Synthesis and cytotoxic activity of N-[(alkylamino)alkyl]carboxamide derivatives of 7-oxo-7H-benz[de]anthracene, 7-oxo-7H-naphtho[1,2,3-de]quinoline, and 7-oxo-7H-benzo[e]perimidine

Article abstract of DOI:10.1016/j.bmc.2005.03.033

7-Oxo-7H-naphtho[1,2,3-de]quinoline-11-carboxamides and analogues were prepared and evaluated for in vitro and in vivo antitumor activity. Chromophore variations included 'deaza' (7-oxo-7H-benz[de]anthracene) and 'diaza' (7-oxo-7H-benzo[e]perimidine) analogues, and side chain variations included chiral α-methyl compounds. The naphthoquinolines were the most cytotoxic, with IC₅₀ values of 5-20 nM, and showed the strongest DNA binding, with high selectivity for G-C rich DNA. The chiral α-methyl analogues were 10-20-fold more cytotoxic than the parent des-methyl compound. Both enantiomers provided substantial growth delays against s.c. colon 38 tumors in mice, with the R-enantiomer more active than the S (tumor growth delays of >35 and 12 days, respectively).

Full text of DOI:10.1016/j.bmc.2005.03.033

Bioorganic & Medicinal Chemistry 13 (2005) 3657–3665
Synthesis and cytotoxic activity of N-[(alkylamino)alkyl]-
carboxamide derivatives of 7-oxo-7H-benz[de]anthracene,
7-oxo-7H-naphtho[1,2,3-de]quinoline, and 7-oxo-7H-
benzo[e]perimidine
Xianyong Bu,a Junjie Chen,a Leslie W. Deady,a,* Clare L. Smith,a Bruce C. Baguley,b
Debra Greenhalgh,b Shangjin Yangb and William A. Dennyb,*
aChemistry Department, La Trobe University, Victoria 3086, Australia
bAuckland Cancer Society Research Centre, School of Medical Sciences, The University of Auckland,
Private Bag 92019, Auckland 1000, New Zealand
Received 3 February 2005; accepted 16 March 2005
Available online 12 April 2005
Abstract—7-Oxo-7H-naphtho[1,2,3-de]quinoline-11-carboxamides and analogues were prepared and evaluated for in vitro and in
vivo antitumor activity. Chromophore variations included ꢀdeazaꢁ (7-oxo-7H-benz[de]anthracene) and ꢀdiazaꢁ (7-oxo-7H-
benzo[e]perimidine) analogues, and side chain variations included chiral a-methyl compounds. The naphthoquinolines were the
most cytotoxic, with IC50 values of 5–20 nM, and showed the strongest DNAbinding, with high selectivity for G-C rich DNA.
The chiral a-methyl analogues were 10–20-fold more cytotoxic than the parent des-methyl compound. Both enantiomers provided
substantial growth delays against s.c. colon 38 tumors in mice, with the R-enantiomer more active than the S (tumor growth delays
of >35 and 12 days, respectively).
ꢀ 2005 Elsevier Ltd. All rights reserved.
1. Introduction
tory properties of these compounds against a panel of
cell lines and, for selected examples, in vivo growth
delay data.
We recently reported on the cytotoxic activity of deriv-
atives of 7-oxo-7H-naphtho[1,2,3-de]quinoline (previ-
ously called dibenz[f,ij]isoquinoline), where the
11-carboxamide 1a in particular showed remarkable
curative activity against colon 38 tumors in mice.1 In
this series, the position of the carboxamide appeared
to be critical for activity, with the 2-, 4-, and 8-carbox-
amides not being as effective.1 We have now extended
these studies to include ꢀdeazaꢁ (7-oxo-7H-benz[de]-
anthracene; 2) and ꢀdiazaꢁ (7-oxo-7H-benzo[e]perimi-
dine; 3) analogues of 1a, as well as varying the nature
of the carboxamide linker chain in system 1. We report
here on the synthesis, DNAbinding and growth-inhibi-
2. Results and discussion
2.1. Chemistry
The ꢀdeazaꢁ compound, 7-oxo-7H-benz[de]anthracene-
11-carboxylic acid (4) was prepared from methyl 2-
iodobenzoate and methyl 8-bromo-1-naphthoate as
reported2,3 (Scheme 1). The starting point for all other
compounds was 1-amino-8-chloroanthraquinone 5a pre-
pared by literature methods, which we outlined previ-
ously.1 Now, in an improved procedure over our
previous route to 5c (Scheme 2), the chloro group was
substituted by reaction with CuCN. This gave access
to nitrile 5b, which then afforded acid 5c on acid hydro-
lysis, and subsequent reaction with acetone in aqueous
sodium hydroxide1 gave acid 6, the precursor to the
ꢀmonoazaꢁ amides 1a–f of Table 1.
Keywords: Naphthoquinolines; GC-selectivity; Cytotoxicity; Colon 38
tumors.
*
Corresponding authors. Tel.: +61 3 9479 2561; fax: +61 3 9479 1399
(L.W.D.); tel.: +64 3737 599x6144; fax: +64 3737 502
9
9
(W.A.D.); e-mail addresses: l.deady@latrobe.edu.au; b.denny@
uckland.ac.nz
0968-0896/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2005.03.033

3658
X. Bu et al. / Bioorg. Med. Chem. 13 (2005) 3657–3665
CO2Me
O
O
CO2Me
Br
i
i
ii
5c
I
+
CO2Me
CO2Me
N
O
CO2Me
NH2
O
CO2Me
5d
O
H
7
NMe2
ii
iii
O
compound 2
of Table 1
iii
v
Compound
3a of Table 1
4
CO2H
Scheme 1. Reagents and conditions: (i) Cu/180 ꢁC/4 h; (ii) concd
H2SO4/100 ꢁC; (iii) SOCl2 then NH2(CH2)2NMe2.
N
N
R
8a: X=CO2Me
8b: X=CO2H
iv
O
O
Scheme 3. Reagents and conditions: (i) MeOH/H2SO4/reflux 16 h; (ii)
Me2NCH(OMe)2/MeCN/reflux 3 h; (iii) NH4OAc/EtOH/reflux 1 h;
(iv) TFA–H2O (1:1.6)/reflux 24 h; (v) SOCl2 then amine.
iii
NH2
O
R
N
CO2H
lysis of the 11-ester functionality in 8a was best carried
out in mildly acid conditions; aqueous trifluoroacetic
acid provided a suitable solubilizing and acidic medium.
5a: R=Cl
6
Me
iv
i
5b: R=CN
ii
5c: R=CO2H
compounds
1a-1f of Table 1
The dimethylformamide reaction gives a single tetra-
cyclic product, but when the analogous dimethylaceta-
mide dimethyl acetal is used, the resulting intermediate
9 can ring close in two different ways (Scheme 4).
Studies with the model compound 5e established that
the course of the reaction was affected by the solvent. In
refluxing ethanol, reaction with the acetal went directly
to the substituted naphtho[1,2,3-de]quinolinone 10a, as
reported previously,5 and this was isolated in 64% yield.
In acetonitrile solvent the reaction was more complex,
and 10a was only a minor component of the initial prod-
uct. The major product, amidine 9a, could be separated
from 10a and, on subsequent reaction with ammonium
Scheme 2. Reagents and conditions: (i) CuCN/DMF/reflux; (ii) 80%
H2SO4/reflux 3 h; (iii) acetone/4% NaOH/reflux 1 h; (iv) CDI/dioxan/
reflux, then amine in dichloromethane or amineÆHCl/BOPÆCl/NEt3/
20 ꢁC 1 h.
For construction of the diaza ring system, the very insol-
uble 5c was converted to the methyl ester 5d. This was
reacted with dimethylformamide dimethyl acetal in
refluxing acetonitrile and gave the intermediate amidine
7 (not isolated), which was reacted with ammonium
acetate in boiling ethanol, as reported for the parent
compound,4 to give 8a in 90% yield (Scheme 3). Hydro-
Table 1. Structural, DNAbinding, and growth-inhibitory data for naphtho[1,2,3- de]quinoline 1a and analogues
O
O
O
10
N
N
NMe2
1
3 N
R
NMe2
O
N
H
O
N
O
N
H
H
Y
Me
3a,b
2
1a-1g
a
No.
R
Y
DNAbinding
dGC
Growth delaysb
Jurkat
c
c
dAT
Ratio
P388
100
LLTC
LA/LC
LD/LC
1a
1b
1c
1d
1e
1f
(CH2)2NMe2
Me
Me
0.37
0.30
0.28
0.71
0.50
0.68
0.87
1.3
4.1
3.9
8.2
2.9
6.1
3.8
16
11
13
29
101
5.9
12
424
42
2.5
12
2.5
18
CH(S-Me)CH2NMe2
CH(R-Me)CH2NMe2
(CH2)3NMe2
4.1
9.0
28
Me
Me
53
4.1
2
5.2
3
4.0
12
5.5
19
1.9
89
858
693
959
(CH2)2NH(CH2)2OH
(CH2)2Npiperidinyl
(CH2)2NMe2
Me
Me
550
180
1500
1530
64
650
296
3.2
1.9
5.1
2.2
1g
2
NMe2
(CH2)2NMe2
2.4
4.9
11
1930
2510
0.9
0.9
3a
3b
H
Me
1.6
1.3
3.4
8.5
48
a Association constants (·106 Mꢀ1) for binding to poly-(dA-dT)Æ(dA-dT) [dAT] and poly-(dG-dC)Æ(dG-dC) [dGC] double-stranded DNA, at 0.01
ionic strength, pH 7.0, and the resulting ratio dGC/dAT.
b IC50 (nM) for growth inhibition of cultured murine lymphocytic leukemia (P388), a cell line derived from the Lewis lung murine carcinoma (LLTC),
and a human acute leukemia (Jurkat) cell line.
c LA/LC and LD/LC; ratios of IC50s for amsacrine-resistant (LA) and doxorubicin-resistant (LD) Jurkat cell lines, compared to the parental line (LC).

X. Bu et al. / Bioorg. Med. Chem. 13 (2005) 3657–3665
3659
oped for 1a based on reverse-phase C18 flash column
chromatography in dilute aqueous TFA/MeCN; this
gave a sample of improved yield and purity. The remain-
ing amides were obtained from 4, 8b, 8e, and 10c by way
of intermediate acid chlorides (not isolated).
O
O
O
i (
ii (
10)
8)
NH2
X
N
O
X
5e: X=H
5d: X=CO2Me
Me NMe2
2.2. Structure–activity relationships
9a: X=H
9b: X=CO2Me
iii
The DNAbinding properties of the series were deter-
mined by competition with ethidium for binding sites
on the synthetic double-stranded copolymers poly-d(A-
T) and poly-d(G-C) as previously described.6 All com-
pounds bound selectively to poly-d(G-C), as has been
found for a number of anticancer drugs containing chro-
mophores attached to a carboxamide side chain.7 The
degree of GC selectivity was affected by the side chain
and the twofold difference in the poly-d(G-C) binding
of the chiral compounds 1b and 1c was of interest.
The in vitro growth-inhibitory activity of the com-
pounds was determined in a panel of tumor cell lines
in culture; a murine leukemia (P388), a murine lung car-
cinoma (Lewis lung)8 and three human leukemia lines.
Of these, JLC is wild-type, while JLA and JLD are resis-
tant to topo II agents (85-fold to amsacrine and 13-fold
to doxorubicin, respectively) because of a reduced level
of topo II.9,10 It is considered that IC50 ratios (JLA/
JLC and JLD/JLC) of less than about twofold suggest a
non-topoisomerase II mediated mechanism of action.11
i
O
O
N
N
X
N
X
Me
8a: X=H
8d: X=CO2Me
8e: X=CO2H
NMe2
10a: X=H
10b: X=CO2Me
10c: X=CO2H
iv
v
vi
vi
compound 3b
of Table 1
compound 1g
of Table 1
Scheme 4. Reagents and conditions: (i) Me2NC(Me)(OMe)2/EtOH/
reflux 3 h; (ii) Me2NC(Me)(OMe)2/MeCN/reflux 3 h; (iii) NH4OAc/
EtOH/reflux 1 h; (iv) TFA–H2O (3:4)/reflux 7 days; (v) 10% NaOH–
MeOH–H2O (1:2:2)/reflux 24 h; (vi) SOCl2 then amine.
acetate in ethanol, followed the path seen for dimethyl-
formamide above, giving the alternative perimidine tet-
racycle 8a in 57% overall yield.
Compounds 1a, 2, and 3b represent three different chro-
mophores, with one, none or two aza atoms, respec-
tively. Of these, 2 was the least cytotoxic, while the
compounds with aza atoms were broadly equivalent
and more potent. The ratio of their binding to AT/GC
DNAroughly paralleled cytotoxicity, with 2 also being
the least selective. The CONH(CH2)2NRR side chain
was previously shown1 to be the best of several explored
in this series, and compounds 1a–g study additional as-
pects of this. The enantiomers 1b and 1c show DNA
binding broadly similar to that of the parent 1a, and
both were much more cytotoxic than 1a, with the
S-enantiomer being slightly better than the R-enantiomer.
Lengthening the carboxamide chain (1d) retained rea-
sonable activity, but weakening the pendant base (1e)
was not favorable. In the ꢀdiazaꢁ series (3a and 3b), the
presence of the 2-methyl group had only a minor effect
on activity. In the aza series, the corresponding ꢀdes-
methylꢁ derivative of 1a was not accessible, but the 2-
NMe2 analogue 1g was much less active. Of the com-
pound studied here, the S-methyl enantiomer 1b stood
out in the Jurkat assays, with its high ratios (12 and
18) suggesting it is primarily a topo II inhibitor.
The same solvent effect was noted with the ester 5d, but
the yields were lower. Thus, reaction with dimethylacet-
amide dimethyl acetal in methanol went directly to the
tetracycle 10b, which separated from the reaction mix-
ture in 28% yield as a mixture of two components. These
were separated by chromatography and, unexpectedly,
both were consistent with structure 10b. The detail of
these presumed atropisomers was not established, and
the mixture was used directly in further reaction. No
solid separated from the analogous reaction mix in aceto-
nitrile and, after evaporation of the volatiles, reaction of
the residue with ammonium acetate in ethanol afforded
the readily isolable 8d in 36% yield. In spite of the mod-
est yields, these reactions did allow sufficient 8d and 10b
to be obtained for the rest of the sequence to be carried
out.
Hydrolysis of the 11-ester functionality in 8d was again
carried out in aqueous trifluoroacetic acid, whereas base
conditions were preferable for the naphthoquinoline
10b. Amide formation was achieved in various ways.
Compounds 1b–f were prepared from 6 and the appro-
priate amine with CDI or BOPÆCl as coupling agents.
The synthesis of these amides used the crude acid 6,
which by NMR was ꢁ50% pure, with a 2:1 ratio of
the desired compound to a major impurity (the 4-methyl
analogue) and many minor impurities. The crude amides
were then purified by preparative reverse-phase HPLC.
This procedure gave pure products, but in low yields
(8–30%) based on the acid 6 in the crude mixture. Sub-
sequently, a modified purification procedure was devel-
2.3. In vivo studies
Compound 1a has been previously evaluated in vivo
against subcutaneously implanted colon 38 tumors in
mice, where it provided cures when administered by a
repeated dose schedule and a growth delay of 17 days
when administered as a single dose of 65 mg/kg.1
Selected derivatives were therefore compared using a single
dose administration schedule. Compounds 1b and 1c
were found to be considerably more dose-potent than

3660
X. Bu et al. / Bioorg. Med. Chem. 13 (2005) 3657–3665
100
(A)
(B)
(C)
100
10
1
100
10
1
10
1
0.1
0.1
0.1
0
5
10 15 20 25 30 35 40
Time (days)
0
5
10 15 20 25 30 35 40
Time (days)
0
5
10
15
20
Time (days)
Figure 1. Growth curves of subcutaneously implanted Colon 38 tumors. (A) Untreated mice (d), 1b at 8.9 mg/kg (s) and 5.9 mg/kg (h). (B)
Untreated mice (d), 1c at 8.9 mg/kg (s) and 5.9 mg/kg (h). (C) Untreated mice (d), 3a at doses of 45 mg/kg (,) and 65 mg/kg (n).
1a, consistent with their in vitro properties. At a dose of
5.9 mg/kg, both 1b and 1c provided growth delays (12
and >35 days, respectively), with some complete tumor
regressions, with the R-enantiomer 1c appearing the
more active. Previous studies with benzophenazinecarb-
oxamides12 also showed a preference for the R-enantio-
mer 11 (XR-11576), which is currently in clinical trial.13
Compound 3a provided growth delays of 6 days when
tested at doses of 45 and 65 mg/kg (Fig. 1). Both chro-
mophore series therefore display in vivo antitumor
activity.
as d values (ppm) relative to Me4Si. Various standard
techniques were used to identify proton-bound carbons
in 13C NMR spectra. EI and LSI (3-nitrobenzyl alcohol
as liquid matrix) mode high-resolution mass spectra were
obtained by Dr. N. Davies, University of Tasmania, Aus-
tralia, or on a Varian VG-70SE spectrometer at nominal
5000 resolution at Auckland, New Zealand. Microanaly-
ses were carried out at the Campbell Microanalytical
Laboratory, University of Otago, New Zealand.
4.1. 7-Oxo-7H-benzo[de]anthracene-11-carboxylic acid
(4)
O
O
This was prepared from methyl 2-iodobenzoate and
methyl 8-bromo-1-naphthoate as reported2,3 (Scheme
1): mp 272–273 ꢁC. H NMR [(CD3)2SO] d 7.71 (t, 1H,
J = 7.6 Hz), 7.79 (t, 1H, J = 7.9 Hz), 7.89–7.96 (m,
2H), 8.27 (d, 1H, J = 8.1 Hz), 8.40 (d, 1H, J = 7.5 Hz),
8.46–8.53 (m, 2H), 8.65 (d, 1H, J = 7.4 Hz).
1
10
1
3 N
NMe2
X
X
NMe2
O
N
O
N
H
H
Me
2: X=CH
3: X=N
1a
N
N
4.2. 8-Aminoanthraquinone-1-carbonitrile (5b)
MeO
Me
CuCN (1.8 g, 20 mmol) was added in one portion to a
warm solution of 1-amino-8-chloroanthraquinone1 (5a)
(2.6 g, 10 mmol) in dimethyl formamide (25 mL), and
the mixture was stirred under reflux for 4 h, then cooled.
Water (200 mL) was added and the solid was filtered off,
refluxed with HCl (70 mL from concd HCl and water
2:3) for 1 h, and filtered while warm. The solid was
washed with water and dried, to give 5b as a red-brown
solid (2.0 g, 80%). This was used directly for the hydro-
lysis reaction, but could be recrystallized from acetic
NMe2
O
N
11
H
3. Conclusions
N-[(Alkylamino)alkyl]carboxamide derivatives of 7-oxo-
7H-naphtho[1,2,3-de]quinoline (1) are shown to have
greater antitumor activity than analogues with an extra
(3), or zero (2) aza functions in the chromophore. The
potency is enhanced by the presence of a chiral a-methyl
group in the side chain, and compounds 1 remain an
interesting series for further development.
1
acid: mp 263–265 ꢁC. H NMR (CDCl3) d 7.10 (d, 1H,
J = 8.1 Hz), 7.60 (t, 1H, J = 8.0 Hz), 7.78 (t, 1H,
J = 9.6 Hz), 7.88 (d, 1H, J = 8.3 Hz), 8.05 (d, 1H,
J = 8.4 Hz), 8.46 (d, 1H, J = 6.8 Hz).
4. Experimental
4.3. 8-Aminoanthraquinone-1-carboxylic acid (5c)
Melting points are uncorrected. NMR spectra were re-
corded on a Bruker AM-300 spectrometer operating at
300.13 MHz, (1H) and 75.47 MHz, (13C) and a Bruker
DRX-400 spectrometer operating at 400.13 MHz, (1H)
and 100.62 MHz, (13C), with chemical shifts are reported
Nitrile 5b (1.8 g) in a mixture of concd H2SO4 (16 mL)
and water (8 mL) was heated under reflux for 3 h, then
cooled and poured onto ice/water. The solid that sepa-
rated was filtered off, washed with water, then stirred
with 4% aqueous NaOH and filtered. The filtrate was

X. Bu et al. / Bioorg. Med. Chem. 13 (2005) 3657–3665
3661
taken to pH 3–4 with concd HCl, and the solid that sepa-
rated was filtered off, washed with water and dried to
give the acid 5c (1.1 g, 57%), with the H NMR spec-
as reported5 gave 10a, mp (EtOH) 178–180 ꢁC (lit.5 mp
1
189 ꢁC) in 64% yield. H NMR (CDCl3) d 3.34 (s, 6H,
1
N(CH3)2), 7.60–7.77 (m, 4H), 8.1 (br s, 1H), 8.25 (d,
1H, J = 8.0 Hz), 8.29 (d, 1H, J = 7.4 Hz), 8.47 (d, 1H,
J = 7.7 Hz). 13C NMR (CDCl3) d 37.8 (N(CH3)2),
104.2 (CH), 118.0 (C), 122.4 (CH), 123.0 (CH), 127.7
(CH), 127.8 (C) 129.1 (CH), 129.3 (CH), 131.6 (C),
132.3 (CH), 132.6 (CH), 133.7 (C), 134.3 (C), 147.4
(C), 157.2 (C), 182.7 (C).
trum of an authentic sample.1
4.4. Methyl 8-aminoanthraquinone-1-carboxylate (5d)
Amixture of acid 5c (1.71 g, 6.4 mmol), absolute MeOH
(50 mL), and concd H2SO4 (0.5 mL) was refluxed for
16 h, then cooled and a little unreacted acid was filtered
off. The filtrate was concentrated to c 20 mL, water was
added and the solid which separated was filtered off and
dried to give the red-brown ester 5d (1.40 g, 78%): mp
4.9. 2-Methyl-7H-benzo[e]perimidin-7-one (8a)
Asolution of 5e (0.40 g, 1.80 mmol) and dimethylacet-
amide dimethyl acetal (1.20 g, 8.96 mmol) in dry MeCN
(20 mL) was heated under reflux for 5 h under an atmo-
sphere of nitrogen, then cooled to ꢀ15 ꢁC and 10a
(0.05 g) was removed by filtration. The filtrate was evapo-
rated to dryness at reduced pressure to leave the inter-
mediate 9a as a brown solid (0.40 g, 77%), which was
1
149–151 ꢁC. H NMR (CDCl3) d 4.02 (s, OCH3), 6.95
(dd, 1H, J = 8.4, 1.0 Hz), 7.45 (t, 1H, J = 7.9 Hz),
7.61–7.66 (m, 2H), 7.74 (t, 1H, J = 7.7 Hz), 8.34 (d,
1H, J = 7.7, 1.4 Hz).
4.5. 2-Methyl-7-oxo-7H-naphtho[1,2,3-de]quinoline-11-
carboxylic acid (6)
1
used in the next step without further purification. H
NMR (CDCl3) d 1.81 (s, 3H, CH3), 3.31 (s, 6H,
N(CH3)2), 7.07 (d, 1H, J = 8.0 Hz), 7.58 (t, 1H,
J = 7.8 Hz), 7.68–7.73 (m, 2H), 7.94 (d, 1H, J =
6.5 Hz), 8.20–8.24 (m, 2H).
This was prepared from 5d as reported, previously,1 and
used directly as the crude product from the cyclization
reaction (ca. 50% pure).
4.6. Methyl 7-oxo-7H-benzo[e]perimidine-11-carboxylate
(8a)
Amixture of 9a (0.40 g) and ammonium acetate (0.52 g,
6.75 mmol) in dry EtOH (8 mL) was heated under reflux
for 2 h then cooled to room temperature. The precipitate
that formed was collected by filtration to yield 8a as a
brown solid (0.25 g, 74%), mp 201–203 ꢁC (lit.1 mp
211–212 ꢁC; lit.14 mp 201–203 ꢁC), with NMR data as
reported.1
Amixture of methyl 8-aminoanthraquinone-1-carboxyl-
ate (5d) (0.14 g, 0.5 mmol) and dimethylformamide di-
methyl acetal (0.3 g, 2.5 mmol) in MeCN (5 mL) was
heated under reflux for 3 h. The solvent was removed
at reduced pressure and to the dark residue was added
ammonium acetate (0.24 g, 3.1 mmol) and dry EtOH
(4 mL). The mixture was heated under reflux for 1 h,
cooled to room temperature and the precipitate that
formed was collected by filtration, to give 8a as a light
tan solid (0.13 g, 90%): mp (MeCN) 229–230 ꢁC. 1H
NMR (CDCl3) d 4.04 (s, OCH3), 7.77–7.81 (m, 2H),
8.12 (t, 1H, J = 8.8 Hz), 8.39 (d, 1H, J = 8.6 Hz), 8.54
(dd, J = 7.1, 1.5 Hz), 8.63 (d, 1H, J = 7.1 Hz), 9.47 (s).
13C NMR (CDCl3) d 52.5 (CH3), 119.5 (C), 127.8 (C),
129.0 (CH), 129.4 (CH), 130.7 (C), 130.8 (C), 131.8
(CH), 132.5 (CH), 133.3 (C), 133.8 (CH), 135.0 (CH),
149 0 (C), 154.9 (CH), 170.0 (C), 180.7 (C). Anal. Calcd
for C17H10N2O3: C, 70.3; H, 3.5; N, 9.7. Found: C, 70.4;
H, 3.4; N, 9.9.
4.10. Methyl 2-dimethylamino-7-oxo-7H-naphtho[1,2,3-
de]quinoline-11-carboxylate (10b)
Asolution of methyl 8-aminoanthraquinone-1-carbox-
ylate (5d) (1.0 g) and dimethylacetamide dimethyl acetal
(2.0 g), in MeOH (15 mL) was heated under reflux for
3 h, then stored in the freezer overnight. The dark solid
which separated was filtered off and washed with a little
cold MeOH to give 0.33 g (28%) of predominantly a
mixture of two compounds (c 2:1). PTLC of a sample
(0.1 g) (silica:chloroform) gave the two components, Rf
0.2 (major) and Rf 0.1 (minor), both of which were con-
sistent with structure 10b:
Major: Orange needles: mp (MeOH) 178–180 ꢁC 1H
NMR (CDCl3) d 3.27 (s, 6H, N(CH3)2), 4.03 (s, 3H,
OCH3), 7.44 (s, 1H), 7.66 (t, 1H, J = 7.7 Hz), 7.70–
7.77 (m, 2H), 8.03 (dd, 1H, J = 8.3, 1.1 Hz), 8.28 (dd,
1H, J = 7.3, 1.1 Hz), 8.61 (dd, 1H, J = 7.8, 1.6 Hz).
13C NMR (CDCl3) d 37.8 (N(CH3)2), 52.6 (OCH3),
108.9 (CH), 118.0 (C), 123.4 (CH), 127.2 (C), 128.8
(CH), 129.3 (CH), 129.9 (CH), 131.3 (C), 132.1 (C),
132.94 (CH), 132.99 (CH), 133.2 (C), 133.3 (C), 147.6
(C), 157.3 (C), 171.1 (C), 182.2 (C). EIMS: m/z 332
(65%), 317 (75), 303 (65), 285 (30), 273 (25), 258 (20),
245 (100), 231 (55). HRMS (EI): Calcd for
C20H16N2O3: 332.1162 (M+). Found: 332.1150.
4.7. 7-Oxo-7H-benzo[e]perimidine-11-carboxylic acid
(8b)
Asolution of ester 8a (0.32 g) in water (4 mL) and triflu-
oroacetic acid (2.5 mL) was heated under reflux for 24 h,
then cooled and added, with stirring, to water (25 mL).
The solid which separated was filtered and washed with
1
water to yield 8b as a red solid (0.27 g, 89%). H NMR
[(CD3)2SO] d 7.83–7.94 (m, 2H), 8.29 (t, 1H, J = 8.0 Hz),
8.40–8.45 (m, 2H), 8.57 (d, 1H, J = 7.2 Hz), 9.47 (s, 1H).
4.8. 2-Dimethylamino-7H-naphtho[1,2,3-de]quinolin-7-
one (10a)
Reaction of 1-aminoanthraquinone (5e) with dimethyl-
acetamide dimethyl acetal under reflux for 3 h in EtOH
Minor: Orange solid: mp (MeCN) 261–264 ꢁC. 1H
NMR (CDCl3) d 3.34 (s, 6H, N(CH3)2), 4.05 (s, 3H,

3662
X. Bu et al. / Bioorg. Med. Chem. 13 (2005) 3657–3665
OCH3), 7.55 (d, 1H, J = 7.0 Hz), 7.69 (s, 1H), 7.71–7.79
(m, 2H), 8.01 (dd, 1H, J = 8.4, 1.1 Hz), 8.24 (d, 1H,
J = 7.4 Hz), 8.38 (d, 1H, J = 8.1 Hz). 13C NMR (CDCl3)
d 39.0 (N(CH3)2), 52.6 (OCH3), 105.8 (CH), 117.0 (C),
123.9 (CH), 124.5 (CH), 127.0 (C), 128.5 (C), 128.6
(CH), 130.3 (CH), 132.5 (CH), 133.2 (C), 135.3 (CH),
170.5 (C), 181.0 (C). EIMS: m/z 332 (65%), 317 (65),
303 (100), 289 (20), 272 (12), 258 (30), 245 (10), 230
(15). HRMS (EI): Calcd for C20H16N2O3: 332.1162
(M+). Found: 332.1161.
7 days, then cooled and added, with stirring, to water
(25 mL). The black solid which separated was filtered
and washed with water to give a first crop of 8e
(0.28 g, 46%). 1H NMR [(CD3)2SO] d 2.82 (s, 3H,
CH3), 7.8–7.9 (m, 2H), 8.15 (t, 1H, J = 7.3 Hz), 8.28
(d, 1H, J = 8.0 Hz), 8.35 (d, 1H, J = 7.5 Hz), 8.41 (d,
1H, J = 7.1 Hz). The filtrate was evaporated to dryness,
the dark residue was stirred with 4% sodium hydroxide
(5 mL), the mixture was filtered and the filtrate was acidi-
fied with hydrochloric acid. The light brown solid which
separated was filtered and washed with water to give a
1
Ammonium acetate (0.4 g, 5.2 mmol) was added to the
dark filtrate, and the mixture was heated under reflux
for 2.5 h, then cooled in the freezer. The dark solid
which separated was filtered off to give 8d (0.06 g). Dilu-
tion of the filtrate with much water provided a solid
(0.05 g) shown by NMR to be largely starting material
5d.
second crop (0.16 g, 26%), with an identical H NMR
spectrum.
4.14. N-[2-(Dimethylamino)ethyl]-7-oxo-7H-benz[de]an-
thracene-11-carboxamide (2)
Asuspension of 4 (0.14 g, 0.51 mmol) in dry benzene
(20 mL) was treated with SOCl2 (0.07 g, 0.59 mmol),
and the mixture was heated under reflux for 2 h, then
cooled to room temperature. N,N-Dimethylethylenedi-
amine (0.11 g, 1.25 mmol) was added over 10 min and
the resulting mixture was stirred for 1 h, then heated
under reflux for 1 h. Volatiles were evaporated at reduced
pressure and the residue was dissolved in chloroform
(30 mL), washed with water (2 · 10 mL), and dried
(Na2SO4). The solvent was removed to give 2 as a pale
yellow solid (0.16 g, 91%): mp (benzene/light petroleum)
4.11. 2-Dimethylamino-7-oxo-7H-naphtho[1,2,3-de]quino-
line-11-carboxylic acid (10c)
Ester 10b (0.3 g) in MeOH (6 mL), water (6 mL), and
10% NaOH (3 mL) was heated under reflux for 24 h
and filtered. The filtrate was concentrated to a small vol-
ume, acidified with hydrochloric acid and a first crop of
10c was filtered off as a black solid (0.1 g). The filtrate
was evaporated and the residue was extracted with hot
EtOH. The extract was filtered while hot, and the yel-
low-brown filtrate was evaporated to leave a second
crop (0.1 g). 1H NMR [(CD3)2SO] d 3.26 (s, 6H,
N(CH3)2), 7.7–7.9 (m, 4H), 8.0–8.2 (m, 2H), 8.40 (d,
1H, J = 7.2 Hz).
1
163–165 ꢁC. H NMR (CDCl3) d 2.11 (s, 6H, NMe2),
2.48 (t, 2H, J = 6.0 Hz, CH2), 3.59 (q, 2H, J = 5.8 Hz,
CH2), 6.57 (br t, 1H, CONH), 7.51 (t, 1H, J = 7.6 Hz),
7.68 (t, 1H, J = 7.9 Hz), 7.69–7.75 (m, 2H), 7.96 (d,
1H, J = 8.0 Hz), 8.18 (d, 1H, J = 8.0 Hz), 8.47 (d, 1H,
J = 7.6 Hz), 8.52 (d, 1H, J = 7.6 Hz), 8.68 (d, 1H,
J = 7.6 Hz).13C NMR (CDCl3) d 37.6 (CH2), 45.0
(CH3), 57.3 (CH2), 125.4 (C), 126.27 (CH), 126.34
(CH), 17.7 (CH), 128.2 (C), 128.6 (CH), 129.3 (CH),
129.8 (CH), 130.6 (CH), 132.3 (C), 132.8 (C), 133.3
(C), 133.7 (CH), 135.5 (CH), 135.8 (C), 171.8 (C),
183.3 (C). Anal. Calcd for C22H20N2O2Æ0.75H2O: C,
73.8; H, 6.05; N, 7.8. Found: C, 74.0; H, 5.6; N, 7.9.
4.12. Methyl 2-methyl-7-oxo-7H-benzo[e]perimidine-11-
carboxylate (8d)
Dimethylacetamide dimethyl acetal (0.7 g, 5 mmol) was
added to a solution of methyl 8-aminoanthraquinone-
1-carboxylate (5d) (0.28 g, 1 mmol) in dry MeCN (5 mL),
and the solution was heated under reflux for 3 h. No
solid separated when this solution was then stored in the
freezer. The solvent was removed at reduced pressure,
ammonium acetate (0.5 g), and EtOH (5 mL) were
added to the dark residue and the mixture was heated
under reflux for 1.5 h. After cooling, the brown solid
was filtered off to give 8d (0.11 g, 36%): mp 211–
4.15. Preparation of amides 1a–f: general
4.15.1. Method A. Asolution of the crude acid 6 (0.5 g,
1.73 mmol) in 50 mL of 1,4-dioxane was heated under
reflux with CDI (0.45 g, 2.8 mmol) for 4 h, cooled to
room temperature, the solvent was removed at reduced
pressure and the residue was dissolved in CH2Cl2
(50 mL), washed with 10% Na2CO3 aqueous solution,
water, and dried. The solvent was evaporated under re-
duced pressure to give the crude imidazolide (0.5 g, 85%)
as a pale brown solid, which was used for next step with-
out further purification. The imidazolide (0.5 g,
1.47 mmol) and the amine (0.5 g, 5 equiv) was dissolved
in dichloromethane (10 mL) and stirred at room temper-
ature for 24 h, and the reaction mixture was then filtered
through a short alumina column with EtOAc/CH2Cl2 to
give the crude amide. This was purified by preparative
HPLC to remove closely running impurities.
1
213 ꢁC. H NMR (CDCl3) d 2.95 (s, 3H, CH3), 4.05 (s,
3H, OCH3), 7.72–7.81 (m, 2H), 8.05 (t, 1H,
J = 8.1 Hz), 8.26 (d, 1H, J = 8.3 Hz), 8.53–8.55 (m,
2H). 13C NMR (CDCl3) d 26.2 (CH3), 52.3 (OCH3),
117.3 (C), 127.7 (C), 128.3 (CH), 128.9 (CH), 130.7
(C), 131.6 (CH), 132.4 (CH), 133.1 (C), 133.7 (CH),
133.9 (C), 134.3 (CH), 149.5 (C), 154.7 (C), 164.3 (C),
170.1 (C), 181.0 (C). EIMS (m/z): 304 (40%), 273
(100), 246 (80). HRMS (EI): Calcd for C18H12N2O3:
304.0848 (M+). Found: 304.0841.
4.13. 2-Methyl-7-oxo-7H-benzo[e]perimidine-11-carbox-
ylic acid (8e)
Asolution of ester 8d (0.64 g) in water (12 mL) and tri-
fluoroacetic acid (9 mL) was heated under reflux for
4.15.2. Method B. Asolution of crude 6 (0.5 g,
1.73 mmol) and the amine hydrochloride in Et3N

X. Bu et al. / Bioorg. Med. Chem. 13 (2005) 3657–3665
3663
(2.0 g) was cooled in an ice-bath and treated with
BOPÆCl (0.54 g, 2.1 mmol) added in one portion. The
reaction was stirred at room temperature for 1 h, then
diluted with CH2Cl2, washed with saturated aqueous
NaHCO3, dried, and concentrated under reduced pres-
sure. The residue was filtered through a short alumina
column with EtOAc/CH2Cl2 and then purified by pre-
parative HPLC to remove closely running impurities.
4.20. N-{2-[(2-Hydroxyethyl)amino]ethyl}-2-methyl-7-
oxo-7H-naphtho[1,2,3-de] quinoline-11-carboxamide (1e)
Reaction of 6 (0.5 g, 1.73 mmol) with 2-(2-aminoethyl-
amino)ethanol by method Agave
1e as a solid
1
(100 mg, 0.27 g, 32% based on available 6); H NMR
[(CD3)2SO] d 8.85 (br s, 1H), 8.56 (dd, J = 7.3, 0.9 Hz,
1H), 8.46 (dd, J = 7.7, 1.5 Hz, 1H), 8.41 (dd, J = 8.2,
0.9 Hz, 1H), 8.14 (s, 1H), 7.84 (t, 1H), 7.80 (dd,
J = 7.4, 1.5 Hz, 1H), 7.78 (t, 1H), 4.45 (br s, 1H), 3.44
(m, 4H), 2.78 (m, 2H), 2.76 (s, 3H), 2.61 (m, 2H).
HRMS(EI) Calcd for C22H22N3O3: 376.1661 (M+).
Found: 376.1662.
4.16. Preparative HPLC
These were performed using a Synergi Polar-RP
(250 · 21.2 mm) column with detection at 254 nM. The
column was eluted by washing for 10 min with a linear
gradient of 35–100% of MeCN (contained 0.1% of
TFA) in water (acidified to pH 3 with TFA). Aliquots
were basified with ammonia, extracted with dichloro-
methane (3 · 50 mL), dried, and concentrated under re-
duced pressure to give the pure products.
4.21. 2-Methyl-7-oxo-N-[2-(1-piperidinyl)ethyl]-7H-
naphtho[1,2,3-de]quinoline-11-carboxamide (1f)
Reaction of 6 (0.5 g, 1.73 mmol) with 1-(2-amino-
ethyl)piperidine by method Agave 1f as a solid
(90 mg, 26% based on available 6); 1H NMR [(CD3)2SO]
d 8.79 (br s, 1H), 8.56 (dd, J = 7.3, 0.9 Hz, 1H), 8.46 (dd,
J = 7.7, 1.5 Hz, 1H), 8.41 (dd, J = 8.2, 0.9 Hz, 1H), 8.14
(s, 1H), 8.05 (t, 1H), 7.80 (dd, J = 7.4, 1.5 Hz, 1H), 7.75
(t, 1H), 3.46 (m, 4H), 2.76 (s, 3H), 2.40 (m, 4H), 1.50 (m,
4H), 1.38 (m, 2H). HRMS(EI) Calcd for C25H26N3O2:
400.2025 (M+). Found: 400.2018.
4.17. N-[(1R)-2-(Dimethylamino)-1-methylethyl]-2-
methyl-7-oxo-7H-naphtho[1,2,3-de]quinoline-11-carbox-
amide (1b)
Reaction of 6 (0.5 g, 1.73 mmol) with (1R)-2-(dimethyl-
amino)-1-methylethylamine hydrochloride (0.5 g, 2.4
mmol) by method B gave 1b as a solid (53 mg, 8%);
1H NMR [(CD3)2SO] d 8.41 (dd, J = 7.3, 0.9 Hz), 8.37
(dd, J = 7.7, 1.5 Hz), 8.26 (dd, J = 8.2, 0.9 Hz), 8.19
(s), 7.79 (t), 7.65 (dd, J = 7.4, 1.5 Hz), 7.51 (t), 6.96 (br
s, NH), 4.31 (m, 1H), 2.78 (s, 3H), 2.54 (m, 1H), 2.37
(m, 1H), 2.29 (s, 6H), 1.47 (d, J = 6.3 Hz, 3H).
HRMS(EI) Calcd for C23H24N3O2: 374.1869 (M+).
Found: 374.1871.
4.22. N-[2-(Dimethylamino)ethyl]-2-methyl-7-oxo-7H-
naphtho[1,2,3-de]quinoline-11-carboxamide (1a): modified
method
Asolution of crude 6 (1.45 g, 5.01 mmol, 50% by NMR)
in dry dioxan (50 mL) was treated with 1,10-carbonyldi-
imidazole (1.22 g, 7.5 mmol), and the mixture was re-
fluxed under N2 for 4 h. The solvent was removed
under reduced pressure, CH2Cl2 (150 mL) was added,
and organic layer was washed with 10% Na2CO3
(80 mL) and brine (80 mL), dried (Na2SO4). Evapora-
tion under reduced pressure gave the intermediate imi-
dazolide as a brown solid (1.43 g). This was dissolved
in dioxane (15 mL) and N,N-dimethylethylenediamine
(5 mL), and refluxed under N2 for 3 h. The solvent
was removed under reduced pressure, and the residue
was diluted with 10% Na2CO3 (150 mL) and extracted
with CH2Cl2 (3 · 80 mL). The combined organic layers
were washed with brine (2 · 100 mL) and dried
(Na2SO4). The resulting solution was washed through
an alumina column (50 · 50 mm), eluting with CH2Cl2/
MeOH (9:1). Reversed-phase C18 flash column chroma-
tography (150 · 20 mm) of this product, eluting with
0.05% aqueous TFA/MeCN (90:10, then 40:60) gave
1a as an orange amorphous solid (516 mg, 58% based
on available 6, 82% pure by HPLC). This product
(200 mg) was dissolved in dioxane and saturated HCl
in dioxan was added dropwise until the solution reached
pH ꢂ2. The solvent was removed under reduced pres-
sure, and the resulting solid was recrystallized from
EtOAc/MeOH to give the hydrochloride salt as an
amorphous orange solid (170 mg, 45% overall yield
4.18. N-[(1S)-2-(Dimethylamino)-1-methylethyl]-2-
methyl-7-oxo-7H-naphtho[1,2,3-de]quinoline-11-carbox-
amide (1c)
Reaction of 6 (0.5 g, 1.73 mmol) with (1S)-2-(dimethyl-
amino)-1-methylethylamine hydrochloride (0.5 g, 2.4
mmol) by method B gave 1c as a solid (25 mg, 8% based
on available 6); 1H NMR [(CD3)2SO] d 8.41 (dd, J = 7.3,
0.9 Hz), 8.37 (dd, J = 7.7, 1.5 Hz), 8.26 (dd, J = 8.2,
0.9 Hz), 8.19 (s), 7.79 (t), 7.65 (dd, J = 7.4, 1.5 Hz),
7.51 (t), 6.96 (br s, NH), 4.31 (m, 1H), 2.78 (s, 3H),
2.54 (m, 1H), 2.37 (m, 1H), 2.29 (s, 6H), 1.47 (d,
J = 6.3 Hz, 3H). HRMS (EI) Calcd for C23H24N3O2:
374.1869 (M+). Found: 374.1858.
4.19. N-[3-(Dimethylamino)propyl]-2-methyl-7-oxo-7H-
naphtho[1,2,3-de]quinoline-11-carboxamide (1d)
Reaction of 6 (0.5 g, 1.73 mmol) with 3-(dimethyl-
amino)propylamine (0.5 g, 4.9 mmol) by method Agave
1
1d as a solid (100 mg, 32% based on available 6); H
NMR [(CD3)2SO] d 8.83 (br s, 1H), 8.56 (dd, J = 7.3,
0.9 Hz, 1H), 8.46 (dd, J = 7.7, 1.5 Hz, 1H), 8.41 (dd,
J = 8.2, 0.9 Hz, 1H), 8.14 (s, 1H), 7.84 (t, 1H), 7.80
(dd, J = 7.4, 1.5 Hz, 1H), 7.78 (t, 1H), 2.76 (s, 3H),
2.50 (m, 2H), 2.27 (m, 2H), 2.12 (s, 6H), 1.72 (m, 2H).
HRMS(EI) Calcd for C23H23N3O2: 373.1790 (M+).
Found: 373.1782 [M+H].
1
based on available 6, 92.5% pure by HPLC); H NMR
[(CD3)2SO] d 10.36 (br s, 1H), 9.14 (br t, J = 5.5 Hz,
1H), 8.57 (dd, J = 7.3, 1.2 Hz, 1H), 8.50 (dd, J = 7.8,
1.6 Hz, 1H,), 8.45 (dd, J = 8.3, 1.2 Hz, 1H), 8.09 (dd,

3664
X. Bu et al. / Bioorg. Med. Chem. 13 (2005) 3657–3665
J = 8.2, 7.4 Hz, 1H), 8.05 (s, 1H), 7.93 (dd, J = 7.5,
1.6 Hz, 1H), 7.85 (t, J = 7.7 Hz, 1H), 3.75 (td, J = 6.4,
5.7 Hz, 2H), 3.34 (td, J = 6.4, 5.8 Hz, 2H), 2.87 (s,
3H), 2.86 (s, 3H), 2.81 (3H, s); HRMS(EI) Calcd for
C22H22N3O2: 360.1712 (M+). Found: 360.1707.
wet but remains crystalline when dry) to give pure 3b
as an orange-brown solid: mp 122–123 ꢁC. H NMR
1
(D2O) d 2.57 (s, 3H, CH3), 2.96 (s, 6H, NMe2), 3.45
(t, 2H, J = 6.1 Hz, CH2), 3.79 (t, 2H, J = 6.1 Hz,
CH2), 7.45–7.75 (m, 5H), 7.88 (br d, 1H). HRMS
(LSI): Calcd for C21H21N4O2: 361.1666 (M+H)+.
Found: 361.1613. Anal. Calcd for C21H20N4O2Æ
C2H2O4Æ1.5H2O: C, 57.9; H, 5.3; N, 11.7. Found: C,
57.6; H, 5.0; N, 11.7.
4.23. N-[2-(Dimethylamino)ethyl]-7-oxo-7H-benzo[e]per-
imidine-11-carboxamide (3a)
Amixture of acid 8b (0.27 g, 1 mmol) in SOCl2 (3 mL)
was kept at 50 ꢁC for 30 min. The SOCl2 was evaporated
under reduced pressure, and residual traces were removed
by azeotroping with benzene. Dichloromethane (10 mL)
was added, the mixture was stirred and cooled on ice, a
solution of N,N-dimethylethylenediamine (0.5 mL) in
dichloromethane was then added, and the solution was
stirred at room temperature for 30 min. It was washed
with water, 10% sodium carbonate, dried, and the solvent
was evaporated to leave a red residue of the amide (0.22 g,
63%). This was taken up in EtOH, a little insoluble mate-
rial was filtered and the filtrate was evaporated. The resi-
due was dissolved in acetone and addition of a drop of
concentrated hydrochloric acid gave a brown, slightly
hygroscopic solid. This was filtered, recrystallized from
2-propanol, and filtered and washed with acetone under
nitrogen to give 3a as the hydrochloride salt (0.12 g),
mp indistinct (darkened and shrank >160 ꢁC). 1H NMR
[(CD3)2SO] d 2.85 (d, J = 4.8 Hz, 6H, +NHMe2), 3.66
(q, 2H, J = 6.5 Hz, CH2), 7.74–7.92 (m, 2H), 8.26 (t,
1H, J = 7.4 Hz), 8.35–8.49 (m, 2H), 8.57 (d, 1H,
J = 7.3 Hz), 9.52 (s, 1H), 10.03 (br s, NH).
4.25. N-[2-(Dimethylamino)ethyl]-2-dimethylamino-7-
oxo-7H-naphtho[1,2,3-de]quinoline-11-carboxamide (1g)
Amixture of acid 10c (0.1 g) in SOCl2 (3 mL) was kept
at 50 ꢁC for 45 min, then treated as for 3a to give the
crude amide (0.040 g, 33%) as a red semisolid. From
PTLC [silica:chloroform/diethylamine (10:1)], an orange
band (Rf 0.32) was extracted with chloroform to give 1g
1
as an orange semisolid (0.01 g). H NMR (CDCl3) d
2.16 (s, 6H, N(CH3)2), 2.49 (t, 2H, J = 5.3 Hz), 3.33 (s,
6H, N(CH3)2), 3.57 (q, 2H, J = 5.3 Hz), 6.72 (br s 1H,
NH), 7.55–7.71 (m, 3H), 7.89 (s, 1H), 7.98 (d, 1H, J =
8.3 Hz), 8.21 (d, 1H, J = 7.6 Hz), 8.50 (d, 1H,
J = 7.6 Hz). HRMS (EI): Calcd for C23H24N4O2:
388.1900 (M+). Found: 388.1899.
4.26. DNA binding studies
Fluorescence was measured in 0.01 ionic strength buf-
fered sodium chloride (pH 7.0) containing ethidium bro-
mide (1.260 lM) and either poly-(dA-dT)Æ(dA-dT) and
poly-(dG-dC)Æ(dG-dC) at a concentration of 1 lM in
nucleotides. Aqueous solutions of the drugs as chloride
salts were added stepwise and fluorescence was mea-
sured at 625 nm with excitation at 595 nm. Drug con-
centrations for 50% reduction in DNA-ethidium
fluorescence (with correction for the fluorescence of free
ethidium) were calculated and binding constants were
calculated from the assumption that ethidium and the
1
The free base of 3a was a red-brown flaky solid. H
NMR (CDCl3) d 2.24 (s, 6H, NMe2), 2.65 (t, 2H,
J = 6.5 Hz, CH2), 3.69 (q, 2H, J = 6.7 Hz, CH2), 6.45
(br s, NH), 7.75–7.77 (m, 2H), 8.09 (t, 1H, J = 7.4 Hz),
8.35 (d, 1H, J = 8.2 Hz), 8.49–8.52 (m, 1H), 8.61 (d,
1H, J = 8.0 Hz), 9.47 (s, 1H). 13C NMR (CDCl3) d
37.3 (CH2), 44.7 (CH3), 57.0 (CH2), 119.5 (C), 127.7
(C), 128.6 (CH), 129.3 (CH), 131.8 (CH), 133.6 (CH),
133.7 (CH), 133.9 (C), 135.1 (CH), 137.4 (C), 149.0
(C), 155.1 (CH), 155.5 (C), 170.4 (C), 180.9 (C).
HRMS(EI): Calcd for C20H18N4O2: 346.1430 (M+).
Found: 346.1432.
6
drugs competed for the same DNAbinding site.
4.27. In vitro cytotoxicity assays
Murine P388 leukemia cells, Lewis lung carcinoma cells
(LLTC), and human Jurkat leukemia cells (JLC), to-
gether with their amsacrine and doxorubicin-resistant
derivatives (JLA and JLD, respectively), were obtained
and cultured as described.15 Growth inhibition assays
were performed by culturing cells at 4.5 · 103 (P388),
and 103 (LLTC and Jurkat) per well in microculture
plates (150 mL per well) for 3 days (P388) or 4 days
(LLTC and Jurkat) in the presence of drug. Cell growth
was determined by [3H]-thymidine uptake over the last
2 h of incubation (P388)16 or by sulforhodamine stain-
ing (LLTC and Jurkat).17 Results represent means of at
least two independent assays.
4.24. N-[2-(Dimethylamino)ethyl]-2-methyl-7-oxo-7H-
benzo[e]perimidine-11-carboxamide (3b)
Acid 8e (0.16 g; the light brown second crop) was re-
acted with SOCl2 (50 ꢁC for 45 min) and then with
N,N-dimethylethylenediamine as above for 3a, to give
1
the crude amide 3b as a dark solid (0.10 g, 50%). H
NMR (CDCl3) d 2.39 (s, 6H, NMe2), 2.80 (br t, 2H,
CH2), 2.90 (s, 3H, CH3), 3.77 (q, 2H, J = 6.7 Hz,
CH2), 6.82 (br s, NH), 7.6–7.75 (m, 2H), 8.00 (t, 1H,
J = 8.3 Hz), 8.20 (d, 1H, J = 8.2 Hz), 8.41–8.48 (m,
2H). This was converted in EtOH to an oxalate salt,
which separated as a sticky solid upon addition of ether.
This product was boiled in EtOH, the liquid was dec-
anted from some sticky residue, and the salt was repre-
cipitated with ether. This was repeated, then the salt was
recrystallized twice from MeCN and filtered in a nitro-
gen stream (the solid rapidly went sticky in air when
4.28. In vivo tumor assays
Procedures followed institutional ethical guidelines.
Colon 38 tumors were grown subcutaneously from
1 mm3 fragments in one flank of BDF1 mice (anesthetized
with pentobarbitone 90 mg/kg). When tumors reached a

X. Bu et al. / Bioorg. Med. Chem. 13 (2005) 3657–3665
3665
7. Palmer, B. D.; Rewcastle, G. W.; Baguley, B. C.; Denny,
W. A. J. Med. Chem. 1988, 31, 707.
8. Wilkoff, L. J.; Dulmadge, E.; Chopra, D. P. Proc. Soc.
Exp. Biol. Med. 1980, 163, 233.
9. Finlay, G. J.; Baguley, B. C.; Snow, K.; Judd, W. J. Natl.
Cancer Inst. 1990, 82, 662.
10. Finlay, G. J.; Holdaway, K. M.; Baguley, B. C. Oncol.
Res. 1994, 6, 33.
diameter of ꢁ4–6 mm (7–8 days), mice were divided into
control and drug treatment groups (5 mice/group), with
similar average tumor volumes in each group. The drugs,
as solutions of the hydrochloride salts in distilled water,
were injected intraperitoneally as a single dose in a vol-
ume of 0.01 mL/g body weight. Tumor diameters were
measured with calipers three times a week and tumor vol-
umes were calculated as 0.52 · a2 · b, where a and b are
the minor and major tumor axes and data plotted on a
semilogarithmic graph as mean tumor volumes
( SEM) versus time after treatment. Growth delay was
calculated as the time taken for a tumor to reach a mean
volume four-fold higher than its pre-treatment volume.
11. Spicer, J. A.; Gamage, S. A.; Atwell, G. J.; Finlay, G. F.;
Baguley, B. C.; Denny, W. A. J. Med. Chem. 1997, 40,
1919.
12. Vicker, N.; Burgess, L.; Chuckowree, I. S.; Dodd, R.;
Folkes, A.; Hardick, D. J.; Hancox, T. C.; Miller, W.;
Milton, J.; Sohal, S.; Wang, S.; Wren, S. P.; Charlton, P.;
Dangerfield, W.; Liddle, C.; Mistry, P.; Stewart, A.;
Denny, W. A. J. Med. Chem. 2002, 45, 721.
13. Sappal, D. S.; McClendon, A. K.; Fleming, J. A.;
Thoroddsen, V.; Connolly, K.; Reimer, C.; Blackman,
R. K.; Bulawa, C. E.; Osheroff, N.; Charlton, P.;
Rudolph-Owen, L. A. Mol. Cancer Ther. 2004, 3, 47.
14. Weidinger, H.; Eilingsfeld, H.; Haese, G. Gr. Patent
1,159,456, 1963. Chem. Abstr. 1964, 60, 14645.
15. Finlay, G. J.; Riou, J.-F.; Baguley, B. C. Eur. J. Cancer
1996, 32A, 708.
References and notes
1. Bu, X.; Deady, L. W.; Finlay, G. J.; Baguley, B. C.;
Denny, W. A. J. Med. Chem. 2001, 44, 2004.
2. Rule, H. G.; Pursell, W.; Brown, R. R. H. J. Chem. Soc.
1934, 168.
3. Rule, H. G.; Smith, F. R. J. Chem. Soc. 1937,
1096.
16. Marshall, E. S.; Holdaway, K. M.; Shaw, J. H. F.; Finlay,
G. J.; Matthews, J. H. L.; Baguley, B. C. Oncol. Res. 1993,
5, 301.
4. Krasnova, L. B.; Popov, S. I.; Dokunikin, N. S. Russ. J.
Org. Chem. 1973, 9, 1523.
17. Skehan, P.; Storeng, R.; Scudiero, D.; Monks, A.;
McMahon, J.; Vistica, D.; Warren, J. T.; Bokesch, H.;
Kenney, S.; Boyd, M. R. J. Natl. Cancer Inst. 1990, 82,
1107.
5. Eiden, F.; Durr, M. Arch. Pharm. (Weinheim) 1979, 312,
¨
708.
6. Baguley, B. C.; Denny, W. A.; Atwell, G. J.; Cain, B. F.
J. Med. Chem. 1981, 24, 170.