Hypoxia-selective antitumor agents. 15. Modification of rate of nitroreduction and extent of lysosomal uptake by polysubstitution of 4- (alkylamino)-5-nitroquinoline bioreductive drugs

Article abstract of DOI:10.1021/jm9607865

Studies have shown that 4-(alkylamino)-5-nitroquinolines possess high selectivity (20-60-fold) for hypoxic tumor cells in vitro, but are not active as hypoxia-selective cytotoxins (HSCs) in vivo. The compounds show inadequate rates of extravascular diffus

Full text of DOI:10.1021/jm9607865

J . Med. Chem. 1997, 40, 1381-1390
1381
Hyp oxia -Selective An titu m or Agen ts. 15. Mod ifica tion of Ra te of
Nitr or ed u ction a n d Exten t of Lysosom a l Up ta k e by P olysu bstitu tion of
4-(Alk yla m in o)-5-n itr oqu in olin e Bior ed u ctive Dr u gs
Bronwyn G. Siim,† Graham J . Atwell,‡ Robert F. Anderson,§ Peter Wardman, Susan M. Pullen,‡
William R. Wilson,† and William A. Denny*,‡
Cancer Research Laboratory, Section of Oncology, Department of Pathology, and Department of Chemistry, The University of
Auckland, Private Bag 92019, Auckland, New Zealand, and Gray Laboratory, P.O. Box 100, Mount Vernon Hospital,
Northwood, Middlesex HA6 2J R, U.K.
Received November 12, 1996X
Studies have shown that 4-(alkylamino)-5-nitroquinolines possess high selectivity (20-60-fold)
for hypoxic tumor cells in vitro, but are not active as hypoxia-selective cytotoxins (HSCs) in
vivo. The compounds show inadequate rates of extravascular diffusion, likely due both to
sequestration of the bisbasic compounds into lysosomes and rapid nitroreduction. A further
series of analogues, designed to counteract these limitations, has been synthesized and
evaluated. Analogues bearing one to three electron-donating substituents on the quinoline
have one-electron reduction potentials up to 100 mV lower than that of the unsubstituted
compound (5), but do not have improved biological activity. The relationship between hypoxic
selectivity and rates of metabolic reduction suggests at least two mechanisms of cytotoxicity
for this series of 5-nitroquinolines. Compounds with high rates of reduction are toxic via oxygen-
sensitive net bioreduction, while compounds which are poor substrates for nitroreduction are
toxic through an oxygen-insensitive non-bioreductive mechanism. As rates of metabolic
reduction are lowered, the non-bioreductive mechanism of toxicity becomes dominant and
hypoxic selectivity is lost. A small series of analogues bearing hydrophilic but neutral side
chains were also prepared. Compounds with a dihydroxypropyl side chain retained cytotoxic
potency and hypoxic cell selectivity in cell culture assays, and had lowered uptake into
lysosomes, but none of three analogues evaluated against KHT tumors in mice showed activity
as an HSC in vivo.
Hypoxic cells in tumors are resistant to ionizing
radiation, and probably to some chemotherapeutic
drugs, and are therefore important targets for drug
1-3
development.
A number of different nitrohetero-
cycles have been shown to act as hypoxia-selective
cytotoxins (HSCs), possessing selective toxicity to hy-
poxic mammalian cells by virtue of their oxygen-
inhibited bioreduction to reactive species. Examples
include 2-nitroimidazoles such as misonidazole (1),1
4-nitropyrazoloacridines such as (2),4 and 1-nitroacridines
such as nitracrine (3).5-7 The 1-nitroacridines are
particularly interesting because of their high potency;
3 shows similar hypoxic selectivity to 1 toward AA8 cells
in culture (ca. 10-fold) but is 100000-fold more potent.8
However, 3 showed no activity against hypoxic cells in
solid tumors in vivo.5,9 Major factors contributing to this
were considered to be limited diffusion due to a combi-
nation of intercalative DNA binding (KDNA ) 2.2 × 105
M-1 at 0.01 ionic strength, pH 7, 22 °C)10 and rapid
reductive metabolism related to a high reduction po-
tential (-303 mV).5,7
(KDNA ) 1.5 × 104 M-1 at 0.01 ionic strength, pH 7, 22
°C) and greatly enhanced hypoxic selectivity (the N-
oxide is also bioreducible).10 In an alternative approach,
involving a more substantial structural change, we
explored structure-activity relationships for cytotoxic
potency and hypoxic selectivity among analogous 5-ni-
troquinolines (e.g., 5), possessing one less aromatic ring
in the chromophore.11-13 The nitroquinoline 5 has a
lower DNA binding affinity than 3 (KDNA ) 3.6 × 103
M-1 at 0.03 ionic strength, pH 7, 25 °C)14 but compa-
rable hypoxic selectivity in AA8 cell cultures (ca. 15-
fold). It also showed superior activity to 3 toward EMT6
spheroids (suggesting improved extravascular diffu-
sional properties).12 However, neither 5 nor its 3- and
8-methyl analogues (7 and 9) were active as HSCs or
as radiosensitizers of hypoxic cells in SCCVII tumors.9,13
The 5-nitroquinolines evaluated to date generally
have reduction potentials from -280 to -320 mV, which
Various strategies to lower the reversible DNA bind-
ing affinity of the nitroacridine 3 while retaining
potency and hypoxic selectivity have been explored.
Masking the side chain cationic charge with an N-oxide
group gave a compound (4) of lower binding affinity
† Section of Oncology, Department of Pathology, The University of
Auckland.
‡ Cancer Research Laboratory, The University of Auckland.
§ Department of Chemistry, The University of Auckland.
Mount Vernon Hospital.
X Abstract published in Advance ACS Abstracts, March 15, 1997.
S0022-2623(96)00786-8 CCC: $14.00 © 1997 American Chemical Society

1382 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 9
Siim et al.
Ta ble 1. Physicochemical and Biological Properties of 4-(Alkylamino)-5-nitroquinolines
growth inhibition assaysc clonogenic assays (AA8)d
rate of reductione
((nmol/min)/106
cells)
AA8 (air)
AA8/UV4
ratio
CT10 (air)
air/N2
ratio
E(1)b
IC50 (µM)
(µM h)
a
no.
X
R
pKa
5f
H
(CH2)3NMe2
(CH2)3NMe2
(CH2)3NMe2
(CH2)3NMe2
(CH2)3NMe2
(CH2)3NMe2
(CH2)3NMe2
(CH2)3NMe2
(CH2)3NMe2
(CH2)3NMe2
(CH2)3NMe2
(CH2)3NMe2
(CH2)3NMe2
(CH2)3NMe2
(CH2)3NMe2
6.39 -286
6.41 -274
6.01 -369
6.15 -319
6.84 -316
6.76 -310
4.22 -520
12.4
6.6
33
170
29
14
17
18
9
13
12
2.3
3.8 ( 0.6
6.0 ( 2.3
1.4 ( 0.2
33 ( 13
17 ( 2
2.6 ( 0.2
1.3 ( 0.3
3.0 ( 0.3
28 ( 6
5.9 ( 1.2
660
770
1870
3300
2100
3090
2100
14
22
47
16
60
21
1.3
1.47
1.75
0.28
0.16
0.96
0.89
0.05
0.22
0.48
< 0.03
0.18
0.16
6f 2-Me
7f 3-Me
8f 6-Me
9f 8-Me
10f 8-OMe
6.8
98
11f 8-NHMe
12 6-Et
(6.2)g -294 ( 11h 108 ( 33
3070 ( 410
2960 ( 260
6630 ( 1890 1.1 ( 0.1
3790 ( 260
1780 ( 290
300
3990
830 ( 350
10500
4.8 ( 1.7
33 ( 2
13 2,3-diMe
14 3,6-diMe
15 3,8-diMe
16 6,8-diMe
17 2-Ph,8-Me
18 6,8-diOMe
19 2,3,8-triMe
6.64 -320
47 ( 6
201 ( 20
277 ( 28
6.07 -367 ( 8
6.10 -334 ( 10
20 ( 8
4.3 ( 1.3
>1
4.8
3.7 ( 2.4
6.3
6.37 -329 ( 12h 74 ( 8
5.43 -342 ( 10
6.25 -304 ( 10
6.67 -342
12 ( 2
186 ( 40
113 ( 13
166 ( 25
33 ( 15
20 2,3-diMe,8-OMe (CH2)3NMe2
21
6.64 -374 ( 10
H
CH2CH(OH)CH2OH 6.81 -276
950 ( 100
34 ( 10
0.71
a
b
Quinoline pKa values were determined in aqueous solution at 25 °C by spectrophotometry; see ref 36. E(1) values for one-electron
reduction were determined by pulse radiolysis; see refs 13 and 17. Values with standard errors are new determinations. c Growth inhibition
assay described in the text. IC50 determined against aerobic AA8 or UV4 cells, using an exposure time of 18 h. AA8/UV4(air) ratio )
hypersensitivity factor (HF) ) IC50(AA8)/IC50(UV4), for an 18 h drug exposure under aerobic conditions. AA8(air/N2) ratio ) IC50(AA8:
d
air)/IC50(AA8:N2). Clonogenic assay described in the text. CT10 is the product of the drug concentration (µM) and exposure time (hr)
needed to reduce cell survival to 10% of controls, using AA8 cells at 106/mL in the clonogenic assay (see text). CT10 ratio ) CT10(air)/
CT10(nitrogen). e Rate of one electron reduction measured as oxygen consumption induced by 200 µM drug in cyanide-inhibited cells.
Data from ref 20. f Data from ref 13. Estimated pKa. Data from ref 20, corrected assuming the E(1) for the redox indicator benzyl
g
h
viologen to be -374 mV.21
Sch em e 1a
is probably at the upper end of the desired range. In
this paper we complete our examination of the class of
5-nitro-4-[[(dimethylamino)propyl]amino]nitroquino-
lines by reporting the synthesis and evaluation of a
series of polysubstituted analogues (13-20). The activi-
ties of these compounds, designed to have lower reduc-
tion potentials, are compared with selected monosub-
stituted derivatives prepared previously. It has also
recently been suggested that high lysosomal uptake of
dibasic nitroquinolines may be a major factor limiting
their ability to diffuse in tumor tissue.14 The present
study thus also evaluates the properties of several
analogues (21-25) bearing hydrophilic but neutral side
chains (Table 2). Two alternative basic side chain
variants (26 and 27) were also investigated.
a
(i) EtOCHdC(CO2Et)2/120 °C/1 h, then Dowtherm A/reflux,
then OH-; (ii) benzophenone/275 °C; (iii) POCl3/reflux; (iv) con-
centrated H2SO4/fuming HNO3/0-10 °C/20 min; (v) H2N(CH2)3-
NMe2/100 °C/3.5 h (N2).
Ch em istr y
Most of the required 4-chloroquinolines were known,
and new ones were prepared by either the Gould-
J acobs15 (Scheme 1) or Conrad-Limpach16 procedures.
Preparation of the alkyl-substituted compounds of Table
1 was by nitration of the requisite 4-chloroquinolines
in fuming HNO3/concentrated H2SO4, followed by dis-
placement of the 4-chloro group with excess N,N-
dimethyl-1,3-propanediamine at 100 °C.13 Analogues
21, 23, and 24, containing hydroxylated side chains,
were prepared by a minor variation of this route,
coupling in DMSO. As noted previously,13 reaction of
methoxy-containing 5-nitroquinolines under these con-
ditions resulted in preferential displacement of the nitro
group, and compounds 18, 20, and 25 were thus
prepared by nitration of the preformed 4-[[(dimethy-
lamino)propyl]amino]quinolines (Schemes 2 and 4). The
2-phenyl analogue (17) was prepared (Scheme 3) by
condensation of 2-methyl-5-nitroaniline with ethyl ben-
zoylacetate, followed by thermal cyclization of the
resulting acrylate (35) to give the quinolone (36), which
was elaborated to 17 by standard methods.
One-electron reduction potentials E(1) (in millivolts)
for the nitroquinolines were determined by pulse radi-
olysis using published procedures,13,17 by establishing
a reversible equilibrium against suitable redox indica-
tors. The equilibrium constant K for the reaction was
determined spectrophotometrically for each compound,
•-
and the E(1) for the ArNO2/ArNO2 redox couples
calculated from this are recorded in Tables 1 and 2. The
solubilities of the compounds were determined in both
water and culture medium by UV spectrophotometry;
all the compounds were soluble to concentrations of at
least 45 mM.

Hypoxiaselective Antitumor Agents
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 9 1383
Ta ble 2. Physicochemical and Biological Properties of 5-Nitroquinolines with Neutral or Alternative Basic Side Chains
growth inhibition assaysc
clonogenic assays (AA8)d
AA8 (air)
IC50 (µM)
AA8/UV4
ratio
CT10 (air)
(µM h)
no.
X
R
pKa
E(1)b
-276
air/N2 ratio
a
21
22
23
24
25
26
27
H
H
H
CH2CH(OH)CH2OH
(CH2)3Me
(CH2)3OH
CH2CH(OH)CH2OH
CH2CH(OH)CH2OH
(CH2)2morpholide
NHPhpCONH(CH2)2NMe2
6.81
7.18
7.06
(7.2)e
(7.1)e
6.78
5.08
33 ( 15
9.6 ( 0.4
26 ( 7
110 ( 24
50 ( 8
18 ( 9
13 ( 2
5.9 ( 1.2
>15.3
15 ( 2
3.7 ( 0.2
11 ( 1
11 ( 4
12 ( 3
950 ( 100
350 ( 120
1100
2450 ( 350
2370 ( 250
860 ( 130
440 ( 120
34 ( 10
20 ( 6
10
7.4 ( 2.6
6.2 ( 1.5
10 ( 1
2.1 ( 0.1
-294 ( 10
-265 ( 10
-302 ( 10
-355 ( 11
-278
8-Me
8-OMe
H
H
-352 ( 10
a
b
Quinoline pKa values were determined in aqueous solution at 25 °C by spectrophotometry; see ref 36. E(1) values for one-electron
reduction were determined by pulse radiolysis; see refs 13 and 17. c Growth inhibition assay described in the text. IC50 determined
against aerobic AA8 or UV4 cells, using an exposure time of 18 h. AA8/UV4(air) ratio ) hypersensitivity factor (HF) ) IC50(AA8)/IC50(UV4)
d
for an 18 h drug exposure under aerobic conditions. AA8(air/N2) ratio ) IC50(AA8:air)/IC50(AA8:N2). Clonogenic assay described in the
text. CT10 is the product of the drug concentration (µM) and exposure time (h) needed to reduce cell survival to 10% of controls, using
AA8 cells at 106/mL in the clonogenic assay (see text). CT10 ratio ) CT10(air)/CT10(nitrogen). e Estimated pKa.
Sch em e 2a
Sch em e 4a
a
(i) H2NCH2CHOHCH2OH/phenol/145 °C/1 h; (ii) Ac2O/pyri-
dine/90-100 °C/10 min; (iii) concentrated H2SO4/KNO3/-5 °C/10
min; (iv) MeOH/2 N aqueous Na2CO3/reflux/5 min.
a
(i) POCl3/reflux; (ii) H2N(CH2)3NMe2/excess phenol/140 °C; (iii)
concentrated H2SO4/fuming HNO3/0-10 °C/5 min.
formation of bulky DNA adducts or crosslinks.18 The
ratio of IC50 values in the two cell lines can thus provide
initial information on mechanisms of cytotoxicity.
Sch em e 3a
Hypoxic selectivity was determined by clonogenic
assay, where AA8 cells in the early plateau phase were
exposed to drugs in continuously-gassed stirred suspen-
sion cultures, and survival was assessed at various
times by determining plating efficiency as described
previously.6,8 Cytotoxicity under aerobic and hypoxic
conditions was determined as the CT10 value (the drug
concentration multiplied by the time required to reduce
cell survival to 10% of controls), and the ratio of these
values was used to assess hypoxic selectivity. Cellular
uptake in aerobic cell cultures, both with and without
the lysosomotropic agent ammonium chloride to raise
lysosomal pH,14 was determined by HPLC. The in vivo
activity of compounds 21, 24, and 25 against the aerobic
and hypoxic subpopulations of cells in KHT tumors was
evaluated by clonogenic assay, using ionizing radiation
to selectively kill the oxygenated tumor cells.19
a
(i) Dowtherm A/reflux; (ii) POCl3/reflux; (iii) H2N(CH2)3NMe2/
100 °C/3.5 h (N2).
Biologica l Stu d ies
Aerobic cytotoxicities were determined against both
AA8 and UV4 cells using a growth inhibition assay7 as
IC50 values (the drug concentration required to reduce
cell numbers to 50% of those in untreated controls on
the same 96-well dish). The UV4 cell line is a repair-
defective mutant, derived from AA8, that is g 2-fold
hypersensitive to agents whose cytotoxicity results from
Resu lts a n d Discu ssion
The physicochemical and biological properties of the
compounds are listed in Tables 1 and 2. The data show
that substitution with multiple electron-donating groups
generally results in a lowering of one-electron reduction
potential (measured at pH 7). As has been shown

1384 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 9
Siim et al.
previously,13,22 this does not always occur with aromatic
weak bases like nitroquinolines, where the ability of
electron-donating substituents to lower reduction po-
tential by inductive effects can be counteracted by their
concomitant ability to increase base strength, resulting
in a higher proportion of the ionized form under physi-
ological conditions. However, the majority of the com-
pounds in Table 1 showed little increase in base
strength; for example the 2,3,8-trimethyl derivative (19)
has a pKa little different to that of the 8-methyl
compound (9). Nevertheless, the structure-activity
relationships for reduction potential among the polysub-
stituted 5-nitroquinolines are not straightforward.
Among the monomethyl analogues (6-9), the 3-deriva-
tive showed the lowest E(1). This may relate both to
the relatively low pKa and (possibly) to steric crowding
among the adjacent 3-, 4-, and 5-substituents. While
the monosubstituted derivatives 6-11 showed widely
varying IC50 values in the growth inhibition assay, all,
with the exception of 11, showed substantial hypoxic
selectivity in the clonogenic assay (15-60-fold).13 The
8-methylamino derivative (11) had the lowest E(1)
(-520 mV) of the compounds investigated in this study,
and although it showed reasonable potency in the
clonogenic assay, it lacked hypoxic selectivity.
The 6-Et analogue (12) was prepared to explore the
effects of larger groups ortho to the nitro group, but no
major differences were found between it and the 6-
methyl analogue 8. Both had relatively high E(1)s
(around -300 mV) and moderate cytotoxic potency and
hypoxic selectivity, although 12 was only half as selec-
tive under hypoxic conditions (ratio 6.5-fold). Deriva-
tives with larger groups at the 6-position (iPr, tBu) were
sought, but could not be prepared (nitration of the
corresponding 4-chloro-6-alkylquinolines gave complex
mixtures of products, presumably due to increased steric
hindrance to nitration at the 5-position). The dimethyl
derivatives 13-16 showed a range of E(1) values (from
-320 to -367 mV). The 2,3-dimethyl derivative (13)
showed high selective toxicity under hypoxic conditions
(34-fold), with the 6,8-dimethyl derivative (16) having
hypoxic selectivity about 10-fold lower. The dimethyl
analogue of lowest reduction potential (the 3,6-dimethyl
derivative 14) showed low potency and no hypoxic
selectivity. It has previously been shown20,23 that 14 is
a poor substrate for nitroreduction. The 2,3,8-trimethyl
analogue 19 also had a low reduction potential (-342
mV) and poor hypoxic selectivity (3.7-fold) although its
cytotoxic potency was high.
F igu r e 1. (A) Relationship between aerobic (O) and hypoxic
(b) cytotoxicity and rates of metabolic reduction, measured as
rates of drug-stimulated oxygen consumption in respiration-
inhibited cells. (B) Relationship between hypoxic selectivity
and rates of metabolic reduction. The solid lines are linear
regressions.
potential is often considered a surrogate measure of the
rate of metabolic reduction, and a correlation between
cytotoxic potency and E(1) can therefore be expected.24
However, we have previously reported20,23 for a subset
of these nitroquinolines (5, 7, 9, 11, 14, and 15) that
E(1) is not the sole determinant of rates of metabolic
reduction as steric interactions have a larger effect on
enzymatic reduction than on E(1). We also showed that
the rate of drug-stimulated oxygen consumption in
cyanide-inhibited AA8 cell suspensions is equal to the
rate of net nitroreduction under hypoxia, and that this
is a useful measure of the rate of the initial (one-
electron) enzymatic reduction in this series.20,23 Con-
sistent with this, if the rate of O2 consumption (Table
1) is used as a measure of enzymatic nitroreduction,
there is a significant positive correlation between rates
of nitroreduction and hypoxic cytotoxic potency (Figure
1A):
Because the 8-methoxy derivative (10) showed a
significant drop in reduction potential, relative to the
unsubstituted compound (5), while retaining high po-
tency and good hypoxic selectivity,13 the related com-
pounds 18 and 20 were evaluated. The 2,3-dimethyl-
8-methoxy derivative 20 had the lowest E(1) of the
polysubstituted derivatives (-374 mV) and the lowest
aerobic toxicity in the clonogenic assay. The 6,8-
dimethoxy derivative (18) had similar potency to 10,
although it, and 20, showed significantly lower hypoxic
selectivity.
log10(hypoxic CT10) ) (1.65 ( 0.13) -
(1.15 ( 0.18) log10(rate of reduction) (1)
r ) 0.890, p < 0.001, F ) 42
whereas the relationship between log10(hypoxic cytotoxic
potency) and E(1) is less strong (r ) 0.573, p ) 0.04, F
) 5.4).
This series of 5-nitroquinolines, with a wide range in
E(1) and hypoxic selectivity, provides an opportunity to
examine the relationship between reduction potential,
rate of reduction, and hypoxic selectivity within a
congeneric series. For bioreductive drugs, reduction
In contrast to hypoxic cytotoxicity, aerobic cytotoxic
potency showed less dependence on rates of nitroreduc-
tion as measured by drug-induced O2 consumption
(Figure 1A), the linear regression of log10(aerobic CT10)
versus log10(rate of reduction) giving a lower gradient

Hypoxiaselective Antitumor Agents
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 9 1385
Ta ble 3. Cellular Uptake for 4-Alkylamino-5-nitroquinolines
and Analogues with Neutral Dihydroxy Sidechains
(coefficient -0.36 ( 0.11; r ) 0.700, p ) 0.008, F ) 11).
If both hypoxic and aerobic potency are linearly related
to log10(rate of reduction), their ratio should also be
linearly related to the latter parameter. This is tested
in Figure 1B, which shows a significant linear relation-
ship:
cellular uptake
a
no.
Ci/Ce
Ci/Ce + 50 mM NH4Clb
5
21
9
24
10
25
12 ( 1
1.4 ( 0.1
6.6 ( 0.3
3.2 ( 0.2
4.7 ( 0.2
2.1 ( 0.1
1.1 ( 0.1
7.5 ( 0.7
32 ( 1
6.8 ( 0.1
29 ( 1
1.1 ( 0.1
log10(CT10 air/CT10 N2) ) (1.50 ( 0.14) +
(0.79 ( 0.19) log10(rate of reduction) (2)
a
Ratio of intracellular (Ci) to extracellular (Ce) drug concentra-
r ) 0.777, p ) 0.002, F ) 17
tions determined after 60 min drug exposure (100 µM) in aerobic
b
AA8 cell cultures (5 × 106 cells/mL). Ratio of intracellular (Ci)
to extracellular (Ce) drug concentrations determined in the pres-
The explained variance in eq 2 is relatively low, but
there is no reason to expect the relationship to be strictly
linear. It can be inferred that there is a non-bioreduc-
tive mechanism of toxicity in this series which domi-
nates when the rate of enzymatic reduction is low,
resulting in loss of hypoxic selectivity. The data of
Figure 1B could also be interpreted as showing an
optimal rate of reduction of approximately (0.5 nmol/
min)/106 cells and that hypoxic selectivity also decreases
at higher rates (possibly because redox cycling is then
rapid enough that reactive oxygen species contribute to
aerobic cytotoxicity).
This analysis demonstrates that the lack of hypoxic
selectivity of nitroquinolines which are poorly reduced
(e.g. 11 and 14) arises because of the contribution of a
second mechanism of toxicity, not dependent on meta-
bolic reduction. The existence of a second mechanism
is also consistent with the lack of differential between
AA8 and UV4 cells for compounds 11 and 14 (Table 1)
indicating that they are not acting as DNA alkylating
agents. The presence of a non-bioreductive mechanism
of toxicity presents a fundamental dilemma in the
nitroquinoline series in that improvement of metabolic
stability (to facilitate extravascular diffusion) can be
expected to result in loss of hypoxic selectivity.
Recent studies14 of a series of 4-[(alkylamino)alkyl]-
5-nitroquinolines, including some of the compounds
described here, showed accumulation in cells to high
concentrations in a manner which did not correlate with
DNA binding affinity. Cellular uptake was inhibited
by ammonium chloride, which raises intralysosomal pH,
and it was suggested that high lysosomal uptake of
these bisbasic nitroquinolines may contribute to their
lack of activity as HSCs in vivo, by limiting their ability
to diffuse to the hypoxic regions of tumors.14 Restricted
tissue diffusion of a related dibasic acridine as a result
of lysosomal sequestration has been demonstrated
recently.25 This suggested that analogues bearing
neutral but hydrophilic side chains would be of interest,
and a small series was prepared and evaluated to see
whether analogues with lowered cellular and lysosomal
uptake could be developed (Table 2). Because no
discernible benefit could be seen by the above variations
in reduction potential, compounds 21-23 employed the
5-nitroquinoline nucleus. All three compounds showed
significant hypoxic selectivity in the clonogenic assay,
with the dihydroxypropyl derivative (21) being the best.
The 8-methyl and 8-methoxy analogues (24 and 25)
were therefore evaluated, but showed lower aerobic
potencies and considerably lower hypoxic selectivity.
Variants 26 and 27, bearing other types of basic side
chain, also did not show interesting levels of hypoxic
selectivity.
ence of 50 mM ammonium chloride.
In order to assess whether a neutral side chain lowers
intralysosomal and intracellular accumulation of the
5-nitroquinolines, ratios of intracellular (Ci) to extra-
cellular (Ce) drug concentrations were determined for
the unsubstituted parent compound (5), the 8-methyl
derivative (9), the 8-methoxy derivative (10), and the
corresponding analogues with dihydroxy side chains (21,
24, and 25, respectively) (Table 3). Cellular uptake of
the compounds (100 µM, 1 h) with basic side chains (5,
9, and 10) was high, ranging from a Ci/Ce ratio of 32 for
9 to 12 for 5. Addition of the lysomotropic agent
ammonium chloride (50 mM) substantially inhibited cell
uptake, lowering ratios of Ci/Ce to 3.2 for 9 and 1.4 for
5. These results confirm the previous report14 that
these bisbasic compounds are accumulated inside cells
to high concentrations and that this largely results from
uptake into acidic lysosomal vesicles. Analogues with
neutral side chains (21, 24, and 25) had lower cell
uptake factors that were largely unaffected by the
addition of ammonium chloride. Although ratios of Ci/
Ce were greater than unity for 21 and 24, the lack of
effect of ammonium chloride on uptake suggests a
different site of subcellular localization for these com-
pounds. The results confirm that replacement of a basic
with a neutral side chain lowers accumulation of the
5-nitroquinolines inside lysosomes.
Since compounds with lowered cell uptake are ex-
pected to have improved extravascular diffusion proper-
ties, the activity of the above neutral side chain com-
pounds was evaluated against KHT tumors at 75% of
the maximum tolerated dose (MTD) using single ip
doses. The MTD of 21 was 562 µmol/kg, which was
7-fold higher than that for the corresponding [(N,N-
dimethylamino)alkyl]amino compound (5, MTD 80 µmol/
kg). Similarly, the MTD of 24 (316 µmol/kg) was greater
than for 9 (100 µmol/kg), while the 8-methoxy compound
25 was the least toxic of the dihydroxy side chain
compounds (MTD 1050 µmol/kg). However, none of the
dihydroxy compounds (21, 24, a n d 25) showed signifi-
cant activity against the KHT tumor, either when
administered alone or at various times before or after
gamma irradiation (15 Gy, drug 2 h before to 5 min after
radiation; data not shown), although the related drug
nitracrine (3) does show activity against hypoxic cells
in KHT tumors.26 The lack of in vivo activity of the
compounds with neutral dihydroxy side chains is com-
parable to results for the analogues (5 and 9) with basic
[(dimethylamino)alkyl]amino side chains against SC-
CVII tumors,9,13 indicating that a neutral side chain
does not confer useful in vivo activity.

1386 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 9
Siim et al.
Similar reaction of 2,4-dimethoxyaniline with diethyl
(ethoxymethylene)malonate gave 6,8-dimethoxy-4(1H)-qui-
nolone, mp (EtOH) 223-224 °C (lit.27 mp 223-225 °C).
Treatment of this with POCl3 as above gave 4-chloro-6,8-
Con clu sion s
The above studies were carried out following indica-
tions that the inactivity of 4-[(alkylamino)alkyl]-5-
nitroquinoline analogues as HSCs in vivo, despite good
in vitro selectivities and potencies, resulted from inad-
equate rates of diffusion due to sequestration of the
bisbasic compounds into lysosomes and rapid nitrore-
duction. Although polysubstitution substantially low-
ered reduction potentials relative to the unsubstituted
compound (5), or the 8-methyl derivative (9), this did
not result in improved biological activity or hypoxic
selectivity. Hypoxic selectivity decreased at low rates
of nitroreduction, suggesting at least two mechanisms
of toxicity for these 5-nitroquinolines; oxygen-sensitive
net bioreduction and an oxygen-insensitive non-biore-
ductive mechanism that is responsible for the cytotox-
icity of compounds which are poor substrates for enzy-
matic reduction. The variable contribution of these two
mechanisms may underlie the variations in hypoxic
selectivity within this series of compounds.
The loss of hypoxic selectivity at low reduction
potentials demonstrates a fundamental problem in
attempting to improve extravascular transport by in-
creasing metabolic stability. The alternative strategy
would be to improve diffusion kinetics by suppressing
entrapment of the drugs in cells, so that metabolic
consumption would present less of a problem. Replace-
ment of the cationic side chain of the parent compound
(5) with a neutral dihydroxypropyl unit gave an ana-
logue (21) which retained good hypoxic selectivity in cell
culture. Cellular uptake and intralysosomal accumula-
tion of (21) and the 8-methyl- (24) and 8-methoxy-
substituted (25) analogues was substantially lower than
for the corresponding analogues (5, 9, and 10) with a
basic side chain. However, any possible gain in ex-
travascular diffusion for the compounds with neutral
side chains was not sufficient to provide useful in vivo
activity. This study has identified methods for modify-
ing hypoxic selectivity and cellular uptake in the 5-ni-
troquinolines, but the failure of these strategies to
provide compounds active against hypoxic cells in
tumors suggests that further development of 5-nitro-
quinolines as HSCs is not warranted.
1
dimethoxyquinoline, mp (petroleum ether) 115-115.5 °C; H
NMR (CDCl3) δ 8.63 (d, J ) 4.7 Hz, 1 H, H-2), 7.49 (d, J ) 4.7
Hz, 1 H, H-3), 7.03 (d, J ) 2.5 Hz, 1 H, ArH), 6.76 (d, J ) 2.5
Hz, 1 H, ArH), 4.08 (s, 3 H, OCH3), 3.95 (s, 3 H, OCH3). Anal.
(C11H10ClNO2) C, H, N.
4-Ch lor o-6-eth yl-5-n itr oqu in olin e (31). Exa m p le of
Dir ect Nitr a tion Meth od (Sch em e 1). The above 4-chlo-
roquinoline (30) (3.0 g, 15.7 mmol) was added slowly to stirred
concentrated H2SO4 (13 mL), keeping the temperature below
10 °C. The solution was then cooled to -5 °C, and a mixture
of fuming HNO3 (1.09 g, 17.3 mmol) and concentrated H2SO4
(2 mL) was then added dropwise to the stirred solution at 0
°C. The reaction mixture was stirred for a further 20 min at
10 °C and then poured into ice/ammonia. The mixture was
extracted with CH2Cl2, and evaporation gave a crystalline
residue which was chromatographed on silica gel. Elution
with CH2Cl2/MeOH (32:1) gave 31 (1.92 g, 52%): mp (petro-
leum ether) 54-55 °C; 1H NMR (CDCl3) δ 8.79 (d, J ) 4.7 Hz,
1 H, H-2), 8.22 (d, J ) 8.6 Hz, 1 H, H-8), 7.71 (d, J ) 8.6 Hz,
1 H, H-7), 7.58 (d, J ) 4.7 Hz, 1 H, H-3), 2.75 (q, J ) 7.6 Hz,
2 H, CH2CH3), 1.35 (t, J ) 7.6 Hz, 3 H, CH2CH3). Anal.
(C11H9ClN2O2) C, H, N.
Similar nitration of 4-chloro-2,3-dimethylquinoline28 gave
a mixture which was chromatographed on silica gel. Elution
with petroleum ether/EtOAc (3:1) gave 4-chloro-2,3-dimethyl-
8-nitroquinoline (3.67 g, 37%): mp (petroleum ether) 114-
114.5 °C; 1H NMR δ 8.35 (dd, J ) 8.5, 1.3 Hz, 1 H, H-7), 7.91
(dd, J ) 7.5, 1.3 Hz, 1 H, H-5), 7.59 (dd, J ) 7.5, 8.5 Hz, 1 H,
H-6), 2.78 (s, 3 H, 2-Me), 2.61 (s, 3 H, 3-Me). Anal. (C11H9-
ClN2O2) C, H, N. Elution with petroleum ether/EtOAc (13:7)
gave a mixture of the 5- and 6-nitro isomers. Recrystallization
of this mixture once from petroleum ether and twice from
MeOH gave pure 4-chloro-2,3-dimethyl-5-nitroquinoline (3.47
1
g, 35%): mp 142-142.5 °C; H NMR (CDCl3) δ 8.15 (m, 1 H,
H-8), 7.66 (m, 2 H, H-6,7), 2.79 (s, 3 H, 2-Me), 2.58 (s, 3 H,
3-Me). Anal. (C11H9ClN2O2) C, H, N.
Similar nitration of 4-chloro-3,6-dimethylquinoline29 gave
4-chloro-3,6-dimethyl-5-nitroquinoline (80%): mp (petroleum
ether) 140-140.5 °C; 1H NMR (CDCl3) δ 8.75 (s, 1 H, H-2),
8.1 1 (d, J ) 8.6 Hz, 1 H, H-8), 7.58 (d, J ) 8.6 Hz, 1 H, H-7),
2.57 (s, 3 H, 6-CH3), 2.48 (s, 3 H, 2-CH3). Anal. (C11H9ClN2O2)
C, H, N.
Similar nitration of 4-chloro-3,8-dimethylquinoline30 gave
4-chloro-3,8-dimethyl-5-nitroquinoline (79%): mp (petroleum
ether) 158-159 °C; 1H NMR (CDCl3) δ 8.83 (s, 1 H, H-2), 7.67
(d, J ) 7.7 Hz, 1 H, H-6), 7.55 (dd, J ) 7.7, 0.9 Hz, 1 H, H-7),
2.83 (d, J ) 0.8 Hz, 3 H, 8-CH3), 2.59 (s, 3 H, 2-CH3). Anal.
(C11H9ClN2O2) C, H, N.
Exp er im en ta l Section
Similar nitration of 4-chloro-6,8-dimethylquinoline31 gave
4-chloro-6,8-dimethyl-5-nitroquinoline (88%): mp (petroleum
ether) 113.5-114 °C; 1H NMR (CDCl3) δ 8.78 (d, J ) 4.66 Hz,
1 H, H-2), 7.56 (d, J ) 4.6 Hz, 1 H, H-3), 7.51 (br s, 1 H, H-7),
2.45 (s, 3 H, 8-CH3), 2.79 (d, J ) 0.8 Hz, 3 H, 6-CH3). Anal.
C11H9ClN2O2) C, H, N, Cl.
Analyses were performed by the Microchemical Laboratory,
University of Otago, Dunedin. Melting points were deter-
mined on an Electrothermal Model 9200 digital melting point
apparatus and are reported as read. NMR spectra were
measured on Bruker AM-200 or AM-400 spectrometers and
referenced to Me4Si.
Similar nitration of 4-chloro-2,3,8-trimethylquinoline32 gave
4-chloro-5-nitro-2,3,8-trimethylquinoline (87%): mp (petro-
leum ether) 110-110.5 °C; 1H NMR (CDCl3) δ 7.59 (d, J ) 7.7
Hz, 1 H, H-6), 7.49 (d, J ) 7.7 Hz, 1 H, H-7), 2.82 (s, 3 H,
P r ep a r a tion of 4-Ch lor o-6-eth ylqu in olin e (30): Ex-
a m p le of Gou ld -J a cobs Meth od (Sch em e 1). Equimolar
amounts of 4-ethylaniline and diethyl (ethoxymethylene)-
malonate were heated together at 120 °C for 1 h and then
heated under reflux in Dowtherm A to effect cyclization.15 The
resulting crude ester was saponified to give 6-ethyl-4(1H)-
quinolone-3-carboxylic acid (28) (81% overall yield), mp (DMF)
257 °C dec. Anal. (C12H11NO3) C, H, N. This acid was
decarboxylated in benzophenone at 275 °C to give 6-ethyl-
4(1H)-quinolone (29) (77%), mp (H2O) 197.5-198.5 °C. Anal.
(C11H11NO) C, H, N. The quinolone was treated with refluxing
POCl3 to give 4-chloro-6-ethylquinoline (30) (92%) as an oil
which was used directly: 1H NMR (CDCl3) δ 8.71 (d, J ) 4.7
Hz, 1 H, H-2), 8.05 (d, J ) 8.7 Hz, 1 H, H-8), 8.00 (d, J ) 1.1
Hz, 1 H, H-5), 7.64 (dd, J ) 8.7, 1.1 Hz, 1 H, H-7), 7.46 (d, J
) 4.7 Hz, 1 H, H-3), 2.89 (q, J ) 7.6 Hz, 2 H, CH2CH3), 1.37
(t, J ) 7.6 Hz, 3 H, CH2CH3).
8-Me), 2.78 (s, 3 H, 2-Me), 2.57 (s, 3 H, 3-Me). Anal. (C12H11
ClNO2) C, H, N.
-
Syn th esis of 4-[3-(Dim eth yla m in o)p r op yl]a m in o]-6-
et h yl-5-n it r oq u in olin e (12). E xa m p le of t h e Gen er a l
Rea ction . A mixture of 4-chloro-6-ethyl-5-nitroquinoline (31)
(2.37 g, 10 mmol) and N,N-dimethyl-1,3-propanediamine (5.1
g, 50 mmol) was heated with stirring at 100 °C for 3.5 h under
N2 and then concentrated under reduced pressure below 70
°C. The residue was chromatographed on alumina, eluting
with CH2Cl2/EtOAc (1:1), to give 12 (1.93 g, 64%): 1H NMR
(free base in CDCl3) δ 8.54 (d, J ) 5.4 Hz, 1 H, H-2), 8.02 (d,
J ) 8.7 Hz, 1 H, H-8), 7.51 (d, J ) 8.7 Hz, 1 H, H-7), 6.57 (d,
J ) 5.4 Hz, 1 H, H-3), 5.67 (br s, 1 H, NH), 3.27 (q, J ) 5.7
Hz, 2 H, NHCH2), 2.66 (q, J ) 7.6 Hz, 2 H, CH2CH3), 2.41, t,

Hypoxiaselective Antitumor Agents
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 9 1387
J ) 6.4 Hz, 2 H, CH2N(CH3)2), 2.27 (s, 6 H, N(CH3)2), 1.84
(quintet, J ) 6.4 Hz, 2 H, CH2CH2CH2), 1.30 (t, J ) 7.6 Hz, 3
H, CH2CH3); dihydrochloride salt mp (MeOH/EtOAc) 190-191
°C. Anal. (C16H22N4O2‚2HCl‚0.25H2O) C, N; H: calcd 6.5,
found 7.0.
3-CH3). Anal. (C12H12ClNO) C, H, N, Cl. A mixture of 33 (1
equiv) and N,N-dimethyl-1,3-propanediamine (1.2 equiv) was
heated in excess phenol at 140 °C, and the crude product was
chromatographed on alumina. Elution with EtOAc gave 2,3-
dimethyl-4-[[3-(N,N-dimethylamino)propyl]amino]-8-methox-
yquinoline (34) (64%), mp (petroleum ether) 67-69 °C; 1H
NMR (CDCl3) δ 7.52 (d, J ) 8.6 Hz, 1 H, H-5), 7.27 (t, J ) 7.8
Hz, 1 H, H-6) 6.91 (d, J ) 7.7 Hz, 1 H, H-7), 5.21 (br s, 1 H,
NH), 4.04 (s, 3 H, OCH3), 3.45 (t, J ) 6.2 Hz, 2 H, NHCH2),
2.70 (s, 3 H, 2-CH3), 2.45 (t, J ) 6.3 Hz, 2 H, CH2N(CH3)2),
2.32 (s, 3 H, 3-CH3), 2.28 (s, 6 H, N(CH3)2), 1.81 (quintet, J )
6.3 Hz, 2 H, CH2CH2CH2). Anal. (C17H25N3O) C, H, N.
Nitration of 34 as described below for the preparation of
18, followed by extraction of the crude product with hot
petroleum ether, concentration of the extract and chromatog-
raphy on alumina, eluting with EtOAc, gave 2,3-dimethyl-4-
[[3-(dimethylamino)propyl]amino]-8-methoxy-5-nitroquino-
line (20) as an oil (76%): dihydrochloride salt mp (MeOH/
Similarly prepared were the following.
2,3-Dim et h yl-4-[[3-(d im et h yla m in o)p r op yl]a m in o]-5-
n itr oqu in olin e (13) (reaction time 9 h, 52%): 1H NMR (free
base in CDCl3) δ 8.06 (dd, J ) 8.45, 1.0 Hz, 1 H, H-8), 7.66
(dd, J ) 7.4, 1.1 Hz, 1 H, H-6), 7.52 (dd, J ) 8.4, 7.4 Hz, 1 H,
H-7), 5.38 (br t, 1 H, NH), 3.09 (q, J ) 6.2 Hz, 2 H, NHCH2),
2.66 (s, 3 H, 2-Me), 2.39 (t, J ) 6.3 Hz, 2 H, CH2NMe2), 2.33
(s, 3 H, 3-Me), 2.24 (s, 6 H, NMe2), 1.66 (quintet, J ) 6.2 Hz,
2 H, CH2CH2CH2); dihydrochloride salt mp (EtOAc/MeOH)
152-155 °C. Anal. (C16H22N4O2‚2HCl) C, H, N, Cl.
3,6-Dim et h yl-4-[[3-(d im et h yla m in o)p r op yl]a m in o]-5-
n itr oqu in olin e (14) (gentle reflux, 8 h, 54%): 1H NMR (free
base in CDCl3) δ 8.61 (s, 1 H, H-2), 8.05 (d, J ) 8.6 Hz, 1 H,
H-8), 7.55 (d, J ) 8.6 Hz, 1 H, H-7), 4.49 (br t, J ) 6.6 Hz, 1
H, NH), 3.10 (q, J ) 6.6 Hz, 2 H, NHCH2), 2.44 (s, 3 H, 6-CH3),
2.41 (s, 3 H, 3-CH3), 2.39 (t, J ) 6.6 Hz, 2 H, CH2NMe2), 2.25
(s, 6 H, N(CH3)2), 1.75 (quintet, J ) 6.5 Hz, 2 H, CH2CH2-
CH2); dihydrochloride salt mp (MeOH/EtOAc) 219-222 °C.
Anal. (C16H22N4O2‚2HCl‚H2O) C, N, Cl; H: calcd 6.7, found
7.3.
3,8-Dim et h yl-4-[[3-(d im et h yla m in o)p r op yl]a m in o]-5-
n itr oqu in olin e (15) (reaction time 6 h, 59%): 1H NMR (free
base in CDCl3) δ 8.57 (s, 1 H, H-2), 7.71 (d, J ) 8.3 Hz, 1 H,
H-6), 7.39 (d, J ) 7.5 Hz, 1 H, H-7), 5.90 (s, 1 H, NH), 3.20 (q,
J ) 6.0 Hz, 2 H, NHCH2), 2.78 (s, 3 H, 8-Me), 2.39 (t, J ) 6.0
Hz, 2 H, CH2N(CH3)2), 2.36 (s, 3 H, 3-CH3), 2.25 (s, 6 H,
N(CH3)2), 1.63 (quintet, J ) 6.0 Hz, 2 H, CH2CH2CH2);
dihydrochloride salt mp (MeOH/EtOAc) 212-214 °C. Anal.
(C16H22N4O2‚2HCl‚1.5H2O) C, H, N, Cl.
6,8-Dim et h yl-4-[[3-(d im et h yla m in o)p r op yl]a m in o]-5-
n itr oqu in olin e (16) (reaction time 2 h, 74%): 1H NMR (free
base in CDCl3) δ 8.58 (d, J ) 5.4 Hz, 1 H, H-2), 7.27 (m, 1 H,
H-7), 6.60 (d, J ) 5.4 Hz, 1 H, H-3), 5.68 (s, 1 H, NH), 3.26 (q,
J ) 4.7 Hz, 2 H, NHCH2), 2.74 (s, 3 H, 6-CH3), 2.40 (t, J ) 6.4
Hz, 2 H, CH2N(CH3)2), 2.38 (s, 3 H, 8-CH3), 2.23 (s, 6 H,
N(CH3)2), 1.87 (m, 2 H, CH2CH2CH2); dihydrochloride salt mp
(MeOH/EtOAc) 214-215 °C dec. Anal. (C16H22N4O2.2HCl) C,
H, N, Cl.
4-[[3-(Dim eth yla m in o)p r op yl]a m in o]-5-n itr o-2,3,8-tr i-
m eth ylqu in olin e (19) (reaction time 8 h, 55%): 1H NMR (free
base in CDCl3) δ 7.58 (d, J ) 7.6 Hz, 1 H, H-6), 7.36 (dd, J )
7.6, 0.6 Hz, 1 H, H-7), 5.20 (br s, 1 H, NH), 3.06 (q, J ) 6.2
Hz, 2 H, NHCH2), 2.77 (s, 3 H, 8-Me), 2.68 (s, 3 H, 2-Me), 2.38
(t, J ) 6.3 Hz, 2 H, CH2N(CH3)2), 2.32 (s, 3 H, 3-Me), 2.24 (s,
6 H, N(CH3)2), 1.65 (quintet, J ) 6.3 Hz, 2 H, CH2CH2CH2);
dihydrochloride salt mp (EtOAc/MeOH) 248-251 °C. Anal.
(C17H24N4O2‚2HCl) C, H, N, Cl.
1
EtOAc) 196-197 °C dec; H NMR [(CD3)2SO] δ 13.12 (s, 1 H,
HCl), 11.12 (s, 1 H, HCl), 8.60 (s, 1 H, NHCH2), 8.33 (d, J )
8.7 Hz, 1 H, H-6), 7.49 (d, J ) 8.8 Hz, 1 H, H-7), 4.19 (s, 3 H,
OCH3), 3.35 (q, J ) 6.0 Hz, 2 H, NHCH2), 2.85 (m, 2 H, CH2-
NH+Me2), 2.74 (s, 3 H, 2-CH3), 2.63 (d, J ) 4.6 Hz, collapsing
to singlet on D2O exchange, 6 H, protonated N(CH3)2), 2.37 (s,
3 H, 3-CH3), 2.02 (quintet, J ) 7 Hz, 2 H, CH2CH2CH2). Anal.
(C17H24N4O3‚2HCl‚0.5H2O) C, H, N, Cl.
6,8-Dim eth oxy-4-[[3-(d im eth yla m in o)p r op yl]a m in o]-5-
n itr oqu in olin e (18). Similar condensation of 4-chloro-6,8-
dimethoxyquinoline with N,N-dimethyl-1,3-propanediamine,
as described above for the preparation of 20, followed by
chromatography of the crude product on alumina and elution
with EtOAc/MeOH (10:1), gave 6,8-dimethoxy-4-[[3-(dimethy-
lamino)propyl]amino]quinoline (51%): mp (EtOAc) 148-149
°C; 1H NMR (CDCl3) δ 8.44 (d, J ) 5.2 Hz, 1 H, H-2), 7.47 (br
s, 1 H, NH), 6.66 (d, J ) 2.4 Hz, 1 H, ArH), 6.52 (d, J ) 2.4
Hz, 1 H, ArH), 6.36 (d, J ) 5.2 Hz, 1 H, H-3), 4.02 (s, 3 H,
OCH3), 3.91 (s, 3 H, OCH3), 3.38 (q, J ) 5.7 Hz, 2 H, NHCH2),
2.57 (t, J ) 5.6 Hz, 2 H, CH2N(CH3)2), 2.36 (s, 6 H, N(CH3)2),
1.91 (quintet, J ) 5.7 Hz, 2 H, CH2CH2CH2). Anal. (C16H23
N3O2) C, H, N.
-
The above compound (2.60 g, 9 mmol) was added slowly to
stirred concentrated H2SO4 (18 mL) below 10 °C. A mixture
of fuming HNO3 (0.4 2 mL, 10 mmol) and concentrated H2SO4
(3 mL) was added dropwise to this solution below 0 °C, and
the mixture was stirred for a further 5 min at 0 °C and then
poured into ice. The mixture was basified with excess 40%
aqueous NaOH and extracted with CH2Cl2 (4 × 150 mL). The
combined extracts were evaporated below 30 °C, and the
residue was chromatograped on alumina (Brockmann activity
II-III). Elution with EtOAc/MeOH (49:1) followed by crystal-
lization of the fraction from EtOAc gave 6,8-dimethoxy-4-[[3-
(dimethylamino)propyl]amino]-5-nitroquinoline (18) (38% yield).
1H NMR (free base in CDCl3) δ 8.50 (d, J ) 5.4 Hz, 1 H, H-2),
6.76 (s, 1 H, H-7), 6.60 (d, J ) 5.4 Hz, 1 H, H-3), 5.64 (br s, 1
H, NH), 4.11 & 4.01 (2xs, 2 × 3 H, 6- and 8-OMe), 3.25 (q, J
) 4.5 Hz, 2 H, NHCH2), 2.41 (t, J ) 6.5 Hz, 2 H, CH2N(CH3)2),
2.26 (s, 6 H, N(CH3)2), 1.83 (quintet, 6.4 Hz, 2 H, CH2CH2-
CH2); dihydrochloride salt mp (EtOAc/MeOH) 184-185 °C.
Anal. (C16H22N4O4‚2HCl) C, N, Cl; H; calcd 5.9, found 6.6.
4-[[3-(Dim eth yla m in o)p r op yl]a m in o]-8-m eth yl-5-n itr o-
2-p h en ylqu in olin e (17) (Sch em e 3). Reaction of 2-methyl-
5-nitroaniline and ethyl benzoylacetate in benzene/1-methyl-
2-pyrrolidinone under the conditions of the Conrad-Limpach
procedure gave ethyl 3-phenyl-3-(2-methyl-5-nitroanilino)-
4-[[2-(4-Mor p h olyl)eth yl]a m in o]-5-n itr oqu in olin e (26).
A mixture of 4-chloro-5-nitroquinoline33 (1.04 g, 5 mmol) and
4-(2-aminoethyl)morpholine (3.7 g, 28 mmol) was heated at
95 °C for 20 min, and the cooled mixture was partitioned
between CH2Cl2 and aqueous Na2CO3. Workup of the organic
layer and chromatography of the residue on alumina in EtOAc
gave 26 (1.18 g, 75%): 1H NMR (free base in CDCl3) δ 8.60 (d,
J ) 5.4 Hz, 1 H, H-2), 8.18 (m, 1 H, H-8), 7.61 (m, 2 H, H-6,7),
6.56 (d, J ) 5.4 Hz, 1 H, H-3), 6.03 (br s, 1 H, NH), 3.82 (t, J
) 4.6 Hz, 4 H, N(CH2CH2)2O), 3.25 (m, 2 H, ArNHCH2CH2),
2.71 (t, J ) 5.0 Hz, 2 H, NHCH2CH2), 2.53 (t, J ) 4.4 Hz, 4 H,
N(CH2CH2)2O); dihydrochloride salt mp (EtOAc/MeOH) 189-
191 °C. Anal. (C15H18N4O3‚2HCl) C, N; H: found 6.49, calcd
5.37.
1
acrylate (35) (36%): mp (EtOH/iPr2O) 114-116 °C; H NMR
(CDCl3) δ 8.89 (s, 1 H, NH), 7.67 (dd, J ) 8.3, 2.3 Hz, 1 H,
H-4), 7.3 (m, 5 H, phenyl), 7.26 (d, J ) 8.3 Hz, 1 H, H-3), 7.09
(d, J ) 2.3 Hz, 1 H, H-6), 5.21 (s, 1 H, CdCH), 4.24 (q, J ) 7
Hz, 2 H, CH2CH3), 2.52 (s, 3 H, 2-CH3), 1.35 (t, J ) 7 Hz, 3 H,
CH2CH3). Anal. (C18H18N2O4) C, H, N. Cyclization of 35 in
refluxing Dowtherm A gave 8-methyl-5-nitro-2-phenyl-4(1H)-
quinolone (36) (57%): mp (MeOH) 232-233 °C; 1H NMR
[(CD3)2SO] δ 10.82 (s, 1 H, NH), 7.32-8.34 (m, 7 H, PhH,
H-6,7), 6.38 (s, 1 H, H-3), 2.70 (s, 3 H, CH3). Anal. (C16H12N2O3)
C, H, N. Treatment of 36 with POCl3 gave 4-chloro-8-methyl-
5-nitro-2-phenylquinoline (37) (74%): mp (petroleum ether)
2,3-Dim et h yl-4-[[3-(d im et h yla m in o)p r op yl]a m in o]-8-
m eth oxy-5-n itr oqu in olin e (20) (Sch em e 2). Reaction of
2-methoxyaniline and ethyl 2-methylacetoacetate by the gen-
eral Conrad-Limpach procedure16 gave 2,3-dimethyl-8-meth-
oxy-4(1H)-quinolone34 (32), mp (MeOH) 295-296 °C. Anal.
(C12H13NO2) C, H, N. Reaction of this with POCl3 as usual
gave 4-chloro-2,3-dimethyl-8-methoxyquinoline32 (33): mp
1
(iPr2O) 135 °C; H NMR (CDCl3) δ 7.74 (d, J ) 8.5 Hz, 1 H,
H-5), 7.47 (t, J ) 8.2 Hz, 1 H, H-6), 7.03 (d, J ) 7.8 Hz, 1 H,
H-7), 4.20 (s, 3 H, OCH3), 2.80 (s, 3 H, 2-CH3), 2.55 (s, 3 H,

1388 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 9
Siim et al.
123-123.5 °C; 1H NMR (CDCl3) d 8.22 (m, 2 H, phenyl), 8.08
(s, 1 H, H-3), 7.66 (d, J ) 7.7 Hz, 1 H, H-6) 7.61 (dd, J ) 7.7
Hz, 0.9 Hz, 1 H, H-7), 7.55 (m, 3 H, phenyl), 2.90 (s, 3 H, CH3).
Anal. (C16H11ClN2O2) C, H, N, Cl.
A suspension of 39 (1.49 g, 6 mmol) in pyridine (20 mL)
containing acetic anhydride (2.0 mL, 21 mmol) was heated at
90-100 °C until homogeneous and for a further 10 min.
Volatiles were removed under reduced pressure, and the
residue was dissolved in water and basified with ammonia.
Recrystallization of the resulting precipitate from benzene/
petroleum ether gave 4-[(2,3-diacetoxypropyl)amino]-8-meth-
oxyquinoline (40) (1.56 g, 78% yield): mp 176-177 °C; 1H NMR
[(CD3)2SO] δ 8.37 (d, J ) 5.3 Hz, 1 H, H-2), 7.69 (d, J ) 8.6
Hz, 1 H, H-5), 7.33 (t, J ) 8.1 Hz, 1 H, H-6), 7.17 (t, J ) 5.9
Hz, 1 H, NH), 7.07 (d, J ) 7.7 Hz, 1 H, H-7), 6.64 (d, J ) 5.3
Hz, 1 H, H-3), 5.25 (m, 1 H, CH2CHCH2), 4.1-4.4 (m, 2 H,
CH2O), 3.89 (s, 3 H, OCH3), 3.53 (m, 2 H, NHCH2), 2.03 (s, 3
H, CH3), 2.00 (s, 3 H, CH3). Anal. (C17H20N2O5) C, H, N.
The diacetate 40 (2.99 g, 9 mmol) was added to stirred
concentrated H2SO4 (20 mL) while the temperature was kept
below -5 °C, and the resulting solution was treated portion-
wise with KNO3 (0.96 g, 9.5 mmol). The reaction mixture was
stirred for a further 10 min at -5 °C, poured into excess ice/
ammonia, and extracted with EtOAc (3 × 150 mL). Evapora-
tion of the combined organic layers under reduced pressure
gave the crude nitrodiacetate (41), which was heated under
Coupling of 37 with N,N-dimethyl-1,3-propanediamine as
for the preparation of 12 (reaction time 1 h) gave 4-[[3-
(dimethylamino)propyl]amino]-8-methyl-5-nitro-2-phenylquin-
oline (17) (62%): 1H NMR (CDCl3) δ 8.15-8.23 (m, 2 H, phenyl
H), 7.60 (d, J ) 7.7 Hz, 1 H, H-6), 7.49 (dd, J ) 5.9, 1.5 Hz, 1
H, H-7), 7.40-7.54 (m, 3 H, phenyl H), 7.06 (s, 1 H, H-3), 6.46
(t, J ) 3.6 Hz, 1 H, NH), 3.40 (q, 2 H, NHCH2), 2.86 (s, 3 H,
8-CH3), 2.41 (t, 2 H, CH2N(CH3)2), 2.26 (s, 6 H, N(CH3)2), 2.87
(quintet, 2 H, CH2CH2CH2); dihydrochloride salt mp (MeOH/
EtOAc) 165-168 °C. Anal. (C21H24N4O2‚2HCl‚H2O) C, H, Cl;
N: calcd 12.3, found 11.8.
4-[(3-Hyd r oxyp r op yl)a m in o]-5-n itr oqu in olin e (23). A
mixture of 4-chloro-5-nitroquinoline (1.04 g, 5 mmol), 3-ami-
nopropanol (0.98 g, 13 mmol), and dry DMSO (4 mL) was
heated at 95-100 °C for 45 min under N2. The mixture was
then cooled, diluted with 2 N aqueous Na2CO3, and extracted
with CH2Cl2 (2 × 100 mL). The combined extracts were
washed with saturated NaCl, dried, and evaporated, and the
residue was chromatographed on alumina. Elution with
EtOAc/MeOH (19:1) gave 23 (0.91 g, 74%): 1H NMR (free base
in CDCl3) δ 8.55 (d, J ) 5.5 Hz, 1 H, H-2), 8.16 (d, J ) 7.9 Hz,
1 H, H-8), 7.54-7.64 (m, 2 H, H-6,7), 6.57 (d, J ) 5.5 Hz, 1 H,
H-3), 5.48 (s, 1 H, NH), 3.87 (t, J ) 5.5 Hz, 2 H, CH2OH), 3.37
(q, J ) 5.7 Hz, 2 H, NHCH2), 2.55 (br s, 1 H, OH), 1.97 (quintet,
J ) 5.9 Hz, 2 H, CH2CH2CH2); hydrochloride salt mp (MeOH/
EtOAc) 195-196 °C. Anal. (C12H13N3O3‚HCl) C, H, N, Cl.
Similar reaction of 4-chloro-5-nitroquinoline with 3-amino-
1,2-propanediol (95 °C, 1 h) and crystallization of the crude
product from EtOAc gave 4-[(2,3-dihydroxypropyl)amino]-5-
nitroquinoline (21) (66%): 1H NMR (free base in CDCl3) δ 8.59
(d, J ) 5.4 Hz, 1 H, H-2), 8.12 (d, J ) 8.4 Hz, 1 H, H-8), 7.94
(d, J ) 7.5 Hz, 1 H, H-6), 7.75 (dd, J ) 7.5, 8.4 Hz, 1 H, H-7),
6.82 (d, J ) 5.4 Hz, 1 H, H-3), 5.67 (br s, 1 H, exchange with
D2O, NH), 5.14 (d, J ) 4.4 Hz, 1 H, exchange with D2O,
CHOH), 4.75 (t, J ) 5.2 Hz, 1 H, exchange with D2O, CH2OH),
3.70-3.80, 3.25-3.50 and 3.10-3.20 (3m, 5 H, CH2CHOHCH2-
OH); hydrochloride salt mp (EtOAc/MeOH) 153-156 °C. Anal.
(C12H13N3O4‚HCl) C, H, N, Cl.
Similar reaction of 4-chloro-8-methyl-5-nitroquinoline with
3-amino-1,2-propanediol (95 °C, 2 h) gave 4-[(2,3-dihydrox-
ypropyl)amino]-8-methyl-5-nitroquinoline (24) (71%): 1H NMR
[(CD3)2SO] δ 8.59 (d, J ) 5.3 Hz, 1 H, H-2), 7.84 (d, J ) 7.7
Hz, 1 H, H-6), 7.62 (d, J ) 7.7 Hz, 1 H, H-7), 6.83 (d, J ) 5.4
Hz, 1 H, H-3), 5.66 (t, J ) 4.3 Hz, 1 H, NH), 5.10 (d, J ) 4.3
Hz, 1 H, CHOH), 4.75 (t, J ) 5.1 Hz, 1 H, CH2OH), 3.69-
3.78, 3.23-3.48, and 3.10-3.19 (3m, 5 H, CH2CHOHCH2OH),
2.70 (s, 3 H, 8-Me); hydrochloride salt mp (MeOH/EtOAc) 175
°C dec. Anal. (C13H15N3O4‚HCl) C, H, N, Cl.
Similar reaction of 4-chloro-5-nitroquinoline with buty-
lamine (100 °C, 1.5 h) gave 4-(butylamino)-5-nitroquinoline
(22) (79%): 1H NMR (CDCl3) δ 8.61 (d, J ) 5.4 Hz, 1 H, H-2),
8.18 (dd, J ) 8.2 Hz, 1.5 Hz, 1 H, H-8), 7.57-7.67 (m, 2 H,
H-6,7), 6.62 (d, J ) 5.5Hz, 1 H, H-3), 5.09 (s, 1 H, NH), 3.22
(q, J ) 6.2 Hz, 2 H, NHCH2), 1.70 (quintet, J ) 7.3 Hz, 2 H,
CH2CH2CH2), 1.47 (sextet, J ) 7.4 Hz, 2 H, CH2CH3), 0.99 (t,
J ) 7.4 Hz, 3 H, CH3); hydrochloride salt mp (MeOH/EtOAc)
214-215 °C. Anal. (C13H15N3O2‚HCl) C, H, N, Cl.
reflux in a mixture of MeOH (12 mL) and 2 N aqueous Na2
-
CO3 (20 mL) for 5 min. The mixture was concentrated to small
volume and cooled, and the resulting precipitate was collected,
washed with water, and dried. Recrystallization from MeOH/
EtOAc gave 4-[(2,3-dihydroxypropyl)amino]-8-methoxy-5-nit-
roquinoline (25) (1.18 g, 45% yield):
1H NMR [(CD3)2SO] δ 8.50
(d, J ) 5.4 Hz, 1 H, H-2), 8.04 (d, J ) 8.7 Hz, 1 H, H-6), 7.15
(d, J ) 8.7 Hz, 1 H, H-7), 6.86 (d, J ) 5.4 Hz, 1 H, H-3), 5.77
(m, 1 H, exchange with D2O, NH), 5.09 (d, J ) 3.5 Hz, 1 H,
exchange with D2O, CHOH), 4.77 (m, 1 H, exchange with D2O,
CH2OH), 4.02 (s, 3 H, OMe), 3.68-3.79, 3.25-3.50, 3.10-3.20
(3m, 5 H, CH2CHOHCH2OH); hydrochloride salt mp (MeOH/
EtOAc) 171.5-172 °C. Anal. (C13H15N3O5‚HCl) C, H, N, Cl.
4-[[4-[N-[2-(Dim eth yla m in o)eth yl]ca r ba m oyl]p h en yl]-
a m in o]-5-n itr oqu in olin e (27). A solution of 4-chloro-5-
nitroquinoline (0.65 g, 3.1 mmol) and N-[2-(dimethylamino)-
ethyl]-4-aminobenzamide (0.71 g, 3.4 mmol) in MeOH (30 mL)
was treated with concentrated HCl (0.6 mL) and heated under
reflux for 2 h. The mixture was diluted with EtOAc, and the
resulting precipitate was recrystallized from MeOH/EtOAc to
give 27 as the dihydrochloride salt (1.01 g, 72%): mp 181-
1
183 °C; H NMR (free base in CDCl3) δ 8.72 (d, J ) 4.6 Hz, 1
H, H-2), 8.31 (d, J ) 8.3 Hz, 1 H, H-8), 7.82 (d, J ) 7.6 Hz, 2
H, H-3′,5′), 7.82 (br d, 1 H, H-6), 7.72 (t, J ) 7.8 Hz, 1 H, H-7),
7.36 (d, J ) 4.6 Hz, 1 H, H-3), 7.16 (d, J ) 7.9 Hz, 2 H, H-2′,6′),
6.92 (s, 1 H, NH), 6.84 (s, 1 H, NH), 3.54 (d, J ) 4.7 Hz, 2 H,
NHCH2), 2.53 (s, 2 H, CH2N(CH3)2), 2.28 (s, 6 H, N(CH3)2).
Anal. (C20H21N5O3‚2HCl‚H2O) C, H, N, Cl.
Red u ction P oten tia ls. Pulse radiolysis experiments were
carried out as described previously,13 using 3 mM drug
solutions in phosphate buffer at pH 7, using either 2-propanol
(typically 0.2 M) or 2-propanol/acetone mixtures as cosolvent.
Redox indicators used were methylviologen (1,1′-dimethyl-4,4′-
bipyridinium dichloride; E(1) ) -447 mV), triquat (7,8-dihydro-
6H-dipyrido[1,2-a:2′,1′-c][1,4]diazepinediium dibromide; E(1)
) -548 mV), or benzylviologen (1,1′-dibenzyl-4,4′-bipyridinium
dichloride; E(1) ) -374 mV), as appropriate. Measured
reduction potentials were corrected for ionic strength effects
using the pKa
values of Tables 1 and 2.
4-[(2,3-Dih ydr oxypr opyl)am in o]-8-m eth oxy-5-n itr oqu in -
olin e (25) (Sch em e 4). A mixture of 4-chloro-8-methox-
yquinoline (38) (1.94 g, 10 mmol), 3-amino-1,2-propanediol
(1.09 g, 12 mmol), and dry phenol (6g) was heated at 145 °C
for 1 h, and then excess phenol was removed under reduced
pressure. The residue was dissolved in water, treated with
charcoal, clarified by filtration, and basified with ammonia to
provide 4-[(2,3-dihydroxypropyl)amino]-8-methoxyquinoline (39)
(2.16 g, 87% yield: mp (MeOH) 255-256 °C; 1H NMR [(CD3)2-
SO] δ 8.33 (d, J ) 4.8 Hz, 1 H, H-2), 7.69 (d, J ) 8.2 Hz, 1 H,
H-5), 7.31 (t, J ) 7.9 Hz, 1 H, H-6), 7.06 (d, J ) 7.6 Hz, H-7),
6.91 (s, 1 H, NH), 6.51 (d, J ) 4.9 Hz, 1 H, H-3), 4.94 (br s, 1
H, CHOH), 4.71 (br s, 1 H, CH2OH), 3.89 (s, 3 H, OCH3),
3.13-3.84 (m, 5 H, CH2CHOHCH2OH). Anal. (C13H16NO3)
C, H, N.
Cell Cu ltu r e Assa ys. AA8 and UV4 cells were maintained
in logarithmic-phase growth in 25 cm3 tissue culture flasks
with subculture twice weekly by trypsinization. The growth
medium was antibiotic-free Alpha MEM with 10% v/v heat-
inactivated (56 °C, 40 min) fetal calf serum. Doubling times
were approximately 14 h for AA8 and 15 h for UV4 cells. Bulk
cultures of AA8 cells were prepared in spinner flasks, using
the above growth medium plus penicillin (100 IU/mL) and
streptomycin (100 µg/mL). Growth inhibition studies were
performed as described in detail elsewhere,7,34 using 200 viable
AA8 or 300 viable UV4 cells plus 5000 lethally-irradiated AA8
feeder cells per well in 96-well tissue culture dishes. The IC50
was determined as the drug concentration needed to reduce
the cell mass (protein content, measured after 72-78 h by
staining with methylene blue and measuring absorbance in a

Hypoxiaselective Antitumor Agents
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 9 1389
(6) Wilson, W. R.; Anderson, R. F.; Denny, W. A. Hypoxia-selective
antitumor agents. 1. Relationships between structure, redox
properties and hypoxia-selective cytotoxicity for 4-substituted
derivatives of nitracrine. J . Med. Chem. 1989, 32, 23-30.
(7) Wilson, W. R.; Thompson, L. H.; Anderson, R. F.; Denny, W. A.
Hypoxia-selective antitumor agents. 2. Electronic effects of
4-substituents on the mechanisms of cytotoxicity and metabolic
stability of nitracrine analogues. J . Med. Chem. 1989, 32, 31-
38.
(8) Wilson, W. R.; Denny, W. A.; Twigden, S. J .; Baguley, B. C.;
Probert, J . C. Selective toxicity of nitracrine to hypoxic mam-
malian cells. Br. J . Cancer 1984, 49, 215-223.
(9) Wilson, W. R.; Siim, B. G.; Denny, W. A.; van Zijl, P. L.; Taylor,
M. L.; Chambers, D. M.; Roberts, P. B. 5-Nitraquine, a new DNA-
affinic hypoxic cell radiosensitizer : comparison with nitracrine.
Radiat. Res. 1992, 131, 257-265.
(10) Wilson, W. R.; van Zijl, P.; Denny, W. A. Bis-bioreductive agents
as hypoxia-selective cytotoxins : N-oxides of nitracrine. Int. J .
Radiat. Oncol. Biol. Phys. 1992, 22, 693-697.
(11) Stefanska, B.; Zirra, J . A.; Peryt, J .; Kaminski, K.; Ledochowski,
A. Research on tumour inhibiting compounds. Part LII. 5- and
8-nitro-4-aminoquinoline derivatives. Rocz. Chem. 1973, 47,
2339-2343.
(12) Denny, W. A.; Wilson, W. R.; Atwell, G. J .; Boyd, M.; Pullen, S.
M.; Anderson, R. F. Nitroacridines and nitroquinolines as DNA-
affinic hypoxia-selective cyotoxins. In Activation of Drugs by
Redox Processes; Adams, G. E., Breccia, A., Fielden, E. M.,
Wardman, P., Eds.; NATO Advanced Study Series; Plenum
Press: New York, 1991; pp 149-158.
(13) Denny, W. A.; Atwell, G. J .; Roberts, P. B.; Anderson, R. F.; Boyd,
M.; Lock, C. J . L.; Wilson, W. R. Hypoxia-selective antitumor
agents. 6. 4-Alkylaminonitroquinolines, a new class of hypoxia-
selective cytotoxic agents. J . Med. Chem. 1992 35, 4832-4841.
(14) Siim, B. G.; Denny, W. A.; Wilson, W. R. Does DNA targeting
affect the cytotoxicity and cell uptake of basic nitroquinoline
bioreductive drugs? Int. J . Radiat. Oncol. Biol. Phys. 1994, 29,
311-315.
(15) Price, C. C.; Roberts, R. M. The synthesis of 4-hydroxyquinolines.
I. Through ethoxymethylenemalonic ester. J . Am. Chem. Soc.
1946, 68, 1204-1208.
(16) Hauser, C. R; Reynolds, G. A. Reactions of â-keto esters with
aromatic amines. Synthesis of 2- and 4-hydroxyquinoline deriva-
tives. J . Am. Chem. Soc. 1948, 70, 2402-2404.
(17) Van Leewen, J . W.; Van Dijk, C.; Veeger, C. A pulse radiolysis
study of the reduction of flavodoxin from Megasphaera elsdenii
by viologen radicals. A conformational change as a possible
regulating mechanism. Eur. J . Biochem. 1983, 135, 601-607.
(18) Hoy, C. A.; Salazar, E. P.; Thompson, L. H. Rapid detection of
DNA-damaging agents using repair-deficient CHO cells. Mutat.
Res. 1984, 130, 321-332.
(19) Hill, R. P.; Gulyas, S.; Whitmore, G. F. Studies of the in vivo
and in vitro cytotoxicity of the drug RSU-1069. Br. J . Cancer
1986, 53, 743-751.
(20) Siim, B. G.; Wilson, W. R. Efficient redox cycling of nitroquinoline
bioreductive drugs due to aerobic nitroreduction in Chinese
hamster cells. Biochem. Pharmacol. 1995, 50, 75-82.
(21) Wardman, P. The reduction potential of benzyl viologen: An
important reference compound for oxidant/radical redox couples.
Free Radical Res. Commun. 1991, 14, 57-67.
(22) Boyd, M.; Boyd, P. D. W.; Atwell, G. J .; Wilson, W. R.; Denny,
W. A. Crystallographic and oxygen-17 NMR studies of nitro
group torsion angles in a series of 4-alkylaminonitroquinolines
designed as hypoxia-selective cytotoxins. J . Chem. Soc., Perkin
Trans. 2 1992, 579-584.
(23) Siim, B. G.; Atwell, G. A.; Wilson, W. R. Metabolic and radiolytic
reduction of 4-alkylamino-5-nitroquinoline bioreductive drugs:
Relationship to hypoxia-selective cytotoxicity. Biochem. Phar-
macol. 1994, 48, 1593-1604.
(24) Adams, G. E.; Stratford, I. J .; Wallace, R. G.; Wardman, P.;
Watts, M. E. Toxicity of nitro compounds toward hypoxic
mammalian cells in vitro: dependence on reduction potential.
J . Natl. Cancer Inst. 1980, 64, 555-560.
microplate photometer) to 50% of the mean value for eight
control cultures on the same 96-well plate.
Clonogenic assays with magnetically-stirred 10 mL suspen-
sion cultures (plateau-phase AA8 cells, 106/mL) were per-
formed by removing samples periodically during continuous
gassing with 5% CO2 in air or N2 as detailed elsewhere.8 Both
cell suspensions and drug solutions in growth medium were
pre-equilibrated under the appropriate gas phase for 60 min
prior to mixing, to ensure essentially complete anoxia through-
out the period of drug contact in hypoxic cultures. Several
drug concentrations were investigated for each agent, and the
concentration × time required to reduce the surviving cell
fraction to 10% (CT10) was determined at each concentration.
CT10 values were not strictly constant, with a trend toward
lower values at higher drug concentrations. To minimize
errors due to this, comparisons between aerobic and hypoxic
cytotoxicity were based on CT10 values at concentrations which
gave similar rates of cell killing (usually one log in about 1 h).
Cell Up ta k e Stu d ies. Cellular uptake was determined in
aerobic cultures of AA8 cells (5 × 106 cells/mL) equilibrated
with or without 50 mM ammonium chloride at 37 °C. Intra-
cellular and extracellular drug concentrations were determined
60 min after addition of 100 µM drug by HPLC following rapid
centrifugation of triplicate 2 mL samples. Intracellular samples
were prepared for HPLC analysis as described previously,23
while aliquots of extracellular medium were injected directly.
The HPLC system has been described in detail elsewhere.23
The mobile phase consisted of formate buffer, pH 4.5 (0.44 M
ammonium formate, 68 mM formic acid), and 80% MeCN.
Compounds were eluted isocratically with the following mobile
phase composition: 90:10 formate buffer/80% MeCN (v/v) for
5, 9, 21, and 24; 88:12 formate buffer/80% MeCN (v/v) for 10
and 25. Quantitation was based on peak areas using absor-
bance at the following wavelength ranges (signal (nm), band-
width (nm)) referenced to absorbance at 550 nm: (370, 10) for
5 and 21; (260, 4) for 9 and 24; (242, 4) for 10 and 25.
Intracellular drug concentrations were corrected for measured
recovery efficiency determined as described previously.23
An titu m or Activity. MTD values were determined in
groups of six male C3H/HeN mice with single ip doses of drug
in water (0.01 mL/g of body weight). The dose was escalated
at 1.33-fold dose intervals, with an observation time of 30 days.
Groups of two mice bearing bilateral KHT tumors in the
gastrocnemius muscle were treated with ip drug at 0.75 of the
MTD either alone, or at various times before or after whole
body irradiation (cobalt 60, 15 Gy, ca. 2.5 Gy/min) of unanes-
thesized, unrestrained mice. Three or four tumors within the
size range 0.5-1.0 mL at treatment were removed from each
group 18 h later, pooled, and assayed for clonogenicity in agar
essentially as described,19 except that tumors were dissociated
using an enzyme cocktail (0.5 mg/mL pronase, 0.2 mg/mL
collagenase, 0.2 mg/mL DNAase I) for 40 min, and total
nucleated cells were counted with an electronic particle
counter.
Ack n ow led gm en t. This work was supported by the
Auckland Division of the Cancer Society of New Zealand
and by Contract CM-47019 from the U.S. National
Cancer Institute.
Refer en ces
(1) Adams, G. E.; Stratford, I. J . Hypoxia-mediated nitroheterocyclic
drugs in the radio- and chemotherapy of cancer. Biochem.
Pharmacol. 1986, 35, 71-76.
(2) Brown, J . M.; Siim, B. G. Hypoxia-specific cytotoxins in cancer
therapy. Seminars Radiat. Oncol. 1996, 6, 22-36.
(3) Denny, W. A., Wilson, W. R.; Hay, M. P. Recent developments
in the design of bioreductive drugs. Br. J . Cancer 1996, 74
(Suppl. XXVII), S32-S38.
(4) Sebolt, J . S.; Scavone, S. V.; Pinter, C. D.; Hamehehle, K. L.;
Von Hoff, D. D.; J ackson, R. C. Pyrazoloacridines, a new class
of anticancer agents with selectivity against solid tumors in vivo.
Cancer Res. 1987, 47, 4299-4304.
(5) Wilson, W. R.; Denny, W. A.; Stewart, G. M.; Fenn, A.; Probert,
J . C. Reductive metabolism and hypoxia-selective cytotoxicity
of nitracrine. Int. J . Radiat. Oncol. Biol. Phys. 1986, 12, 1235-
1238.
(25) Hicks, K. O.; Ohms, S. J .; van Zijl, P. L.; Zwi, L. J .; Denny, W.
A.; Hunter, P. J .; Wilson, W. R. Quantitation of the extravascular
diffusion of a DNA intercalator using multicellular membranes
as an in vitro model. Br. J . Cancer, submitted.
(26) Lee, H. H.; Wilson, W. R.; Ferry, D. M.; van Zilj, P.; Pullen, S.
M.; Denny, W. A. Hypoxia-selective antitumor agents. 13. Effects
of acridine substitution on the hypoxia-selective cytotoxicity and
metabolic reduction of the bis-bioreductive agent nitracrine
N-oxide. J . Med. Chem. 1996, 39, 2508-2517.
(27) Pharmaceutical compositions containing quinoline compounds.
British Patent 942,524 (Norwich Pharmacal. Co.), 1963; Chem.
Abstr. 1964, 60, 5468f.
(28) Wohnlich, E. 2-Aminopropiophenone, its acyl derivatives, and
their condensation to 2-and 4-hydroxyquinolones. Arch. Pharm.
1913, 251, 526-550.

1390 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 9
Siim et al.
(29) Steck, E. A.; Hallock, L. L.; Holland, A. J . Quinolines. I. The
synthesis of 3-methyl-4-(1′-methyl-4′-diethylaminobutylamino)-
quinoline and some 6-substituted derivatives. J . Am. Chem. Soc.
1946, 68, 129-132.
(33) Gouley, R. W.; Moersch, G. W.; Mosher, H. J . Heterocyclic basic
compounds. X, 4,8-diaminoquinoline and derivatives. J . Am.
Chem. Soc. 1947, 69, 303-306.
(34) Mathur, K. C.; Sharma, H. N. Synthesis of possible antiamoebic
agents. VIII. 2-[2-(Dimethylamino)ethyl]-3-alkyl-8-methoxyquin-
olines. Indian J . Appl. Chem. 1972, 35, 71-72.
(35) Finlay, G. J .; Baguley, B. C.; Wilson, W. R. A semiautomated
microculture method for investigating growth inhibitory effects
of cytotoxic compounds on exponentially growing carcinoma cells.
Anal. Biochem. 1984, 139, 272-277.
(36) Cain, B. F.; Atwell, G. J .; Denny, W. A. Potential Antitumor
Agents. Part 16. 4′-(Acridin-9-ylamino)methanesulfonanilides.
J . Med. Chem. 1975, 18, 1110-1117.
(30) Steck, E. A.; Hallock, L. L.; Holland, A. J . Quinolines. II. The
synthesis of 8-substituted 3-methyl-4-(1′-methyl-4′-diethylami-
nobutylamino)quinolines. J . Am. Chem. Soc. 1946, 68, 132-133.
(31) Burckhalter, J . H.; Tendick, F. H.; J ones, E. M.; J ones, P. A.;
Holcomb, W. F.; Rawlins, A. L. Aminoalkylphenols as antima-
larials. II. (Heterocyclic-amino)-R-amino-o-cresols. The synthesis
of camoquine. J . Am. Chem. Soc. 1948, 70, 1363-1373.
(32) Mallams, A. K.; Israelstam, S. S. The reaction between â-
ketoesters and arylamines in the presence of polyphosphoric
acid. II. Ethyl acetoacetate and its a-alkyl derivatives and
arylamines. J . Org. Chem. 1964, 29, 3548-3554.
J M9607865