrobenzene
(7), 4-({4-[Bis(2-chloroethyl)amino]phenyl}sulfinyl)phenylamine
(8), and 4-{[4-(Acridin-9-ylamino)phenyl]sulfinyl}-1-
[bis(2-chloroethyl)amino]benzene Hydrochloride (9) and Their
Corresponding Sulfides and Sulfones Against V-79 Cells
IC50
relative
compd
(µM)a
cytotoxicityb
25 (the corresponding sulfide of 7)
7 (the sulfoxide)
28 (the corresponding sulfone of 7)
26 (the corresponding sulfide of 8)
8 (the sulfoxide)
29 (the corresponding sulfone of 8)
27 (the corresponding sulfide of 9)
9 (the sulfoxide)
2.25
21.88
26.73
2.49
30.03
26.90
1.93
1
0.1
0.08
1
0.08
0.09
1
26.20
22.24
0.07
0.09
F igu r e 1. Survival of V-79 cells following treatment with 3
and 7. Cells were subjected to aerobic or hypoxic conditions
for 3 h. Survival was measured by using a clonogenic assay:
(O) aerobic, (0) hypoxic. The data points represent the mean
( SE of at least triplicate determinations for each of three
separate experiments.
30 (the corresponding sulfone of 9)
a
IC50
: concentration (µM) of drug required to reduce cell
survival to 50% of controls in a SRB assay from mean of triplicate
determinations of one experiment. Relative cytotoxicity is ex-
pressed as the ratio of the IC50 values of the corresponding sulfide
to sulfoxide or to sulfone.
b
fold). This high selectivity, however, stemmed primarily
from the low cytotoxicity under aerobic conditions, and
there was no further enhancement of cytotoxicity under
hypoxic conditions. It is unclear whether the acridine
moiety provided any contribution to the overall cyto-
toxicity profile through a possible DNA intercalation.
Compound 6 was neither hypoxia-selective nor signifi-
cantly cytotoxic. This was consistent with its structural
inability to cross-link DNA upon bioreduction. The fused
diphenyl sulfoxides, dibenzothiophen-5-one nitrogen
mustards (10, 11), showed low hypoxia selectivity (1.3-
3-fold). Compound 10 was fairly cytotoxic under both
aerobic (IC90 210 µM) and anaerobic (IC90 161 µM)
conditions. On the other hand, 11 showed much lower
cytotoxicity under both aerobic (IC90 4673 µM) and
anaerobic (IC90 1522 µM) conditions.
It has been reported that the enzyme systems in-
volved in sulfoxide reduction apparently show different
substrate specificity.15,31 To further understand the
relationship between the hypoxia selectivity and the
sulfoxide structures, selected model sulfoxide com-
pounds such as methylphenyl sulfoxide, diphenyl sul-
foxide, and dibenzothiophen-5-one were studied for their
metabolism with emphasis on bioreduction. Selected
sulfoxide nitrogen mustards were also included in the
study for comparison purpose. The chemical reduction
potential data of the model sulfoxides indicated (Table
3) that dibenzothiophen-5-one was the most easily
reduced, followed by diphenyl sulfoxide and methylphe-
nyl sulfoxide. On the other hand, diphenyl sulfoxide was
the most easily bioreduced by the rat S-9 fraction,
followed by methylphenyl sulfoxide and dibenzothiophen-
5-one (Table 3). Dibenzothiophen-5-one was poorly
reduced by aldehyde oxidase despite its low chemical
redox potential. Diphenyl sulfoxide was further studied
for its bioreduction profiles in detail (Table 3). Under
aerobic conditions, the formation of diphenyl sulfone
was observed only in the presence of NADPH, suggest-
ing that NADPH-dependent enzymes were responsible
for the oxidation. Under anaerobic conditions, diphenyl
sulfoxide was reduced to diphenyl sulfide in the pres-
ence of benzaldehyde, 2-hydroxypyrimidine, or acetal-
dehyde. On the other hand, NADPH did not appear to
be involved in catalyzing this reduction. In addition, the
reduction was inhibited by menadione, a specific alde-
Ta ble 2. Hypoxia-Selective Cytotoxicity of
Sulfoxide-Containing Aromatic Nitrogen Mustards
IC90 (µM)a
IC90 (µM)a
hypoxic
hypoxic
compd air
N2 selectivityb compd air
N2 selectivityb
1
3
4
5
6
944 340
3
3
11
17
7
8
9
10
11
3680 182
20
15
27
1.3
3
953 319
2012 189
3094 184
1577 2208
2797 185
4950 180
210 161
4673 1522
0.7
a
IC90: drug concentration (µM) required to reduce cell survival
to 10% of controls using V-79 cells in the clonogenic assay.
Hypoxic selectivity is expressed as the ratio of IC90 values in air
b
and N2 [IC90(air)/IC90(nitrogen)].
the cytotoxicities of the corresponding sulfones (28-30)
were similar to those of the sulfoxides (less than 2-fold).
Differences in partition behavior of the sulfides com-
pared to the sulfoxides/sulfones may also contribute to
the overall uptake and hence intracellular drug avail-
ability for cell damage.
Having shown that the sulfides are much more
cytotoxic than the sulfoxides and sulfones, the sulfoxides
(1, 3-11) were tested for their hypoxia-selective cyto-
toxicity. A clonogenic assay with V-79 cells was used
as described in the Experimental Section. The selection
of this cell line was based on the literature information30
that V-79 cells contain DT-diaphorase, NADPH cyto-
chrome c reductase, and xanthine oxidase, where the
latter two are known to be involved in the reduction of
sulfoxides to the corresponding sulfides. The relative
cytotoxicity profiles along with the hypoxia selectivity
of the sulfoxides are summarized in Table 2. Illustrative
data are shown in Figure 1 for a representative com-
pound of high hypoxia selectivity (7) and of low hypoxia
selectivity (3) of the compounds tested. In general, the
alkylphenyl sulfoxide nitrogen mustards (1, 3) showed
low hypoxia selectivity (∼3-fold), which was due to the
combination of relatively high cytotoxicity under aerobic
conditions and low cytotoxicity under hypoxic condi-
tions. On the other hand, the diaryl sulfoxide nitrogen
mustards (4, 5, 7-9) were less cytotoxic under aerobic
conditions and more cytotoxic under hypoxic conditions
than the alkylphenyl sulfoxides and therefore showed
greater hypoxia selectivity (11-27-fold). Compound 9,
an acridine-substituted diphenyl sulfoxide nitrogen
mustard, showed the greatest hypoxia selectivity (27-

Sulfoxide-Containing Aromatic Nitrogen Mustards
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4163
Ta ble 3. Bioreduction Activitya and Reduction Potentials of Methylphenyl Sulfoxide, Diphenyl Sulfoxide, and Dibenzothiophen-5-one
enzyme activityb
N2, controlc
methylphenyl sulfoxide
diphenyl sulfoxide
dibenzothiophen-5-one
0
0
0
N2, without cofactors
N2, benzaldehyde
N2, 2-hydroxypyrimidine
N2, acetaldehyde
N2, NADPH
0
0
0
10.2 ( 0.5 (sulfide)
18.0 ( 0.5 (sulfide)
4.4 ( 0.7 (sulfide)
17.4 ( 0.9 (sulfide)
6.0 ( 0.2 (sulfide)
0
0
0
N2, menadioned
air, controlc
air, without cofactors
air, benzaldehyde
air, NADPH
0
0
11.6 ( 1.4 (sulfone)
2.3
Ee (-V)
2.4
1.82
a
The bioreduction experiments were conducted by incubation of the sulfoxides with rat liver S-9 fractions under aerobic (S-9 open to
b
air) and anaerobic (S-9 pregassed with N2) conditions, and the metabolites sulfide and sulfone were analyzed by HPLC. Enzyme activity
values are expressed as sulfide or sulfone formation in nmol/mg protein, and mean ( SD was from three experiments. The substance in
parentheses was the metabolite detected. c The sulfoxides were incubated with boiled rat S-9 fraction as control in the presence of either
d
NADPH under aerobic conditions or benzaldehyde under anaerobic conditions. The S-9 fraction was preincubated with menadione for
30 min under aerobic conditions, then used in the presence of benzaldehyde. e Reduction potential (SCE) values were determined by
differential pulse polarography.
Ta ble 4. Formation of Sulfide Metabolites from Incubations of the Corresponding Sulfoxide Prodrugs with Rat S-9 Fractions
bioreduction (%)a
N2, controlb
1e
5
8
9
10
11
0
0
0
0
0
0
0
0
N2, without cofactors
N2, benzaldehyde
N2, 2-hydroxypyrimidine
N2, acetaldehyde
N2, NADPH
2.3
1.29 ( 0.15
0.46 ( 0.05
1.68 ( 0.10
2.02 ( 0.03
0.60 ( 0.05
0.60 ( 0.01
0.33 ( 0.01
5.00 ( 0.43
2.28 ( 0.08
4.50 ( 0.43
1.34 ( 0.06
4.42 ( 0.38
1.79 ( 0.04
4.3
0.37 ( 0.03
1.37 ( 0.03
N2, menadionec
air, controlb
air, without cofactors
air, NADPH
0
0
0
0
0
0
0
0
0
0
2.55
1.07 ( 0.09
0
0
0
0
0
1.0
2.6
air, benzaldehyde
Ed (-V)
2.60
2.63
2.63
2.03
1.96
a
The bioreduction experiments were conducted by incubation of the sulfoxides with rat liver S-9 fractions under aerobic or anaerobic
conditions, and the metabolites were analyzed by TLC densitometry. The values, representing percentage of sulfide formation from the
b
corresponding sulfoxide, were from three experiments with the same batch of S-9 fractions. See footnote c in Table 3. c See footnote d
d
in Table 3. See footnote e in Table 3. e The bioreduction data was determined by HPLC reported previously.19
hyde oxidase inhibitor. All these findings suggested that
liver aldehyde oxidase, not other NADPH-dependent
enzymes, was the enzyme involved in the reduction of
diphenyl sulfoxide under anaerobic conditions. Under
aerobic conditions, diphenyl sulfoxide was not reduced
to the corresponding sulfide in the presence of benzal-
dehyde. This was consistent with the literature observa-
tion32 that sulfoxide reduction by aldehyde oxidase is
inhibited by oxygen. The feature that diphenyl sulfoxide
is bioreduced to the sulfide by aldehyde oxidase only
under anaerobic conditions may explain why the diaryl
sulfoxide nitrogen mustards (4, 5, 7-9) showed sub-
stantial hypoxia-selectivity characteristics as stated
above.
benzaldehyde, 2-hydroxypyrimidine, or acetaldehyde.
The bioreduction was inhibited in the presence of
menadione, indicating that aldehyde oxidase is involved
in the reduction. In addition, 9 was also reduced in the
presence of NADPH or without cofactor, and the biore-
duction was not completely inhibited by menadione,
suggesting some additional enzyme(s) may be involved
in the reduction. Regardless, both 8 and 9 were reduced
exclusively under anaerobic conditions. Despite numer-
ous attempts, we were unable to detect the sulfide
metabolite (20) from 5. The reason is unknown. Inter-
estingly, however, 5 produced 7-fold further enhanced
hypoxia-selective cytotoxicity against V-79 cells in the
presence of benzaldehyde (0.25 mM) under hypoxic
conditions. Benzaldehyde, at this concentration, had no
effect on the controls (both aerobic and hypoxic). This
finding suggested that V-79 cells contain aldehyde
oxidase and a significantly enhanced reduction of the
sulfoxide to the cytotoxic sulfide takes place in the
presence of a suitable electron donor for aldehyde
oxidase. Compounds 10 and 11 are fused diphenyl-
substituted sulfoxide nitrogen mustards. Compound 10
was reduced to its sulfide 32 in the presence of benzal-
dehyde or 2-hydroxypyrimidine, indicating involvement
of aldehyde oxidase in the reduction. The reason 10 was
highly toxic under aerobic conditions is not clear.
Compound 11 was not significantly reduced to its sulfide
34 in the presence of benzaldehyde or 2-hydroxypyri-
midine, suggesting that 11 may be not a good substrate
As reported previously,20 compound 1, an alkylphenyl-
substituted sulfoxide nitrogen mustard, was reduced to
its sulfide 2 in the presence of NADPH (Table 4)
suggesting that an NADPH-dependent enzyme may be
involved in the bioreduction. The fact that this reduction
took place under both aerobic and anaerobic conditions
may explain why this compound was more cytotoxic
under aerobic conditions and less hypoxia-selective than
the diaryl counterparts. The metabolism data of some
selected sulfoxide nitrogen mustards is summarized in
Table 4. Compounds 5, 8, and 9 are diphenyl-substituted
sulfoxide nitrogen mustards and all showed substantial
hypoxia selectivity. Under anaerobic conditions, both 8
and 9 were significantly reduced to their corresponding
sulfides (26 and 27, respectively) in the presence of

4164 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Sun et al.
Ta ble 5. Formation of 8 from Incubation of 7 with Rat S-9
Fractions
methylphenyl sulfoxide, suggesting that diaryl sulfox-
ides may be the better substrates for aldehyde oxidase
than alkylaryl sulfoxides.
bioreduction (%)a
incubation conditions
N2
air
Exp er im en ta l Section
boiled S-9b
0
0
Ch em istr y. Melting points were determined on a Thomas-
Hoover capillary melting point apparatus and were uncor-
rected. Elemental analyses were performed by Atlantic Mi-
crolab, Inc., Norcross, GA. The following instruments were
used: IR, Perkin-Elmer model 281; 1H NMR, Varian EM-360L
CW, 60 MHz (TMS as internal standard). Silica gel GF plates
(Analtech) were used for TLC (250 µm, 2.5 × 10 cm) and
preparative TLC (1000 µm, 20 × 20 cm). Silica gel (40 µm;
Baker) was used for flash column chromatography. All chemi-
cals and solvents were reagent grade and were purchased from
commercial vendors.
4-P r op ylth iop h en yla m in e (12). To a boiling suspension
of 4-nitro-1-propylthiobenzene33,34 (11.2 g, 56.8 mmol) in a
solution of 37% HCl (35 mL) in water (30 mL) were added tin
turnings (11.0 g) upon which the suspension became a clear
light brown solution. The solution was heated under reflux
with vigorous stirring for 15 min. Activated charcoal (0.10 g)
was added and the mixture was allowed to continue refluxing
for an additional 15 min. The mixture was filtered hot and
the filtrate was cooled and diluted with cold water (100 mL).
Aqueous NaOH (40%) was added till the mixture became
strongly alkaline. The resulting gray product was filtered and
purified with charcoal give 10.1 g of white solid which was
further purified on a flash column using hexane/EtOAc (4:1)
as eluent to give 8.13 g (85.6% yield) of 12 as a yellow oil: TLC
Rf 0.37 in hexane/EtOAc (4:1); 1H NMR (CDCl3) δ 1.00 (t, J )
6 Hz, 3H), 1.32-1.93 (m, 2H), 2.75 (t, J ) 6 Hz, 2H), 3.58 (s,
2H), 6.58 (d, J ) 8 Hz, 2H), 7.24 (d, J ) 8 Hz, 2H).
without cofactors
NADPH
NADH
benzaldehyde
2-hydroxypyrimidine
acetaldehyde
menadionec
1.75 ( 0.09
42.10 ( 4.33
9.00 ( 0.43
17.93 ( 0.62
8.50 ( 0.75
12.92 ( 0.52
1.60 ( 0.03
1.14 ( 0.14
16.00 ( 1.00
9.75 ( 0.66
1.05 ( 0.17
a
The bioreduction experiments were conducted by incubation
of the sulfoxides with rat liver S-9 fractions under aerobic (air) or
anaerobic (N2) conditions, and the metabolites were analyzed by
TLC densitometry. The values, representing percentage of sulfide
formation from sulfoxide, were from three experiments with the
same batch of S-9 fractions. See footnote c in the Table 3. c See
b
footnote d in the Table 3.
of aldehyde oxidase. This might be a possible reason to
explain the relative lack of hypoxia-selective cytotoxicity
of 11. The bioreductivity of the sulfoxides did not
parallel their chemical reduction potentials. Compound
11, for example, was the most easily reduced chemically
(E ) -1.96 V), but it was apparently a poor substrate
for aldehyde oxidase and one of the least hypoxia-
selective compounds (Table 4).
Compound 7 contains both nitro and sulfoxide groups
in the molecule. The metabolism profiles of 7 are
summarized in Table 5. Under our experimental condi-
tions, reduction of the nitro group to the amine metabo-
lite (8) was observed in most cases. The reductions of
sulfoxide group however occurred only in the presence
of NADPH under aerobic conditions. Compound 7 was
significantly reduced to 8 in the presence of NADPH
under both aerobic and anaerobic conditions in the
amount of 16% and 42% bioreduction, respectively.
NADH, albeit less effective, was also involved under
both conditions. These suggest that the NADPH/NADH-
dependent enzyme system is responsible for reduction
of the nitro group in 7. Compound 7 was also reduced
to 8 in the presence of benzaldehyde, 2-hydroxypyrimi-
dine, and acetaldehyde in the amount of 17.93%, 8.50%,
and 12.92%, respectively. These reductions were ob-
served only under hypoxic conditions and were inhibited
by menadione. These findings indicate that aldehyde
oxidase may also be involved in the reduction of the
nitro group in 7. The metabolism study of 7 suggests
compound 7 might be the prodrug of 8. The nitro group
is first reduced to form 8, followed by reduction of the
sulfoxide to form the reactive sulfide nitrogen mustard
26.
In conclusion, these results support that properly
designed sulfoxide-containing nitrogen mustards can
serve as potential hypoxia-directed bioreductive cyto-
toxins. Diaryl sulfoxide nitrogen mustards showed much
greater hypoxia-selective in vitro cytotoxicity than alky-
laryl sulfoxide mustards. Metabolism study with rat S-9
fraction showed that bioreduction of compound 8 under
anaerobic conditions can be inhibited by menadione, an
inhibitor of aldehyde oxidase, and enhanced by electron
donors of aldehyde oxidase (e.g., benzaldehyde, acetal-
dehyde, and 2-hydroxypyrimidine). This suggests that
aldehyde oxidase may be involved in the reduction of
the sulfoxides. Diphenyl sulfoxide is more easily reduced
with rat S-9 fraction under anaerobic conditions than
2-[(2-H yd r oxyl)(4-p r op ylt h iop h en yl)a m in o]et h a n -1-
ol (13). 4-Propylthiophenylamine (12) (8.12 g, 48.6 mmol) and
4 equiv of 2-chloroethanol (14.3 mL, 17.15 g, 196 mmol) were
added to a suspension of CaCO3 (10.0 g, 99.9 mmol) in water
(250 mL). The cloudy mixture was heated under reflux with
vigorous stirring, protected from light, for 24 h. An additional
4 equiv of 2-chloroethanol, CaCO3 (10.0 g) were added in two
2-equiv portions over the following 48 h. The reaction mixture
was then cooled, adjusted to pH 7.0 with 10% aqueous NaOH,
and extracted with EtOAc (4 × 200 mL). The combined organic
layer was dried over anhydrous Na2SO4 and evaporated to give
8.18 g of brown oil. The crude material was purified on a flash
column with eluent EtOAc/hexane (from ratio 2:1 to 3:1) to
give 3.80 g (30.7% yield) of 13 as a light brown oil: TLC Rf
1
0.46 in EtOAc/hexane (3:1); H NMR (CDCl3) δ 0.92 (t, J ) 6
Hz, 3H), 1.32-1.68 (m, 2H), 2.74 (t, J ) 6 Hz, 2H), 3.15-3.94
(m, 8H), 6.55 (d, J ) 8 Hz, 2H), 7.30 (d, J ) 8 Hz, 2H).
Bis(2-ch lor oeth yl)(4-p r op ylth iop h en yl)a m in e (14). A
solution of 13 (3.80 g, 14.9 mmol) in POCl3 (9.0 mL) was heated
under reflux for 40 min, then the hot reaction mixture was
added, with stirring, to a mixture of ice/water (150 mL). The
mixture was extracted with EtOAc (4 × 100 mL) and washed
with 10% NaHCO3 solution (3 × 200 mL). The combined
organic layer was dried over anhydrous Na2SO4 and evapo-
rated to give 3.23 g (70.0% yield) of 14 as a brown oil: TLC Rf
1
0.84 in hexane/EtOAc (2:1); H NMR (CDCl3) δ 0.95 (t, J ) 6
Hz, 3H), 1.32-1.82 (m, 2H), 2.74 (t, J ) 6 Hz, 2H), 3.32-3.75
(m, 8H), 6.60 (d, J ) 8 Hz, 2H), 7.34 (d, J ) 8 Hz, 2H). Anal.
(C13H19Cl2NS) C, H, Cl, N, S.
4-[Bis(2-ch lor oe t h yl)a m in o]-1-(p r op ylsu lfin yl)b e n -
zen e (3). To a solution of 14 (1.00 g, 3.42 mmol) in 5 mL TFA
was added H2O2 in TFA (1.0 mL, 1 equiv), prepared by adding
30% H2O2 (8.6 mL) to TFA (16.4 mL), while stirring at 0 °C.
The reaction was allowed to proceed for 2 h. The ice bath was
removed and the reaction was allowed to continue for an
additional 30 min. TFA was evaporated to give 0.78 g (74.2%
yield) of 3 as a light yellow solid: mp 45-48 °C; TLC Rf 0.41
1
in EtOAc/hexane (3:1); H NMR (CDCl3) δ 0.95 (t, J ) 6 Hz,
3H), 1.40-1.84 (m, 2H), 2.85 (t, J ) 6 Hz, 2H), 3.38-3.79 (m,
8H), 6.45 (d, J ) 8 Hz, 2H), 7.20 (d, J ) 8 Hz, 2H). Anal.
(C13H19Cl2NOS) C, H, Cl, N, S.

Sulfoxide-Containing Aromatic Nitrogen Mustards
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4165
2-[(2-Hyd r oxyeth yl)(4-p h en ylth ioph en yl)a m in o]eth a n -
1-ol (16) was prepared from 4-phenylthiophenylamine (15),
obtained from reduction of 4-nitro-1-phenylthiobenzene ac-
cording to the procedure for 12, on a 50.0-mmol (10.1-g) scale
as in the procedure for 13 to give 9.60 g of the crude product
as a brown oil. The oil was purified on a flash column using
eluent EtOAc/hexane (from ratio 1:1 to 2:1) to give 2.26 g
(15.6% yield) of 16 as a brown oil: TLC Rf 0.25 in EtOAc/
hexane (2:1); 1H NMR (CDCl3) δ 3.26-4.00 (m, 8H), 6.52-6.77
(m, 4H), 7.19 (s, 5H).
oil was purified on preparative TLC plates using EtOAc/
hexane (1:1) as eluent to give 0.23 g (70.1% yield) of 6 as a
white crystalline solid: TLC Rf 0.35 in EtOAc/hexane (1:1);
1H NMR (CDCl3) δ 3.22-3.75 (m, 8H), 6.78 (d, J ) 9 Hz, 4H),
7.33 (d, J ) 9 Hz, 4H). Anal. (C16H18Cl2N2OS) C, H, Cl, N, S.
2-{(2-H yd r oxye t h yl)[4-(4-n it r op h e n ylt h io)p h e n yl]-
am in o}eth an -1-ol (24). To a solution of 4-amino-4′-nitrodiphe-
nyl sulfide (23) (25.0 g, 102 mmol) in THF (200 mL) and AcOH
(250 mL) was added 10% of ethylene oxide in THF (315 mL,
715 mmol) 45-mL portions over the following 7 days at room
temperature. The mixture was then neutralized with Na2CO3/
H2O and extracted with EtOAc for three times. The EtOAc
extract was washed with water, dried over anhydrous sodium
sulfate, and then evaporated to remove EtOAc. The crude
product was crystallized with EtOAc to give 19.5 g of yellow
product 24. The mother liquid was evaporated, and the residue
was purified on a flash column using CH2Cl2/CH3COCH3 (8:
2) as eluent, then recrystallized with EtOAc/hexane to obtain
7.8 g of 24 (80.4% total yield): mp 99.5-100.5 °C; TLC Rf 0.41
in CH2Cl2/CH3COCH3 (8:2); 1H NMR (CDCl3) δ 3.63-3.90 (m,
8H), 6.75 (d, J ) 9 Hz, 2H), 7.00-7.45 (m, 4H), 8.00 (d, J ) 9
Hz, 2H). Anal. (C16H18N2O4S) C, H, N, S.
B is (2-c h lo r o e t h y l)[4-(4-n it r o p h e n y lt h io )p h e n y l]-
a m in e (25) was prepared from 24 on a 29.9-mmol (10.0-g)
scale as in the similar procedure for 14 to give 8.90 g of 25
(80.2% yield) as a yellow crystalline after crystallization with
EtOAc/hexane: mp 112.5-113.5 °C; TLC Rf 0.51 in hexane/
EtOAc (5:1); 1H NMR (CDCl3) δ 3.60-3.80 (m, 8H), 6.70 (d, J
) 9.6 Hz, 2H), 6.95-7.45 (m, 4H), 8.00 (d, J ) 9.6 Hz, 2H).
Anal. (C16H16Cl2N2O2S) C, H, N, Cl, S.
Bis(2-ch lor oeth yl)(4-p h en ylth iop h en yl)a m in e (17) was
prepared from 16 on
a 7.82-mmol (2.26-g) scale in the
procedure for 14 to give 2.11 g of the crude product as a light
brown oil. The crude was purified on a flash column using
hexane/EtOAc (4:1) as eluent to give 2.03 g (79.7% yield) of
17 as a light brown oil: TLC Rf 0.69 in hexane/EtOAc (4:1);
1H NMR (CDCl3) δ 3.32-3.78 (m, 8H), 6.40-7.48 (m, 4H), 7.11
(s, 5H). Anal. (C16H17Cl2NS) C, H, Cl, N, S.
4-[Bis(2-ch lor oe t h yl)a m in o]-1-(p h e n ylsu lfin yl)b e n -
zen e (4) was prepared from 17 on a 3.07-mmol (1.00-g) scale
in the procedure for 3 to give 0.84 g of the crude product as a
dark brown viscous oil. The crude oil was purified on a flash
column using EtOAc/hexane (3:1) to give 0.71 g (67.6% yield)
of 4 as an ivory-colored crystalline product: mp 122-124 °C;
TLC Rf 0.68 in EtOAc/hexane (3:1); 1H NMR (CDCl3) δ 3.36-
3.77 (m, 8H), 6.51-7.59 (m, 4H), 7.20 (s, 5H). Anal. (C16H17
Cl2NOS) C, H, Cl, N, S.
-
2-[(4-{4-[Bis(2-h ydr oxyeth yl)am in o]ph en ylth io}ph en yl)-
(2-h yd r oxyeth yl)a m in o]eth a n -1-ol (19) was prepared from
4-(4-aminophenylthio)phenylamine27 (18) (6.50 g 30.0 mmol)
and 12 equiv of 2-chloroethanol as in the similar procedure
for 13 to give 6.60 g crude product. The crude was purified on
a flash column to give 3.15 g (26.8% yield) of 19 as an off-
white solid: mp 135-137 °C; TLC Rf 0.60 in EtOH/EtOAc (1:
4); 1H NMR (CDCl3/DMSO-d6, 3:1) δ 3.16-3.68 (m, 16H), 6.60
(d, J ) 9 Hz, 4H), 7.60 (d, J ) 9 Hz, 4H).
4-({4-[Bis(2-ch lor oet h yl)a m in o]p h en yl}su lfin yl)-1-n i-
tr oben zen e (7) was prepared from 25 on a 36.4-mmol (13.5-
g) scale as in the procedure for 3 to give 13.7 g (97.1% yield)
of 7 as yellow crystalline after crystallization with EtOAc/
hexane: mp 107-108 °C; TLC Rf 0.45 in EtOAc/hexane (1:1);
1H NMR (CDCl3) δ 3.60-3.77 (m, 8H), 6.70 (d, J ) 9.6 Hz,
Bis(2-ch lor oe t h yl)[4-{4-[b is(2-ch lor oe t h yl)a m in o]-
p h en ylth io}p h en yl]a m in e (20) was prepared from 19 on a
3.82-mmol (1.50-g) scale in the procedure for 14 to give 1.24 g
of the crude product as a light brown oil. The crude was
purified on a silica gel column to give 1.20 g (67.3% yield) of
20 as an off-white solid: mp 65-66 °C; TLC Rf 0.51 in hexane/
2H), 7.25-7.80 (m, 4H), 8.30 (d, J ) 9.6 Hz, 2H). Anal. (C16H16-
Cl2N2O3S) C, H, Cl, N, S.
4-({4-[Bis(2-ch lor oeth yl)a m in o]p h en yl}su lfon yl)-1-n i-
tr oben zen e (28). To a solution of 7 (1.00 g, 2.58 mmol) in
130 mL acetone and 13 mL of water was added a mixture of
25.8 mL 30% (wt) of hydrogen peroxide (252 mmol) and 3.9
mL 0.3 M of ammonium molybdate. The reaction mixture was
stirred at room temperature for 5 h and evaporated in vacuo
at 30 °C. The residue was partitioned between CH2Cl2 and
water. The water layer was extracted with CH2Cl2. The organic
extract was dried over anhydrous MgSO4 and evaporated to
give the residue which was recrystallized with EtOAc/hexane
to give 1.00 g (96.0% yield) of 28 as light yellow crystals: mp
132.5-133.5 °C; TLC Rf 0.75 in EtOAc/hexane (1:1); 1H NMR
(CDCl3) δ 3.63-3.80 (m, 8H), 6.71 (d, J ) 9.6 Hz, 2H), 7.70-
8.40, (m, 6H). Anal. (C16H16Cl2N2O4S) C, H, Cl, N, S.
1
EtOAc (5:1); H NMR (CDCl3) δ 3.46-3.89 (m, 16H), 6.58 (d,
J ) 9 Hz, 4H), 7.28 (d, J ) 9 Hz, 4H). Anal. (C20H24Cl4N2S) C,
H, Cl, N, S.
1-[Bis(2-ch lor oeth yl)a m in o]-4-[{4-[bis(2-ch lor oeth yl)-
a m in o]p h en yl}su lfin yl]ben zen e (5) was prepared from 20
on a 1.99-mmol (0.93-g) scale in the procedure for 3 to give
0.75 g (78.0% yield) of 5 as an off-white solid: mp 130.5-132
°C; TLC Rf 0.26 in EtOAc/hexane (3:2); 1H NMR (CDCl3) δ
3.40-3.95 (m, 16H), 6.70 (d, J ) 9.6 Hz, 4H), 7.28 (d, J ) 9.6
Hz, 4H). Anal. (C20H24Cl4N2OS) C, H, Cl, N, S.
2-[(4-{4-[Bis(2-h ydr oxyeth yl)am in o]ph en ylth io}ph en yl)-
a m in o]eth a n -1-ol (21) was prepared from 4-(4-aminophe-
nylthio)phenylamine (18) (5.00, 23.1 mmol) and 6 equiv of
2-chloroethanol as in the similar procedure for 13 to give 6.23
g of a brown oil. The crude oil was purified on a flash column
using EtOAc/hexane (5:1) as eluent to give 0.77 g (11.0% yield)
of 21 as a pale yellow oil: TLC Rf 0.23 in EtOAc/hexane (4:1);
1H NMR (CDCl3) δ 2.92-3.45 (m, 8H), 6.53 (d, J ) 9 Hz, 4H),
7.12 (d, J ) 9 Hz, 4H).
(2-Ch lor oeth yl)[4-{4-[(2-ch lor oeth yl)am in o]ph en ylth io}-
p h en yl]a m in e (22) was prepared from 21 on a 21.4-mmol
(0.75-g) scale as in the procedure for 14 to give 0.63 g of the
crude product as a golden yellow oil. The crude oil was purified
on preparative TLC plates using hexane/EtOAc (3:1) as eluent
to give 0.46 g (52.3% yield) of 22 as a dark brown oil: TLC Rf
0.43 in hexane/EtOAc (3:1); 1H NMR (CDCl3) δ 3.18-3.68 (m,
8H), 6.70 (d, J ) 9 Hz, 4H), 7.22 (d, J ) 9 Hz, 4H). Anal.
(C16H18Cl2N2S) C, H, Cl, N, S.
4-({4-[B is (2-c h lo r o e t h y l)a m in o ]p h e n y l}s u lfin y l)-
p h en yla m in e (8). Ion powder (19.1 g, 100 mesh) was acti-
vated by refluxing it with 1.5 mL of distilled water and 1 drop
of concentrated hydrochloric acid for 0.5 h under vigorous
stirring. Ethanol (68.4 mL) and 7 (13.7 g, 35.3 mmol) were
added into the mixture at 40-50 °C. The reaction mixture was
slowly heated until reflux, then vigorously stirred for 15 mim.
Several drops of 10% NaOH was added to the reaction mixture
to pH 8. The hot reaction mixture was then filtered, and the
residue was washed with EtOH for 3 times. After evaporation
of EtOH, the residue was diluted with water, and extracted
with EtOAc for 3 times. The EtOAc extract was dried over
anhydrous Na2SO4 and evaporated in vacuum. The crude was
purified on a flash column using EtOAc/CH2Cl2 (6:4) as eluent,
then recrystallized with EtOAc/hexane to give 9.09 g (72.1%
yield) of 8 as an off-white crystalline: mp 136.5-138 °C; TLC
Rf 0.49 in EtOAc; 1H NMR (CDCl3) δ 3.60-3.73 (m, 8H), 6.60-
6.75 (m, 4H), 7.20-7.50 (m, 4H). Anal. (C16H18N2Cl2OS) C, H,
Cl, N, S.
1-[(2-Ch lor oeth yl)a m in o]-4-[{4-[(2-ch lor oeth yl)a m in o]-
p h en yl}su lfin yl]ben zen e (6) was prepared from 22 on a
0.91-mmol (0.31-g) scale as in the procedure for 3 to give 0.29
g of the crude product as a light gray viscous oil. The crude
[4-(4-Am in op h e n ylt h io)p h e n yl]b is(2-ch lor oe t h yl)-
a m in e (26) was prepared from 25 (2.00 g, 5.39 mmol)
according to the procedure for 8 to give 1.17 g (63.8% yield) of

4166 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Sun et al.
26 as a light brown solid: mp 77-79 °C; TLC Rf 0.64 in CH2-
Cl2/Me2CO (49:1); 1H NMR (CDCl3) δ 3.45-3.75 (m, 8H), 6.50-
6.70, (m, 4H), 7.10-7.30 (m, 4H). Anal. (C16H18Cl2N2S) C, H,
Cl, N, S.
4-({4-[B is (2-c h lo r o e t h y l)a m in o ]p h e n y l}s u lfo n y l)-
p h en yla m in e (29) was prepared from 28 (1.00 g, 2.48 mmol)
according to the procedure for 8 to give 0.76 g (70.1% yield) of
29 as an off-white solid after flash column purification using
CH2Cl2/EtOAc (95:5) as eluent: mp 153.5-155 °C; TLC Rf 0.41
in EtOAc/hexane (1:1); 1H NMR (CDCl3/DMSO-d6, 2:1) δ 3.55-
3.80 (m, 8H), 6.50-6.80 (m, 4H), 7.35-7.75 (m, 4H). Anal.
(C16H18Cl2N2O2S) C, H, Cl, N, S.
2,8-Bis[bis(2-ch lor oeth yl)am in o]diben zoth ioph en e (34)
was prepared from 33 on a 3.53-mmol (1.38-g) scale in the
procedure for 14 to give 0.44 g (29.1% yield from 2,8-
diaminodibenzothiophene) of 34 as a light brown solid: mp
160-163 °C dec; TLC Rf 0.58 in EtOAc/hexane (2:8); 1H NMR
(CDCl3) δ 3.10-4.45 (m, 16H), 6.85-7.95 (m, 6H). Anal.
(C20H22Cl4N2S) C, H, Cl, N, S.
2,8-Bis[b is(2-ch lor oet h yl)a m in o]d ib en zot h iop h en -5-
on e (11) 50-60% m-chloroperbenzoic acid (mCPBA) (0.29 g,
1.08 mmol) in 6 mL of CH2Cl2 was added dropwise to a solution
of 34 (0.5 g, 1.08 mmol) in 8 mL of CH2Cl2 with stirring at 0
°C in an ice bath, and the reaction mixture was stirred for
additional 10 min at 0 °C. The reaction solution was then
washed sequentially with water (2 × 50 mL), saturated
aqueous Na2SO3 (4 × 50 mL), saturated aqueous NaHCO3 (2
× 50 mL), and brine (2 × 50 mL). After drying over anhydrous
Na2SO4, the solvent was removed to give 0.51 g of the crude
product. The crude material was purified on a flash column
using EtOAc/CH2Cl2 (1:9) as eluent to give 0.34 g (65.4% yield)
of 11 as a light brown solid: mp 185-187 °C dec; TLC Rf 0.28
in EtOAc/CH2Cl2 (3:7); 1H NMR (CDCl3) δ 3.00-4.30 (m, 16H),
6.55-8.21 (m, 6H). Anal. (C20H22Cl4N2OS) C, H, Cl, N, S.
{4-[4-(Acr idin -9-ylam in o)ph en ylth io]ph en yl}bis(2-ch lo-
r oeth yl)a m in e Hyd r och lor id e (27). A solution of 26 (0.52
g, 1.53 mmol) and 9-chloroacridine35 (0.30 g, 1.40 mmol) in
N-methyl-2-pyrrolidinone (13.5 mL) was treated with a few
drops of concentrated HCl and then stirred at room temper-
ature for 1.5 h. The mixture was diluted with EtOAc, and the
resulting precipitate was collected and washed with EtOAc to
give a crude product. This crude product was mixed with 10
mL of N-methyl-2-pyrrolidinone and stirred for 1 h. Ethyl
acetate (250 mL) was gradually added for complete precipita-
tion, and the resulting precipitate was collected and washed
with EtOAc to give 0.70 g (89.4% yield) of 27 as an orange
solid: mp 261-262 °C dec; TLC Rf 0.73 in EtOAc/hexane (1:
In Vitr o Cytotoxicity Assessm en t. 1. Clon ogen ic As-
sa y. The published procedure of Fracasso and Sartorelli37 with
some modifications was used. A confluent flask of Chinese
hamster lung transformed V-79 transformed cells was scraped
to collect the monolayer and the cells were counted by trypan
blue dye exclusion. Glass milk dilution bottles were then
seeded with 500 000 cells in 10 mL of medium (MEM supple-
mented with 10% fetal bovine serum, Pen G sodium, strepto-
mycin sulfate, amphotericin, and essential and nonessential
amino acids and vitamins) and incubated for 20 h in a 95%
air/5% humidified atmosphere. One set of six bottles was
sealed with sterilized rubber septa and fitted with 21-gauge
needles for gas inflow and outflow. Each bottle was gassed
individually, all outlet tubings were connected to individual
flow meters (average flow rate ) 0.15 L/min) with additional
tubings immersed in water to allow visible monitoring of gas
flow as well as to prevent back flow of air into the cultures.
To produce hypoxia, cultures were gassed continuously for 2
h at 37 °C (maintained in a water bath) with a humidified
mixture of 95% N2/5% CO2. Normally aerated cultures were
maintained in air until drug treatment. At this time, appropri-
ate dilutions of the drug in ethanol/water or DMF were added
directly to the cultures, without breaking the hypoxia, by
injecting 0.25 mL of drug through the rubber septa. Cells in
the aerobic set were exposed to drug for 3 h under normal
aerated conditions. The hypoxic set was gassed 10 min after
injection of the drug, needles were pulled out, and the sealed
bottles were incubated for 3 h in a normally aerated incubator.
After drug exposure, the cells were washed twice with 5 mL
sterile HBSS and treated with 2.5% trypsin (Gibco Labs) for
2-3 min. Cells were collected by centrifugation, resuspended
in 5 mL fresh media, and counted by trypan blue dye exclusion.
Appropriate dilutions were made, and 500 cells were plated
in triplicate in a total of 10 mL medium. Dishes were incubated
for 6-8 days to allow colony formation. Colonies were then
rinsed twice with 0.9% saline, fixed with ethanol, stained with
crystal violet and counted. Results are reported as the number
of colonies surviving chemical treatment per number of
colonies in the solvent-treated control. The IC50 or IC90 values
were determined by semilogarithmically plotting the drug
concentration versus cell viability as determined by the
number of colonies surviving the treatments.
1
1); H NMR (CDCl3/DMSO-d6, 1:4) δ 3.70-3.90 (m, 8H), 6.80
(d, J ) 9 Hz, 2H), 7.20-7.60 (m, 8H), 7.95-8.35 (m, 6H). Anal.
(C29H26 Cl3N3S) C, H, Cl, N, S.
4-{[4-(Acr idin -9-ylam in o)ph en yl]su lfin yl}-1-[bis(2-ch lo-
r oeth yl)a m in o]ben zen e h yd r och lor id e (9) was prepared
from 8 (1.00 g, 2.81 mmol) and 9-chloroacridine (0.55 g, 2.57
mmol) according to the procedure for 27 to give 1.30 g (88.4%
yield) of 9 as an orange solid: mp 250-251 °C dec; TLC Rf
0.81 in EtOAc; 1H NMR (CDCl3/DMSO-d6, 1:4) δ 3.70-3.85
(m, 8H), 6.88 (d, J ) 10.2 Hz, 2H), 7.45-4.60 (m, 8H), 8.00-
8.40 (m, 6H). Anal. (C29H26Cl3N3OS) C, H, Cl, N, S.
4-{[4-(Acr idin -9-ylam in o)ph en yl]su lfon yl}-1-[bis(2-ch lo-
r oeth yl)a m in o]ben zen e h yd r och lor id e (30) was prepared
from 29 (0.38 g, 1.02 mmol) and 9-chloroacridine (0.20 g, 0.94
mmol) according to the procedure for 27 to give 0.52 g (94.7%
yield) of 30 as an orange solid: mp 270.5-271.5 °C dec; TLC
Rf 0.34 in EtOAc/hexane (1:1); 1H NMR (CDCl3/DMSO-d6, 1:3)
δ 3.65-3.90 (m, 8H), 6.88 (d, J ) 8.4 Hz, 2H), 7.40-8.40 (m,
14H). Anal. (C29H26Cl3N3O2S) C, H, Cl, N, S.
2-[Bis(2-h yd r oxyet h yl)]a m in od ib en zot h iop h en e (31)
was prepared from 2-aminodibenzothiophene36 (4.10 g, 20.5
mmol) and 2.8 equiv of 2-chloroethanol according to the
procedure for 13 to give 6.00 g of 31 as the crude oil. This oil
was used to synthesize 32 without further purification: TLC
Rf 0.54 in EtOAc; 1H NMR (CDCl3/DMSO-d6, 6:1) δ 3.43-4.10
(m, 8H), 6.75-8.30 (m, 7H).
2-[Bis(2-ch lor oeth yl)a m in o]d iben zoth iop h en e (32) was
prepared from 31 on a 24.0-mmol (6.00-g) scale as in the
procedure for 14 to give 5.80 g of a brownish solid. This crude
product was recrystallized with CH2Cl2/hexane to give 3.50 g
(45.5% yield from 2-aminodibenzothiophene) of 32 as a light
brownish solid: mp 101-102 °C; TLC Rf 0.49 in EtOAc/hexane
1
(1:10); H NMR (CDCl3/DMSO-d6, 10:1) δ 3.40-4.05 (m, 8H),
6.60-8.20 (m, 7H). Anal. (C16H15Cl2NS) C, H, Cl, N, S.
2-[Bis(2-ch lor oeth yl)am in o]diben zoth ioph en -5-on e (10)
was prepared from 32 on a 3.04-mmol (1.00-g) scale in the
procedure for 3 to give a brown oil which was purified on a
flash column eluting with EtOAc/CH2Cl2 from ratio 1:20 to 1:4
to give 0.55 g (53.0% yield) of 10 as a brown solid: mp 170-
172 °C dec; TLC Rf 0.31 in EtOAc/hexane (1:1); 1H NMR
2. Su lfor h od a m in e B (SRB) Assa y. 100 µL of Chinese
hamster lung V-79 cells (about 7 × 103 cells/100 µL) was
preincubated for 24 h; 100 µL media, which contains variable
concentrations of drugs, was added to each well and then the
plate was incubated under the oxic conditions for 3 h. The cells
were washed 5 times with HBSS and 200 µL of fresh media
was added. The plate was incubated for 48 h. The subsequent
TCA-fixation and SRB assay were done as reported38 with a
slight modification. The bound dye was solubilized with 10 mM
Tris buffer for 5 min, and then 10-µL aliquots of the solution
(CDCl3) δ 3.65-5.30 (m, 8H), 6.30-7.65 (m, 7H). Anal. (C16H15
Cl2NOS) C, H, Cl, N, S.
-
2,8-Bis[b is(2-h yd r oxyet h yl)a m in o]d ib en zot h iop h en e
(33) was prepared from 2,8-diaminodibenzothiophene36 (0.70
g, 3.25 mmol) and 30 equiv of 2-chloroethanol in the procedure
for 13 to give 1.38 g of 33 as a dark brown thick oil. This oil
was used to synthesize 34 without further purification: TLC
Rf 0.26 in EtOAc/EtOH (9:1); 1H NMR (CDCl3) δ 2.90-5.05
(m, 16H), 6.73-8.22 (m, 6H).

Sulfoxide-Containing Aromatic Nitrogen Mustards
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22 4167
(4) Stratford, I. J .; Workman, P. Bioreductive drugs into the next
millenium. Anticancer Drug Des. 1998, 13, 519-528.
(5) Sinhababu, A. K.; Thakker, D. R. Review: Prodrugs of anticancer
agents. Adv. Drug Deliv. Rev. 1996, 19, 241-273.
(6) Walton, M. I.; Wolf, C. R.; Workman, P. Molecular enzymology
of the bioactivation of hypoxic cell cytotoxins. Int. J . Radiat.
Oncol. Biol. Phys. 1989, 16, 983-986.
(7) Workman, P.; Walton, M. I. Enzyme-directed bioreductive drug
development. In Selective activation of drugs by redox process;
Adams, G. E., Breccia, A., Fielden, E. M., Wardman, P., Eds.;
Plenum: New York, 1991; pp 173-191.
(8) Patterson, A. V.; Saunders, M. P.; Chinje, E. C.; Talbot, D. C.;
Harris, A. L.; Stratford, I. J . Overespression of human NAD-
PH: cytochrome c (P450) reductase confers enhanced sensitivity
to both tirapazamine (SR 4233) and RSU 1069. Br. J . Cancer
1997, 76, 1338-1347.
(9) Patterson, A. V.; Saunders, M. P.; Chinje, E. C.; Patterson, L.
H.; Stratford, I. J . Enzymology of tirapazamine metabolism: a
review. Anti-Cancer Drug Des. 1998, 13, 541-573.
(10) Rauth, A. M.; Goldberg, Z.; Misra, V. DT-diaphorase: possible
roles in cancer chemotherapy and carcinogenesis. Oncol. Res.
1997, 9, 339-350.
(11) Workman, P. Bioreductive mechanisms. Int. J . Radiat. Oncol.
Biol. Phys. 1992, 22, 631-637.
(12) Duggan, D. E.; Hooke, K. F.; Noll, R. M.; Hucker, H. B.; Van
Arman, C. C. Comparative disposition of sulindac and metabo-
lites in five species. Biochem. Pharmacol. 1978, 27, 2311-2320.
(13) Davis, P. J .; Guenthner, L. E. Sulindac oxidation/reduction by
microbial cultures; microbial models for mammalian metabolism.
Xenobiotica 1985, 15, 845-857.
(14) Duggan, D. E. Sulindac: therapeutic implications of the prodrug/
pharmacophore equilibrium. Drug Metab. Rev. 1981, 12, 325-
337.
(15) Lee, S. C.; Renwick, A. G. Sulphoxide reduction by rat and rabbit
tissues in vitro. Biochem. Pharmacol. 1995, 49, 1557-1565.
(16) Tatsumi, K.; Kitamura, S.; Yamada, H. Sulfoxide reductase
activity of liver aldehyde oxidase. Biochim. Biophys. Acta 1983,
747, 86-92.
(17) Yoshihara, S.; Tatsumi, K. Sulfoxide reduction catalyzed by
guinea pig liver aldehyde oxidase in combination with one-
electron reducing flavoenzymes. J . Pharmacobio-Dye 1985, 8,
996-1005.
were spotted on a TLC plate. Optical density (OD) was
determined by the use of UNISCAN video densitometer with
the visible light.
Deter m in a tion of Ch em ica l Red u ction P oten tia l by
P ola r ogr a p h y. The reduction potential of the sulfoxides
compounds was determined by differential pulse polarography.
Metrohm E626 polarecord and an E505 polarographic stand,
which consisted of a dropping mercury electrode (cathode), a
saturated calomelelectrode (SCE.), and a platinum wire as
auxiliary electrode (anode), were used to measure all polaro-
grams. The testing compounds (2 mM) and TBAB (tetrabuty-
lammonium bromide; 0.1 M) as an electrolyte were dissolved
in 50 mL of anhydrous DMF. The solution was purged with
N2 for 10 min and then the potential was scanned from -1.00
to -3.00 V. The peak potential was record on the polarogram,
and each reduction potential value was based on the average
of three readings. All the results were compared with 0.1 M
of TBAB in DMF as a blank.
In Vitr o Meta bolism Stu d y. The rat liver S-9 fraction was
prepared essentially as previously reported.20,39 The incubation
was carried out in a 20-mL vial containing 2-4 mL of rat liver
S-9 fraction, 5-10 µmol of testing compound (0.1 M in EtOH
or DMF), cofactors (2 equiv for NADPH or NADH, 1 equiv for
benzaldehyde, 2-hydroxypyrimidine, or acetaldehyde), and
phosphate buffer to a final volume 3-6 mL. The boiled S-9
fraction and the S-9 fraction without adding any cofactors were
used as controls. For the aldehyde inhibition study, the S-9
fraction was preincubated with 1 equiv menadione for 30 min
under aerobic conditions prior adding testing compounds and
benzaldehyde. Aerobic experiments were performed by leaving
the vials open to the atmosphere during incubation. Anaerobic
incubations were performed by purging the incubation vials
with nitrogen for 10 min prior to the injecting the testing
compounds. Incubations were done at 37 °C for 30 min in a
Dubnoff metabolic shaking incubator.
HPLC or TLC densitometry was used to determine the
metabolites. For HPLC determination, the incubation was
terminated by addition of 2 mL EtOH. The supernatant was
filtered through a syringe filter (0.2 µm, HT Tuffryn mem-
brane) and directly used for HPLC. HPLC analysis was done
on a Waters µ-Bondapak C-18 analytical column (3.9 × 150
mm) with isocratic elution (MeCN/H2O, 10:25) and gradient
system (MeCN/H2O) at flow rate 2 mL/min. For TLC deter-
mination, the incubation was terminated by addition of 6 mL
CH2Cl2. The resulting mixture was centrifuged (3000 rpm, 15
min), and the denatured protein precipitate was extracted with
5 mL of CH2Cl2 3 times. The combined supernatants were
dried with anhydrous sodium sulfate and evaporated. The
residue was then reconstituted with 0.5 mL of CH3Cl. The
metabolites were quantitatively determined on TLC plates
(silica gel GF plates, 250 µm; Analtech) with a Uniscan Video
densitometer (Analtech) under 254-nm UV light. The standard
curves were obtained by using standard sulfides incubated
with boiled S-9 fraction following the same experimental
conditions as that for sulfoxides. The substrate activities are
expressed as their corresponding sulfide metabolite formation/
mg protein or percentage bioreduction.
(18) Yoshihara, S.; Tatsumi, K. Metabolism of diphenyl sulfoxide in
perfused guinea pig liver-involvement of aldehyde oxidase as a
sulfoxide reductase. Drug Metab. Dispos. 1990, 18, 876-881.
(19) Anders, M. W.; Ratnayake, J . H.; Hanna, P. E.; Fuchs, J . A.
Involvement of thioredoxin in sulfoxide reduction by mammalian
tissues. Biochem. Biophys. Res. Commun. 1980, 97, 846-851.
(20) Kwon, C.-H.; Blanco, D. R.; Baturay, N. p-(Methylsulfinyl)phenyl
nitrogen mustard as a novel bioreductive prodrug selective
against hypoxic tumors. J . Med. Chem. 1992, 35, 2137-2139.
(21) Palmer, B. D.; Wilson, W. R.; Pullen, S. M.; Denny, W. A.
Hypoxia-selective antitumor agents. 3. Relationships between
structure and cytotoxicity against cultured cells for substituted
N, N-bis(2-chloroethyl)anilines. J . Med. Chem. 1990, 33, 112-
121.
(22) White, I. N. H.; Suzanger, M.; Mattocks, A. R.; Bailey, E.;
Farmer, P. B.; Conners, T. A. Reduction of nitromin to nitrogen
mustard: unscheduled DNA synthesis in aerobic or anaerobic
rat hepatocytes, J B1, BL8 and Walker carcinoma cell lines.
Carcinogenesis 1989, 10, 2113.
(23) Valu, K. K.; Gourdie, T. A.; Boritzki, T. J .; Gravatt, G. L.;
Baguley, B. C.; Wilson, W. R.; Wakelin, L. P. G.; Woodgate, P.
D.; Denny, W. A. DNA-directed alkylating agents. 3. Structure-
activity relationships for acridine-linked aniline mustards: con-
sequences of varying the length of the linker chain. J . Med.
Chem. 1990, 33, 3014-3019.
(24) Venier, C. G.; Squires, T. G.; Chen, Y. Y.; Hussmann, G. P.; Shei,
J . C.; Smith, B. F. Peroxy trifluoroacetic acid. A convenient
reagent for the preparation of sulfoxides and sulfones. J . Org.
Chem. 1982, 47, 3773-3774.
(25) Davis, F. A.; J enkins, L. A.; Billmers, R. L. Chemistry of Sulfenic
acids. Reason for the high reactivity of sulfenic acids. Stabiliza-
tion by intramolecular hydrogen bonding and electronegativity
effects. J . Org. Chem. 1986, 51, 1033-1040.
Ack n ow led gm en t. We are grateful to Mr. K.-K. Min
and Dr. J .-K. Kim of Medicinal Research Center, CKD
Corp., Korea, for some of the cytotoxicity testing. This
work was supported by the National Cancer Institute,
NIH, under Grant CA 63618-01.
(26) Hardy, P. M.; Rydon, H. N.; Thompson, R. C. The use of
â-methylthioethyl esters for the protection of carboxyl groups
in peptide synthesis: removal through the â-methylsulphonyl-
ethyl ester. Tetrahedron Lett. 1968, 20, 2525-2526.
(27) Raiziss, G. W.; Clemence, L. W.; Sevrac, M.; Moetsch, J . C.
Chemistry and chemotherapy of 4,4′-diaminodiphenyl sulfone,
4-amino-4′-hydroxydiphenyl sulfone and related compounds. J .
Am. Chem. Soc. 1939, 61, 2763-2764.
(28) Gamage, S. A.; Tepsiri, N.; Wilairat, P.; Wojcik, S. J .; Figgitt,
D. P.; Ralph, R. K.; Denny, W. A. Synthesis and in vitro
evaluation of 9-anilino-3,6-diaminoacridines active against mul-
tidrug-resistant strain of the malaria parasite plasmodium
falciparum. J . Med. Chem. 1994, 37, 1486-1494.
Refer en ces
(1) Brown, J . M. Tumor hypoxia: problems and opportunities. In
Encyclopedia of Cancer; Bertino, J . R., Ed.; Academic Press: San
Diego, London, 1997; Vol. III, pp 1883-1898.
(2) Brown, J . M. The potential benefit of hypoxia cytotoxins in radio-
oncology. In Blood perfusion and microenvironment of human
tumors, implications for clinical radiooncology; Molls, M., Vau-
pel, P., Eds.; Springer-Verlag: Berlin, Heidelberg, New York,
1998; pp 33-62.
(3) Rauth, A. M.; Melo, T.; Misra, V. Bioreductive therapies: an
overview of drugs and their mechanisms of action. Int. J . Radiat.
Oncol. Biol. Phys. 1998, 42, 755-762.

4168 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 22
Sun et al.
(29) Denny, W. A.; Wilson W. R. Bioreductive mustards: a paradigm
for hypoxia-selective prodrugs of diffusible cytotoxins (HPDCs).
Cancer Metastasis Rev. 1993, 12, 135-152.
(30) Keyes, S. R.; Fracasso, P. M.; Heimbrook, D. C.; Rockwell, S.;
Sligar, S. G.; Sartorelli, A. C. Role of NADPH: cytochrome c
reductase and DT-diaphorase in the biotransformation of mito-
mycin C. Cancer Res. 1984, 44, 5638-5643.
(31) Renwick, A. G. Sulfoxides and sulfones. In Sulfur-containing
drugs and related organic compounds; Damani, L. A., Ed.; Ellis
Horwood Ltd.: London, 1989; Vol. 1, Part B, pp 134-154.
(32) Yoshihara, S.; Tatusumi, K. Guinea pig liver aldehyde oxidase
as a sulfoxide reductase: its purification and characterization.
Arch. Biochim. Biophys. 1985, 242, 213-224.
(33) Benaichouche, M.; Bosser, G.; Paris, J .; Plichon, V. Ralative
nucleophilicities of aryldisulphide and thiolate ions in dimethy-
lacetamide estimated from their reaction rates with alkyl
halides. J . Chem. Soc., Perkin Trans. 2 1990, 8, 1421-1424.
(34) Amaral, M. J . S. A. The 2-(p-nitrophenylthio) ethyl group for
carboxy-group protection in peptide synthesis. J . Chem. Soc. C
1969, 2495-2497.
(35) Cain, B. F.; Seelye, R. N.; Atwell, G. J . Potential antitumor
agents. 14. Acridylmethanesulfonanilides. J . Med. Chem. 1974,
17, 922-930.
(36) Gilman, H.; Nobis, J . F. Some aminodibenzothiophenes. J . Am.
Chem. Soc. 1949, 71, 274-276.
(37) Fracasso, P. M.; Sartorelli, A. C. Cytotoxicity and DNA lesions
produced by mitomycin C and porfiromycin in hypoxic and
aerobic EMT6 and Chinese hamster ovary cells. Cancer Res.
1986, 46, 3939-3944.
(38) Skehan, P.; Storeng, R.; Scudiero, D.; Monks, A.; McMahon, J .;
Vistica, D.; Warren, J . T.; Bokesch, H.; Kenney, S.; Boyd, M. R.
New colorimetric cytotoxicity assay for anticancer-drug screen-
ing. J . Natl. Cancer Inst. 1990, 82, 1107-1112.
(39) Kwon, C.-H.; Moon, K.-Y.; Baturay, N.; Shirota, F. N. Chemically
stable, lipophilic prodrugs of phosphoramide mustard as poten-
tial anticancer agents. J . Med. Chem. 1991, 34, 588-592.
J M9904957