Mustard prodrugs for activation by Escherichia coli nitroreductase in gene-directed enzyme prodrug therapy

Article abstract of DOI:10.1021/jm960794l

Twenty nitrogen mustard analogues derived from 5-(aziridin-l-yl)-2,4- dinitrobenzamide (CB 1954, 1) were evaluated as candidate prodrugs for gene- directed enzyme prodrug therapy (GDEPT) in Chinese hamster V79 cell lines engineered to express Escherichia

Full text of DOI:10.1021/jm960794l

1270
J . Med. Chem. 1997, 40, 1270-1275
Mu sta r d P r od r u gs for Activa tion by Esch er ich ia coli Nitr or ed u cta se in
Gen e-Dir ected En zym e P r od r u g Th er a p y
Frank Friedlos, William A. Denny,† Brian D. Palmer,† and Caroline J . Springer*
Cancer Research Campaign Centre for Cancer Therapeutics, Institute of Cancer Research, 15 Cotswold Road, Sutton,
Surrey SM2 5NG, U.K., and Cancer Society Research Laboratory, Faculty of Medicine and Health Sciences,
The University of Auckland, Private Bag 92019, Auckland 1000, New Zealand
Received November 15, 1996X
Twenty nitrogen mustard analogues derived from 5-(aziridin-1-yl)-2,4-dinitrobenzamide (CB
1954, 1) were evaluated as candidate prodrugs for gene-directed enzyme prodrug therapy
(GDEPT) in Chinese hamster V79 cell lines engineered to express Escherichia coli nitrore-
ductase (NR). Structural variations within the series included the use of N-dihydroxypropyl
and (N-dimethylamino)ethyl carboxamide side chains, the use of chloro, bromo, mesyl, and
iodo leaving groups on the mustards, and regioisomeric changes. The compounds were assayed
for cytotoxicity (IC50) with the NR-expressing and controls of non-NR-expressing cell lines.
The proportion of NR-expressing cells required in a mixture for nonexpressing cells to experience
50% of their cytotoxicity (termed the TE50) was used to assess the compounds’ ability to induce
a bystander effect. This study suggests that 5-[N,N-bis(2-bromoethyl)amino]-2,4-dinitroben-
zamide (8), 5-[N,N-bis(2-iodoethyl)amino]-2,4-dinitrobenzamide (9), 2-[N,N-bis(2-bromoethyl)-
amino]-3,5-dinitrobenzamide (13), and 2-[N,N-bis(2-iodoethyl)amino]-3,5-dinitrobenzamide (14)
showed considerable improvements over 1, exhibiting greater potency, higher IC50 ratios, and
lower TE50s, and are thus superior prodrugs to 1 for GDEPT.
In tr od u ction
Recently, the aerobic nitroreductase (NR) isolated
from Escherichia coli B has also been shown to be
capable of activating 1 efficiently,6 although it is unlike
DT diaphorase in size, sequence, and prosthetic group.
Both the 4- and 2-nitro groups of 1 are reduced to the
corresponding hydroxylamines 2 and 3 at equal rates
by the bacterial NR. The 2-hydroxylamine 3, although
more toxic than 1, is not a cross-linking agent and is
not as cytotoxic as 2.3 Nevertheless, the rate of reduc-
tion of 1 is about 60-fold faster by NR than by rat DT
diaphorase (E. coli Kcat ) 360 min-1, rat DT diaphorase
Kcat ) 4.1 min-1),5,6 and thus the more cytotoxic
4-isomer 2 is still produced 30 times faster.
An attractive concept for improving the selectivity of
cancer chemotherapy is by tumor-specific activation of
noncytotoxic prodrugs to active drugs by either endog-
enous or specifically-introduced exogenous enzymes.
One such strategy is gene-directed enzyme prodrug
therapy (GDEPT), where a gene encoding a foreign
enzyme is expressed in target cells. Several methods
for delivery of the genes to the target tumor, under the
control of tumor-selective promoters, have been pro-
posed: liposomes, retroviral vectors, adenoviral vectors,
and cationic lipids.1 These cells are then able to
selectively activate a nontoxic prodrug. The released
drug can then kill the target cells in which it is
generated, together with the cells in the surrounding
vicinity.
A classic example of a prodrug activated by an
endogenous enzyme is CB 1954 [5-(aziridin-1-yl)-2,4-
dinitrobenzamide, 1], which is capable of effecting
complete cures of the Walker 256 rat carcinoma,2
because this tumor expresses a high level of the enzyme
DT diaphorase [NAD(P)H dehydrogenase (quinone) (EC
1.6.99.2)]. This enzyme selectively reduces the 4-nitro
group of 1 to the 4-hydroxylamine 2, which is then
further activated by acetylation to form a cytotoxic DNA
interstrand cross-linking agent.3,4 Other rat cell lines,
expressing similar levels of DT diaphorase, were sensi-
tive to 1 but not human cell lines because human DT
diaphorase is less efficient than the rat enzyme in
activating the prodrug (Kcat ) 0.64 min-1 compared with
4.1 min-1).5 This probably explains why 1 was not
useful clinically (E. Wiltshawe, unpublished informa-
tion).
The dinitrobenzamide mustard SN 23862 (4) under-
goes an even more facile reductive activation than 1 by
the NR enzyme (Kcat ) 1580 min-1 7
) and is not a
substrate for rat DT diaphorase.8 It is reduced by NR
exclusively at the 2-position to give the corresponding
hydroxylamine 5. However, this mechanism of activa-
tion is different from that of 1. Because 4 already
possesses a difunctional alkylating moiety, this reduc-
tive step may be fully activating to a DNA cross-linking
* Address correspondence to this author. Tel: 00-44-(0)181-643-
8901. Fax: 00-44-(0)181-770-7899.
† The University of Auckland.
X Abstract published in Advance ACS Abstracts, March 15, 1997.
S0022-2623(96)00794-7 CCC: $14.00 © 1997 American Chemical Society

Mustard Prodrugs for GDEPT
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 8 1271
Sch em e 1a
a
(i) HN(CH2CH2OH)2/20 °C; (ii) MsCl/Et3N/0 °C; (iii) NaI/
EtOAc/reflux (for 14), NaBr/DMF/20 °C (for 13).
Sch em e 2a
F igu r e 1. Variation in transmission of toxicity from activator
NR-expressing T79-A3 cells to target non-NR-expressing T78-1
cells with percentage of activator cells following treatment with
(A) 1 (b), 4 (9), 8 (2), 9 (1), 11 ([) and (B) 13 (b), 14 (9), 15
(2), 16 (1). The percentage of NR-expressing cells for which
the transmission efficiency was 50% (the TE50) was calculated
by interpolation.
a
(i) SOCl2/reflux, then 3-aminopropane-1,2-diol; (ii) acetone/
HClO4/20 °C; (iii) HN(CH2CH2OH)2/20 °C; (iv) MsCl/Et3N/0 °C,
then NaI/EtOAc/reflux, then 2 N HCl/20 °C.
Sch em e 3a
efficiently to kill surrounding non-enzyme-expressing
tumor cells by a “bystander effect”. In these studies,
particular attention is paid to quantitating such by-
stander effects.
Ch em istr y
Many of the mustards selected for this study have
been reported previously as hypoxia-selective cytotox-
ins.8,10,11 Syntheses of new analogues prepared to
complete the series are outlined in Schemes 1-3. The
iodo and bromo mustards 13 and 14 were prepared via
the diol 22, which was obtained in good yield by reaction
of 2-chloro-3,5-dinitrobenzamide (21) with diethanola-
mine at room temperature (Scheme 1). Iodide or
bromide displacement of the derived dimesylate (1-
aminopropane-2,3-diol) furnished the required mustards
cleanly. The isomeric iodo mustards 11 and 16 bearing
the solubilizing dihydroxypropyl carboxamide side chain
were prepared as shown in Schemes 2 and 3, using an
acetonide protecting group for the side chain diol
functionality during the elaboration of the bis(2-hy-
droxyethyl)amino moiety to the iodo mustard.
a
(i-iv) As for Scheme 2.
species via electron release to the mustard. Some
analogues of 4 are reductively activated by the NR
enzyme at even higher rates.7 For these reasons the
NR enzyme, in conjunction with 4, appeared a good
potential enzyme/prodrug combination for GDEPT. In
this paper we report the synthesis of a series of
analogues of 4 and evaluate their potential as GDEPT
prodrugs in a transfected Chinese hamster cell line
expressing the E. coli NR enzyme. Since it is probable
that only a small proportion of the cells in a tumor will
express the exogenous enzyme,9 it is important that the
activated form of the prodrug be capable of diffusing
Resu lts a n d Discu ssion
The compounds were evaluated for IC50 and TE50 in
T78-1 (NR-negative) and T79-A3 (NR-positive) cells. We
sought to identify compounds that were of similar or

1272 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 8
Ta ble 1. 5-Amino-2,4-dinitrobenzamides
Friedlos et al.
IC50 (µM)
T79-A3
0.89
15
no.
formula
X, Y
R
T78-1
IC50 ratio
TE50
1
4
6
7
8
9
10
11
A
B
B
B
B
B
B
B
CONH2
CONH2
CONH2
CONH2
CONH2
CONH2
158
895
177
63
>295
>110
2532
494
1
3.7
Cl, Cl
Cl, Ms
Ms, Ms
Br, Br
I, I
>1000
>1000
633
3.4
9.3
0.25
0.23
0.1
0.3
71
Cl, Cl
I, I
CONHCH2CHOHCH2OH
CONHCH2CHOHCH2OH
>5000
89
1.9
>55
417
793
0.45
Ta ble 2. 2-Amino-3,5-dinitrobenzamides
IC50 (µM)
no.
X, Y
R
T78-1
>1000
17.5
T79-A3
IC50 ratio
TE50
12
13
14
15
16
Cl, Cl
Br, Br
I, I
Cl, Cl
I, I
CONH2
CONH2
CONH2
7.5
0.19
0.1
13
>133
92
580
75
0.22
0.4
12
58
919
86
CONHCH2CHOHCH2OH
CONHCH2CHOHCH2OH
4
21.5
7
Ta ble 3. 3-Amino-2,6-dinitrobenzamides and Other Structures
greater potency than 1 to T79-A3, had a larger IC50
ratio, and had a lower TE50 (greater bystander effect).
Compounds 4 and 6-11 having mustard substituents
in the same relative positions as the aziridine of 1 (5-
amino-2,4-dinitrobenzamides) constituted the largest
grouping of compounds studied (Table 1). The com-
pounds which satisfied the desired conditions of both
being more cytotoxic than 1 to T79-A3 cells and having
a larger IC50 ratio were the bis-bromo and bis-iodo
carboxamide analogues 8 and 9, respectively. These
same compounds also had the lowest TE50s in the group
and are thus likely to be superior to 1 as GDEPT
prodrugs. The dihydroxypropyl substitution of the
carboxamide yielded compounds which were less toxic
to either cell type, negating the benefits of increased
aqueous solubility. Nevertheless, 11 had a greater IC50
ratio and lower TE50 than 1. Of the remaining com-
pounds of this group, 4 had a poor potency, IC50 ratio,
and TE50, and 6, 7, and 10 were insufficiently potent to
allow a full data set to be collected.
IC50 (µM)
IC50
ratio
no. formula
R
T78-1 T79-A3
17
18
19
20
A
A
B
C
CONH2
>1000
722
>1000
895
CONH(CH2)2NME2
CONH2
0.8
>17
>5
>1000
>1000
58
CONH2
207
dihydroxypropyl substitution of the carboxamide with
respect to reducing potency in 15 and 16 was less than
that seen in 10 and 11 but resulted in a marked increase
in TE50 value, to greater than 1.
In compounds 12-16 the carboxamide was adjacent
to the mustard (2-amino-3,5-dinitrobenzamides, Table
2). These were generally more cytotoxic toward NR-
expressing cells than the corresponding 5-amino-2,4-
dinitro analogues, which was not unexpected from their
larger Kcat values,7 indicating more facile bioreduction.
Only 14 had an IC50 ratio greater than 1, but both 13
and 14 had TE50s lower than 1, similar to their 5-amino-
2,4-dinitrobenzamide counterparts. The effect of the
When the carboxamide was sited between the nitro
groups (3-amino-2,6-dinitrobenzamides 17) (Table 3),
the compound was noncytotoxic at the limits of solubility
to both cell types. If this carboxamide were substituted
with a cationic (dimethylamino)ethyl group, the result-
ing analogue 18 became measurably cytotoxic to both
cell lines but with no difference between them, an effect
expected from previous studies7 which suggested that
such charged analogues are not substrates for NR.

Mustard Prodrugs for GDEPT
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 8 1273
compounds using desorption electron impact ionization at 70
eV or chemical ionization with NH3 as carrier gas. All mustard
products were judged to be >98% pure by reverse-phase HPLC
analysis with diode array detection.
CB 1954 was supplied by Professor M. J arman (ICR), and
compounds 4, 6-10, 12, 15, and 17-20 have been reported
previously.8,10,11
When the 4-nitro group of 4 was replaced by a similarly
electron-withdrawing but nonreducible methanesulfonyl
group, the resulting compound 19 was selectively cyto-
toxic toward the NR-expressing cells, suggesting that
it was a substrate for NR but that the modification
conferred no particular benefit. Replacement of the
aromatic ring in these compounds with a 5-nitrothiazole
heterocycle gave an analogue (20) with a significantly
lower reduction potential compared to 4 (-488 versus
-371 mV, respectively; H. H. Lee, unpublished results).
This compound showed only a modest selectivity for the
NR-expressing cells.
2-[N,N-Bis(2-iod oet h yl)a m in o]-3,5-d in it r ob en za m id e
(14), Sch em e 1. A mixture of 2-chloro-3,5-dinitrobenzamide
(21)10 (1.73 g, 7.04 mmol) and diethanolamine (1.48 g, 14.10
mmol) in p-dioxane (50 mL) and methanol (2 mL) was stirred
at room temperature for 18 h, and the solution was concen-
trated directly onto silica gel and chromatographed. Elution
with ethyl acetate gave 2-[N,N-bis(2-hydroxyethyl)amino]-3,5-
dinitrobenzamide (22) (1.84 g, 83%): mp 143 °C (ethyl acetate/
petroleum ether); 1H NMR [(CD3)2SO] δ 8.66 (d, J ) 2.8 Hz, 1
H, H-4), 8.46 (br s, 1 H, CONH2), 8.34 (d, J ) 2.8 Hz, 1 H,
H-6), 8.12 (br s, 1 H, CONH2), 4.96 (t, J ) 5.5 Hz, 2 H, OH),
3.57 (dt, J ) 5.5, 5.5 Hz, 4 H, CH2OH), 3.21 (t, J ) 5.5 Hz, 4
H, NCH2); 13C NMR δ 167.91 (s), 147.69 (s), 143.06 (s), 140.26
(s), 138.74 (s), 133.38 (s), 128.61 (d), 123.56 (d), 58.00 (t), 54.22
(t). Anal. (C11H14N4O7) CHN.
Methanesulfonyl chloride (0.56 mL, 7.00 mmol) was added
dropwise at 0 °C to a stirred solution of 22 (1.00 g, 3.18 mmol)
and Et3N (1.11 mL, 7.90 mmol) in CH2Cl2 (250 mL). After 10
min, excess reagent was removed by stirring with aqueous
NaHCO3 for 30 min, and the organic layer was separated and
washed with water (four times). The solution was dried and
worked up to give crude dimesylate 23, which was used
directly. This was dissolved in ethyl acetate (100 mL), sodium
iodide (10 g) was added, and the mixture was refluxed with
vigorous stirring for 1 h. Water was added, and the organic
layer was separated, dried, and chromatographed on silica gel.
Elution with ethyl acetate/petroleum ether (1:1) gave 2-[N,N-
bis(2-iodoethyl)amino]-3,5-dinitrobenzamide (14) (0.94 g,
55%): mp 160-161 °C (ethyl acetate/petroleum ether); 1H
NMR [(CD3)2SO] δ 8.73 (d, J ) 2.7 Hz, 1 H, H-4), 8.30 (d, J )
2.7 Hz, 1 H, H-4), 8.30 (d, J ) 2.7 Hz, 1 H, H-6), 8.18, 7.98 (2
× br, 2 H, CONH2), 3.49 (t, J ) 7.3 Hz, 4 H, CH2N), 3.29 (t, J
) 7.3 Hz, 4 H, CH2I); 13C NMR δ 167.14 (s), 145.31 (s), 144.63
(s), 140.97 (s), 136.27 (s), 127.10 (d), 122.10 (d), 54.80 (t), 3.06
(t). Anal. (C11H12I2N4O5) CHN.
The use of NaBr in place of NaI in the above reaction
afforded 2-[N,N-bis(2-bromoethyl)amino]-3,5-dinitrobenzamide
(13) as an orange foam: 1H NMR [(CD3)2SO] δ 8.74 (d, J )
2.7 Hz, 1 H, H-6), 8.21, 7.99 (2 × br, 2 H, CONH2), 3.58, 3.52
(2 × t, J ) 6.2 Hz, 8 H, NCH2CH2Br); 13C NMR δ 167.15 (s),
145.58 (s), 145.26 (s), 141.16 (s), 136.46 (s), 127.17 (d), 122.11
(d), 54.14 (t), 30.01 (t). Found [M + H]+ 442.9225, 440.9238,
438.9258 (CIMS); C11H12Br2N4O5 requires 442.9212, 440.9232,
438.9253.
N-(2,3-Dih ydr oxypr opyl)-5-[N,N-bis(2-iodoeth yl)am in o]-
2,4-d in itr oben za m id e (11), Sch em e 2. 5-Chloro-2,4-dini-
trobenzoic acid (24) (13.00 g, 0.053 mol) was heated under
reflux in thionyl chloride (100 mL) and DMF (1 drop) for 18 h
and then concentrated to dryness under reduced pressure. The
resulting crude acid chloride was dissolved in diethyl ether
(250 mL), cooled to -20 °C, and treated in one portion with a
solution of 1-aminopropane-2,3-diol (9.60 g, 0.105 mol) in water
(25 mL). After a further 30 min at 20 °C, the precipitate was
removed by filtration, washed well with water, and crystallized
from ethyl acetate/petroleum ether to give N-(2,3-dihydrox-
ypropyl)-5-chloro-2,4-dinitrobenzamide (25) (10.44 g, 62%): mp
152 °C; 1H NMR [(CD3)2SO] δ 8.89 (t, J ) 5.7 Hz, 1 H, CONH),
8.82 (s, 1 H, H-3), 8.13 (s, 1 H, H-6), 4.89 (d, J ) 5.0 Hz, 1 H,
CHOH), 3.62 (m, 1 H, CH2OH), 3.47-3.37 (m, 2 H, CH(OH)CH2-
OH), 3.13 (m, 2 H, CONHCH2); 13C NMR δ 162.79 (s), 147.05
(s), 145.00 (s), 136.27 (s), 132.52 (d), 130.23 (s), 122.01 (d), 69.95
(d), 63.66 (t), 42.86 (t). Anal. (C10H10ClN3O4) CHN.
Su m m a r y
The bacterial NR expression system was designed to
evaluate compounds as potential GDEPT prodrugs. The
ratio of the IC50 values in the NR-expressing T79-A3
and nonexpressing T78-1 lines is a composite measure
of the degree of reductive metabolism due to the foreign
enzyme and the relative cytotoxicities of the prodrug
and its reduced (activated) form. The coculture assay
provides a robust and reproducible measure of the
ability of the activated drug to produce a bystander
effect, i.e., to pass from the cells where it is generated
and kill neighboring, nonexpressing cells. An ideal
GDEPT prodrug will possess the properties of stability
to endogenous activation, high aqueous solubility, great
potency of the activated drug, a large differential in
cytotoxicity between prodrug and drug, and a large
bystander effect. The latter two properties are modeled
here by the IC50 ratio and the TE50 value, respectively,
and the potency of the active drug is reflected in the
IC50 to the NR-expressing cells. By these criteria,
modifications to the carboxamide group which increase
solubility are largely detrimental, decreasing cytotox-
icity (thus negating the benefit of the enhanced solubil-
ity) and improving neither the IC50 ratio nor the TE50
.
Regioisomeric changes could sometimes produce small
improvements in these parameters, but these were
modest compared to the advantages gained by manipu-
lations to the mustard in the 5-amino-2,4-dinitroben-
zamide and 2-amino-3,5-dinitrobenzamide series. Com-
pounds 8, 9, 13, and 14 are of greater potency than 1
and also have larger IC50 ratios and lower TE50s. The
best of these (8) requires only 0.1% of the cells to be
expressing NR for the nonexpressing cells to experience
50% of their cytotoxicity. Compound 11 also has a
higher IC50 ratio and lower TE50 than 1 but has a lower
potency. Thus a range of compounds (8, 9, 13, and 14)
have been identified as prodrugs superior to 1 for use
in GDEPT.
Exp er im en ta l Section
Ch em istr y. Analyses were carried out in the Microchemi-
cal Laboratory, University of Otago, Dunedin, NZ. Melting
points were determined on an Electrothermal 9200 melting
point apparatus. NMR spectra were obtained on Bruker AC-
200 or AM-400 spectrometers and are referenced to Me4Si.
Thin-layer chromatography was carried out on aluminum-
backed silica gel (230-400 mesh). Petroleum ether refers to
the fraction boiling at 40-60 °C. Products which were not
crystalline retained trace amounts of EtOAc tenaciously,
making successful combustion analysis difficult. Satisfactory
high-resolution mass spectral data were obtained for these
Concentrated perchloric acid (10 mL) was added dropwise
to a solution of 25 (3.00 g, 9.38 mmol) in acetone (150 mL).
After 30 min the solution was poured into saturated aqueous
NaHCO3 and extracted into EtOAc and the extract worked up
and chromatographed on silica gel. Ethyl acetate eluted the
acetonide 26 as an oil (3.20 g, 95%): 1H NMR (CDCl3) δ 8.62
(s, 1 H, H-3), 7.75 (s, 1 H, H-6), 6.70 (t, J ) 5.5 Hz, 1 H,

1274 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 8
Friedlos et al.
CONH), 4.35 (m, 1 H, CHO), 4.12 (dd, J ) 8.6, 6.6 Hz, 1 H,
CHHO), 3.76 (dd, J ) 8.6, 6.3 Hz, 1 H, CHHO), 3.73 (m, 1 H,
CONHCHN), 3.46 (m, 1 H, CONHCHH), 1.42, 1.35 (2s, 6 H,
C(CH3)2); 13C NMR δ 163.57 (s), 147.42 (s), 144.19 (s), 136.39
(s), 133.19 (s), 132.48 (d), 122.30 (d), 109.63 (s), 73.92 (d), 66.62
(t), 42.56 (t), 26.73 (q), 24.91 (q). Found [M + H]+ 360.0609,
362.0580 (CIMS); C13H14ClN3O7 requires 360.0598, 362.0569.
A solution of 26 (3.20 g, 8.89 mmol) and diethanolamine
(1.96 g, 0.019 mol) in p-dioxane (200 mL) was stirred at 20 °C
for 18 h, and the solution was concentrated directly onto silica
gel and chromatographed. Elution with ethyl acetate gave the
diol 27 (3.41 g, 89%) as an oil: 1H NMR [(CD3)2SO] δ 8.77 (t,
J ) 6.0 Hz, 1 H, CONH), 8.47 (s, 1 H, H-3), 7.30 (s, 1 H, H-6),
4.80 (t, J ) 5.3 Hz, 2 H, OH), 4.19 (m, 1 H, CHO), 4.20 (dd, J
) 8.5, 6.2 Hz, 1 H, CHHO), 3.73 (dd, J ) 8.3, 5.8 Hz, 1 H,
CHHO), 3.57 (dt, J ) 5.6, 5.3 Hz, 4 H, CH2OH), 3.42 (t, J )
5.6 Hz, 4 H, CH2N), 3.38 (m, 2 H, CONHCH2), 1.36, 1.28 (2s,
6 H, C(CH3)3); 13C NMR δ 165.27 (s), 147.74 (s), 136.88 (s),
133.67 (s), 124.62 (d), 119.58 (d), 108.35 (s), 73.80 (d), 66.70
(t), 58.04 (t), 54.08 (t), 41.57 (t), 26.76 (q), 25.32 (q). Found
[M + H]+ 429.1538 (CIMS); C17H24N4O9 requires 429.1621.
42.33 (t), 26.73 (q), 25.20 (q). Found [M + H]+ 429.1650
(CIMS); C17H24N4O9 requires 429.1621.
Finally, halogenation of 31 gave N-(2,3-dihydroxypropyl)-
2-[N,N-bis(2-iodoethyl)amino]-3,5-dinitrobenzamide (16) as a
foam: 1H NMR [(CD3)2SO] δ 8.72 (d, J ) 2.7 Hz, 1 H, H-4),
8.68 (t, J ) 5.7 Hz, 1 H, CONH), 8.32 (d, J ) 2.7 Hz, 1 H,
H-6), 3.65-3.10 (m, 15 H); 13C NMR δ 165.44 (s), 145.17 (s),
144.81 (s), 140.93 (s), 136.33 (s), 127.49 (d), 122.11 (d), 69.92
(d), 63.89 (t), 54.67 (t), 43.00 (t), 3.07 (t). Found [M + H]+
608.9341 (CIMS); C14H18I2N4O7 requires 608.9343.
Biologica l Meth od s. 1. Tr a n sfection a n d P r op er ties
of E. coli NR -E xp r essin g Ch in ese Ha m st er V79 Cells.
Chinese hamster V79 cells were transfected with an expression
vector encoding E. coli NR driven by the human cytomega-
lovirus promoter cloned into a puromycin resistance vector or
with an empty vector. A pair of cell lines expressing (T79-
A3) or not expressing (T78-1) NR was selected. T78-1 cells
were of equal sensitivity to 1 as parental V79 cells, but T79-
A3 cells were approximately 2000 times more sensitive, a
property which was stably maintained for >18 months (Fried-
los et al., unpublished results).
2. Cytotoxicity (IC50) a n d Bysta n d er Effect (TE50
)
To a solution of 27 (2.30 g, 5.37 mmol) in CH2Cl2 (200 mL)
and triethylamine (1.87 mL, 0.013 mol) at -10 °C was added
methanesulfonyl chloride (0.95 mL, 0.012 mmol). After 10
min, saturated aqueous NaHCO3 solution was added, and the
mixture was stirred vigorously for 30 min. The organic layer
was worked up to give crude dimesylate which was dissolved
in ethyl acetate (200 mL) containing NaI (20 g), and the
mixture was refluxed with stirring for 30 min. Water was
added, and the organic layer was worked up to give an oil
which was dissolved in tetrahydrofuran (200 mL) containing
2 N HCl (100 mL). After 2 h at room temperature, the mixture
was diluted with brine and extracted with EtOAc, and the
extract was worked up and chromatographed on silica gel.
EtOAc eluted N-(2,3-dihydroxypropyl)-5-[N,N-bis(2-iodoethyl)-
amino]-2,4-dinitrobenzamide (11) as a yellow oil (1.80 g,
55%): 1H NMR [(CD3)2SO] δ 8.71 (t, J ) 5.8 Hz, 1 H, CONH),
8.52 (s, 1 H, H-3), 7.38 (s, 1 H, H-6), 3.68 (t, J ) 7.0 Hz, 4 H,
CH2N), 3.64 (m, 2 H), 3.43 (m, 1 H), 3.39 (t, J ) 7.0 Hz, 4 H,
CH2I), 3.14 (m, 2 H, CONHCH2); 13C NMR δ 164.48 (s), 145.73
(s), 137.91 (s), 137.22 (s), 136.52 (s), 124.16 (d), 121.11 (d), 70.01
(d), 63.68 (t), 53.03 (t), 42.68 (t), 2.90 (t). Found [M + H]+
608.9337 (CIMS); C14H18I2N4O7 requires 608.9343.
N-(2,3-Dih ydr oxypr opyl)-2-[N,N-bis(2-iodoeth yl)am in o]-
3,5-d in itr oben za m id e (16), Sch em e 3. This was prepared
in an analogous series of reactions and in similar yields from
2-chloro-3,5-dinitrobenzoic acid (28). Amidation of this with
1-aminopropane-2,3-diol as above gave N-(2,3-dihydroxypro-
pyl)-2-chloro-3,5-dinitrobenzamide (29): mp 118-120 °C (EtOAc/
petroleum ether); 1H NMR [(CD3)2SO] δ 8.98 (d, J ) 2.6 Hz, 1
H, H-4), 8.81 (t, J ) 5.7 Hz, 1 H, CONH), 8.56 (d, J ) 2.6 Hz,
1 H, H-6), 4.93 (d, J ) 5.1 Hz, 1 H, CHOH), 4.62 (t, J ) 5.7
Hz, 1 H, CH2OH), 3.65 (m, 1 H, CHOH), 3.45 (m, 1 H,
CONHCHH), 3.38 (m, 2 H, CH2OH), 3.17 (m, 1 H, CONH-
CHH); 13C NMR δ 163.09 (s), 148.43 (s), 145.84 (s), 140.38 (s),
128.35 (s), 126.03 (d), 120.47 (d), 69.94 (d), 63.73 (t), 42.87 (t).
Assa ys. To assess the cytotoxicity and bystander capability
of the compounds, a 96-well plate respreading assay was
designed (Friedlos et al., submitted for publication). In brief,
T78-1 (target) cells were loaded in quadruplicate at 105 cells/
well together with T79-A3 (activator) cells, starting at 4 × 104
cells/well, reducing in 12 stages of 2× to 20 cells/well. After
allowing 4 h to attach, this produced confluent monolayers.
The compounds were prepared by dissolving in DMSO (100/
500 mM) and diluting to 3.7/18.5 mM in medium. Upon
addition to row 1 this yielded an initial concentration of 1/5
mM. Serial dilutions (10 of 3.7×) were performed in situ,
giving a final concentration of 0.0018-0.09 µM. After 24 h
exposure, the drug-containing medium was removed, and the
cells were trypsinized (100 µL). The trypsin was poured out,
and the plates were thoroughly blotted onto a pad of sterile
tissue, thus removing all but a thin film of liquid containing
residual cells. The plates were then refilled with 200 µL of
fresh medium and allowed to grow up. After 4 days growth
(ca. 7-8 generations), the control wells were confluent imply-
ing that the dilution step had been about 100×, and the plates
were fixed and stained with sulforhodamine-B; the extinction
at 590 nm was read, and results are expressed as percentage
of control growth. The IC50s were evaluated by interpolation.
Thus for each compound, one IC50 was generated for each
proportion of activator cells present.
A value termed the transmission efficiency (TE) was calcu-
lated as
0
0
TE ) (IC50 - IC50N)/(IC50 - IC50100) × 100
where the superscripts 0 and 100 are the IC50 values for 0%
and 100% activators, respectively, and superscript N is the
value for each N% activators, and plotted against the percent
activator cells (Figure 1). A single value (TE50) could be
derived for each compound which is the percentage of activa-
tors at which the value of the transmission efficiency is 50%.
The differing shapes of the profiles probably reflect indepen-
dent variation in the parameters contributing to this complex
response.
Reaction of 29 with acetone/perchloric acid gave the ac-
etonide 30 as an oil: 1H NMR (CDCl3) δ 8.68 (d, J ) 2.7 Hz,
1 H, H-4), 8.58 (d, J ) 2.7 Hz, 1 H, H-6), 6.66 (t, J ) 5.2 Hz,
1 H, CONH), 4.37 (m, 1 H, CHO), 4.13 (dd, J ) 8.5, 6.5 Hz, 1
H, CHHO), 3.81-3.74 (m, 2 H, CHHO, CONHCHH), 3.54 (m,
1 H, CONHCHH), 1.44, 1.35 (2s, 6 H, C(CH3)2); 13C NMR δ
163.10 (s), 148.92 (s), 145.90 (s), 139.87 (s), 130.13 (s), 126.47
(d), 121.03 (d), 109.69 (s), 73.85 (d), 66.53 (t), 42.59 (t), 26.74
(q), 24.81 (q). Found [M + H]+ 360.0588, 362.0555 (CIMS);
Ack n ow led gm en t. This work was supported by the
Cancer Research Campaign, U.K. We thank Professor
K. R. Harrap for support and Professor I. Nicolescu-
Duvaz for helpful discussions.
C
13H14ClN3O7 requires 360.0598, 362.0569.
Reaction of 30 with diethanolamine gave the diol 31 as an
oil: 1H NMR [(CD3)2SO] δ 9.14 (t, J ) 5.7 Hz, 1 H, CONH),
8.67 (d, J ) 2.7 Hz, 1 H, H-4), 8.33 (d, J ) 2.7 Hz, 1 H, H-6),
4.93 (t, J ) 5.4 Hz, 2 H, OH), 4.25 (m, 1 H, CHO), 4.04 (dd, J
) 8.4, 6.4 Hz, 1 H, CHHO), 3.69 (dd, J ) 8.3, 5.7 Hz, 1 H,
CHHO), 3.57 (dt, J ) 5.7, 5.4 Hz, 4 H, CH2OH), 3.19 (m, 2 H,
CONHCH2), 1.37, 1.28 (2s, 6 H, C(CH3)2); 13C NMR δ 166.21
(s), 147.81 (s), 143.05 (s), 138.74 (s), 133.06 (s), 128.75 (d),
123.60 (d), 108.52 (s), 73.71 (d), 66.30 (t), 58.07 (t), 54.11 (t),
Refer en ces
(1) Deonarain, M. P.; Spooner, R. A.; Epenetos, A. A. Genetic
delivery of enzymes for cancer therapy. Gene Ther. 1995, 2, 235-
244.
(2) Cobb, L. M.; Connors, T. A.; Elson, L. A.; Khan, A. H.; Mitchley,
B. V. C.; Ross, W. J . C.; Whisson, M. E. 2,4-Dinitro-5-ethylene-
iminobenzamide (CB1954): a potent and selective inhibitor of
the growth of the Walker carcinoma 256. Biochem. Pharmacol.
1969, 18, 1519-1527.

Mustard Prodrugs for GDEPT
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 8 1275
(3) Knox, R. J .; Friedlos, F.; J arman, M.; Roberts, J . J . A new
cytotoxic, DNA interstrand crosslinking agent, 5-aziridin-1-yl)-
4-hydroxylamino-2-nitrobenzamide, is formed from 5-(aziridin-
1-yl)-2,4-dinitrobenzamide (CB1954) by a nitroreductase enzyme
in Walker carcinoma cells. Biochem. Pharmacol. 1988, 37, 4661-
4669.
(4) Knox, R. J .; Friedlos, F.; Marchbank, T.; Roberts, J . J . Bioacti-
vation of CB 1954: Reaction of the Active 4-Hydroxylamino
Derivative with Thioesters to Form the Ultimate DNA-DNA
interstrand Crosslinking Species. Biochem. Pharmacol. 1991, 42,
1691-1697.
(8) Palmer, B. D.; Wilson, W. R.; Cliffe, S.; Denny, W. A. Hypoxia-
selective antitumour agents. 5. Synthesis of water-soluble ni-
troaniline mustards with selective cytotoxicity for hypoxic
mammalian cells. J . Med. Chem. 1992, 35, 3214-3222.
(9) Huber, B. E.; Austin, E. A.; Richards, C. A.; Davis, S. T.; Good,
S. S. Metabolism of 5-fluorocytosine to 5-fluorouracil in human
colorectal tumor cells transduced with the cytosine deaminase
gene: significant antitumour effects when only a small percent-
age of tumour cells express cytosine deaminase. Proc. Natl. Acad.
Sci. 1994, 91, 8302-8306.
(5) Boland, M. P.; Knox, R. J .; Roberts, J . J . The differences in
kinetics of rat and human DT diaphorase result in a differential
sensitivity of derived cells lines to CB1954 (5-(aziridin-1-yl)-2,4-
dinitrobenzamide). Biochem. Pharmacol. 1991, 41, 867-875.
(6) Anlezark, G. M.; Melton, R. G.; Sherwood, R. F.; Coles, B.;
Friedlos, F.; Knox, R. J . The Bioactivation of 5-(aziridin-1-yl)-
2,4-dinitrobenzamide (CB1954) I. Purification and properties of
a nitroreductase enzyme from Escherischia coli - a potential
enzyme for antibody-directed enzyme prodrug therapy (ADEPT).
Biochem. Pharmacol. 1992, 44, 2289-2295.
(7) Anlezark, G. M.; Melton, R. G.; Sherwood, R. F.; Wilson, W. R.;
Denny, W. A.; Palmer, B. D.; Knox, R. J .; Friedlos, F.; Williams,
A. Bioactivation of dinitrobenzamide mustards by an E. Coli B
nitroreductase. Biochem. Pharmacol. 1995, 50, 609-618.
(10) Palmer, B. D.; Wilson, W. R.; Anderson, R. F.; Boyd, M.; Denny,
W. A. Hypoxia-selective antitumor agents. 14. Synthesis and
hypoxic cell cytotoxicity of regioisomers of the hypoxia-selective
cytotoxin 5-[N,N-bis(2-chloroethyl)amino]-2,4-dinitro-benzamide.
J . Med. Chem. 1996, 39, 2518-2528.
(11) Atwell, G. J .; Boyd, M.; Palmer, B. D.; Anderson, R. F.; Pullen,
S. M.; Wilson, W. R.; Denny, W. A. Synthesis and evaluation of
4-substituted analogues of 5-[N,N,-bis(2-chloroethyl)amino]-2-
nitrobenzamide as bioreductively-activated prodrugs using an
E. coli nitroreductase. Anti-Cancer Drug Des. 1996, 11, 553-557.
J M960794L