5-Non-amino aromatic substituted naphthalimides as potential antitumor agents: Synthesis via Suzuki reaction, antiproliferative activity, and DNA-binding behavior

Article abstract of DOI:10.1016/j.bmc.2010.11.055

Amonafide is a naphthalimide derivative with antitumor activity and has failed to enter clinical phase III, because of its high-variable and unpredictable toxicity. In order to develop selective, efficient, and safe drugs, applying the 'nonfused' aromatic

Full text of DOI:10.1016/j.bmc.2010.11.055

Bioorganic & Medicinal Chemistry 19 (2011) 961–967
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
5-Non-amino aromatic substituted naphthalimides as potential
antitumor agents: Synthesis via Suzuki reaction, antiproliferative
activity, and DNA-binding behavior
Lijuan Xie a,c, Jingnan Cui b, Xuhong Qian c, , Yufang Xu c, Jianwen Liu c, Ruian Xu a
⇑
a Institute of Molecular Medicine, Huaqiao University, Chenghua North Road 269, Quanzhou 362021, China
b State Key Laboratory of Fine Chemicals, Dalian University of Technology, PO Box 89, Zhongshan Road 158, Dalian 116012, China
c Shanghai Key Laboratory of Chemical Biology, School of Pharmacy, East China University of Science and Technology, PO Box 544, 130 Meilong Road, Shanghai 200237, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Amonafide is a naphthalimide derivative with antitumor activity and has failed to enter clinical phase III,
because of its high-variable and unpredictable toxicity. In order to develop selective, efficient, and safe
drugs, applying the ‘nonfused’ aromatic system strategy, a series of 5-non-amino aromatic substituted
naphthalimides as replacement for amonafide were designed and were synthesized from naphthalic anhy-
dride by three steps including bromination, amination, and Pd(PPh3)4 catalyzed Suzuki reaction. These new
naphthalimide derivatives, except 4b, not only exhibited better activity than amonafide against HeLa and
P388D1 cell lines in vitro under the same experimental conditions, but also could avoid the side effect of
amonafide due to their structure, which lacks an easy acetylated arylamine at the 5 position. The DNA-bind-
ing behavior of the naphthalimide derivatives was also investigated, and the results suggested that they
bind to DNA via intercalation and 4a and 4g intercalated into DNA in different fashion.
Received 4 October 2010
Revised 23 November 2010
Accepted 23 November 2010
Available online 30 November 2010
Keyword:
5-Aromatic substituted naphthalimide
Amonafide
Suzuki reaction
DNA-binding
Ó 2010 Published by Elsevier Ltd.
1. Introduction
tor with lowest possible association constants may be better distri-
bution in vivo and generally present a broader spectrum of activity
Naphthalimide (benz[de]isoquinoline-1,3-dione) derivatives
have been extensively investigated as antitumor agents for many
years.1–4 Amonafide (see Fig. 1), a DNA-intercalator, was the first
compound that reached the clinical trial stage in this family and
showed good antitumor activity against advanced breast cancer.
However, amonafide has so far failed to enter clinical phase III be-
cause its amino group was easily metabolized to N-acetyl-amona-
fide (Fig. 1) by enzyme N-acetyltransferase 2 (NAT2), which caused
a high-variable, unpredictable toxicity (Fig. 1) among individu-
als.5,6 As a strategy to modify amonafide, the N-substituted amona-
fide derivatives were synthesized by Kiss and co-workers7 and us8
(Fig. 1A and B), respectively. These derivatives not only revealed
similar or better antitumor activity than amonafide in vitro or
in vivo and could avoid the side effect of amonafide due to their
structure, which lacks an easy acetylated arylamine. Whereas,
the above-mentioned amonafide derivatives had shortcoming, on
the one hand the numbers of the derivatives was limited according
to the Kiss’s synthetic method, on the other hand the yield was low
according to our synthetic method.
than structurally related compounds with higher binding con-
stants, and was applied to design some tricyclic carboxamides pre-
senting potent cytotoxicity where the third phenyl ring is
appended but not fused to the chromophore. Based on the results
obtained by Denny et al.9 Braña et al. designed and reported five
‘nonfused’ arylnaphthalimides10 (Fig. 1) with good antitumor
activity. These ‘nonfused’ arylnaphthalimides are in apparent con-
trast with an accepted concept that an efficient approach to en-
hance the antitumor activity of naphthalimides was to fuse one
or more aromatic heterocycles with the naphthalene core.3 Mean-
while, in the Braña’s report it aroused our attention that among the
arylnaphthalimides, except the 5-phenyl substituted dimeric com-
pound, the others 5-aromatic substituted compounds were not
studied in detail, while the 5-phenyl substituted compound 4a
exhibited similar antitumor activity to amonafide and may be seen
as the ‘nonfused’ azonafide (Fig. 1). Moreover, existing research re-
sults confirmed that the substituent in 5-position of the naphtha-
lene ring is key to antitumor activity of the naphthalimides.1,3
Thus, on the one hand as a part of our continuing search for safe
and potential anticancer naphthalimides, on the other hand as an
attempt to explore whether the ‘nonfused’ strategy was an efficient
measure to design anticancer naphthalimides, we aimed at the
development of 5-non-amino aromatic substituted naphthalimides
as replacement for amonafide.
The ‘minimal intercalation’ was brought forward by Denny
et al.9 which was hypothesized that an anticancer DNA-intercala-
⇑
Corresponding author. Tel.: +86 21 64253589; fax: +86 21 64252603.
E-mail address: xhqian@ecust.edu.cn (X. Qian).
0968-0896/$ - see front matter Ó 2010 Published by Elsevier Ltd.
doi:10.1016/j.bmc.2010.11.055

962
L. Xie et al. / Bioorg. Med. Chem. 19 (2011) 961–967
O
N
N
N-acetyl-amonafide: R =
X
O
N
O
O
N
O
R'
R' = alkyl
A
R =
N
H
alkyl
R =
B
NH2
NHR
N-substituted amonafide derivatives
amonafide
N
N
N
R
O
Ar =
or
NH
O
N
O
O
N
O
S
N
O
O
N
O
Y
Y
2
Ar
target compound
azonafide
Y (C5) = phenyl Y (C5) = phenyl
Y (C5) = 2-furyl Y (C5) = 2-furyl
Y (C6) = phenyl
arylnaphthalimides
Figure 1.
According to the Braña’s method, the 5-phenyl substituted
In the synthesis procedure for target compounds, the third step
reaction was the most crucial and difficult, thus the coupling reac-
tion of 3 with phenylboric acid was chosen as the model reaction to
optimize the reaction conditions and detailed in Table 1. Initially,
the catalyst was evaluated. The cheaper Ni catalyst Ni(PPh3)2Cl2
has been explored for the model reaction (entry 1). The result
revealed that Ni(PPh3)2Cl2 failed to give the target compound 4a
and suggested that the Ni-based catalysts or the chosen catalyst
were not suitable for the substrate. For the Pd-catalyst have been
most widely used in the Suzuki reaction of organoboron com-
pounds with organic halides,11,12 PdCl2, Pd(PPh3)2Cl2, and
Pd(PPh3)4 were assayed respectively (entries 2–4). The yield was
the highest when Pd(PPh3)4 was used as catalyst. The results could
be related to the phenylboronic acid are inert to PdC12 and
Pd(PPh3)2Cl2, or the dissociation of the PPh3 is the easiest among
the three Pd-catalyst in reductive elimination procedure of the Su-
zuki reaction mechanism.11 Next, in order to determine suitable
reaction time, the yield was analyzed by HPLC detection during dif-
ferent reaction time (entries 4–6). According to the experimental
data, the yield approached a constant as the increase of reaction
time and 28 h was a suitable reaction time. This phenomenon
could be attributed to that the reaction involved Suzuki reaction
between 3 and phenylboronic acid, self-coupling reaction and
debrominated reaction of 3 reached the equilibrium along with
the reaction time. The debrominated product of 3 was found,
which gave evidence to the above explanation. It is well known
that base also play an important role in the Suzuki reaction, how-
ever, at present, the choice of base is still empirical, and no general
naphthalimide 4a was synthesized from naphthalic anhydride by
three steps including bromination, Stille coupling reaction and
amination. In Stille coupling procedure, the expensive and poison-
ous phenyl tributiltin was used as raw material, and was excessive
compared with naphthalic anhydride, moreover, the Pd(PPh3)4 was
also used as catalyst. The fact that 5-alkylamino substituted
amonafide derivatives8 (Fig. 1A) were successfully synthesized in-
spired us to improve the synthesis method for the target com-
pounds. The Suzuki–Miyaura cross-coupling reaction (also named
Suzuki reaction) has evolved into a most effective strategy for car-
bon–carbon (C–C) bond formation,11–13 and is less expensive and
less toxic compared with Stille reaction. In recent years, Suzuki
reaction has been greatly developed, and some previously inacces-
sible compounds were synthesized, nevertheless the scope of sub-
strate was mainly limited to simple aryl.11–13 In this paper, the
synthesis of 5-non-amino aromatic substituted naphthalimides
(Fig. 1) by Pd-catalyst Suzuki reaction, antiproliferative activity
in vitro and the DNA-binding are reported.
2. Results and discussion
2.1. Synthesis
The target compounds 4a–g were synthesized starting from
naphthalic anhydride by three steps including bromination, ami-
nation and Suzuki reaction by Pd-catalyst, and the synthetic route
was illustrated in the Scheme 1.
N
N
O
N
O
O
O
O
O
O
O
O
N
O
a
c
b
Br
Br
R
S
1
2
3
4a-g
F
F
4a: R =
4b: R =
4c: R =
4d: R =
4e: R =
4f: R =
4g: R =
F
CH3
CF3
NO2
Scheme 1. Reagents and conditions: (a) Br2, concd HNO3, 70 °C, 2 h, room temperature overnight; (b) NH2(CH2)2N(CH3)2, C2H5OH, reflux, 2 h; (c) aromatic boric acid with
substituted groups, Pd(PPh3)4, toluene reflux, 28 h.

L. Xie et al. / Bioorg. Med. Chem. 19 (2011) 961–967
963
Table 1
experiment are similar or lower than reported perylene bisimides
derivatives. This difference implied that the Suzuki reaction of
the perylene bisimides derivatives easily process compared with
the 5-position naphthalimide or our reaction condition would
deserved to be optimized.
The optimization of the reaction conditions in coupling reaction of 3 with phenyl-
boronic acid
Entry
Catalyst
Base
Solvent
t (h)
Yield (%)
1a
2b
3b
4b
5b
6b
7b
8b
9b
Ni(PPh3)2Cl2
PdCl2/PPh3
Pd(PPh3)2Cl2
Pd(PPh3)4
Pd(PPh3)4
Pd(PPh3)4
Pd(PPh3)4
Pd(PPh3)4
Pd(PPh3)4
K3PO4
K3PO4
K3PO4
K3PO4
K3PO4
K3PO4
K3PO4
K2CO3
t-BuOK
Toluene
Toluene
Toluene
Toluene
Toluene
Toluene
1,4-Dioxane
Toluene
Toluene
48
48
48
48
36
28
28
28
28
0
35
37
46
48
48
45
45
40
2.2. Antiproliferative activity in vitro
The in vitro antiproliferative activity of the 5-non-amino aro-
matic substituted naphthalimides against HeLa (human cervical
carcinoma cells) and P388D1 (murine lymphoid neoplasm cells)
were assessed by MTT tetrazolium dye assay17 and summarized
in Table 2, while amonafide was used as positive control. The
IC50 value of 4a and amonafide against HeLa cell in our study is
in agreement with reported data by Braña et al.10 Based on the
IC50 value of 4a, 4b, 4e, and 4f, we hypothesized that electro-with-
drawing group residing on the substituted phenyl could improve
the antitumor activity of the new naphthalimides. The hypothesis
was strengthened by synthesis and assay IC50 value of compound
4c and 4d against HeLa cell. It is well known that thiophene is a
bioisostere for benzene. 4g, a ‘nonfused’ thiophene substituted
naphthalimide, exhibited stronger antiproliferative activity than
amonafide. Referring to the Braña’s report that the furyl substi-
tuted naphthalimide10 is better bioactivity than amonafide, this
seems to indicate that the aromatic heterocycle substituted naph-
thalimide could possess better bioactivity. The experimental re-
sults displayed that all new naphthalimides except 4b exhibited
similar or stronger antiproliferative activity than amonafide
against HeLa cell and P388D1, moreover, the antiproliferative
activity against P388D1 cell was stronger than that against HeLa
under the same experimental conditions. Therefore, it can be
deduced that the ‘nonfused’ strategy can be applied to develop
the anticancer naphthalimides.
a
Reaction conditions:
(0.75 mmol), solvent (3 mL).
3
(0.5 mmol) Ni(PPh3)2Cl2 (0.015 mmol), base
b
Reaction conditions: 3 (0.5 mmol), Pd(PPh3)4 (0.015 mmol), base (0.75 mmol),
solvent (3 mL).
rule for their selection has been established.14 The K3PO4, K2CO3,
and t-BuOK have been investigated as base. The K3PO4 gave a bet-
ter yield (entries 6, 8 and 9). Last, as proper selection of solvent is
also essential for the reaction, the toluene and 1,4-dioxane were
also tested (entries 7 and 8). The results indicated that the effect
of toluene and 1,4-dioxane on the yield was similar, and toluene
was used as solvent due to its lower toxicity. On the basis of above
studies, it can be concluded that the Suzuki reaction carry out
using the Pd(PPh3)4 as catalyst and K3PO4 as the base in toluene
under reflux for 28 h. The desired compounds were synthesized
by the optimized conditions, and the results were described in
Table 2. To our delight, the reaction times of 5-pehnyl naphthali-
mide derivative were dramatically reduced as much as 42 h and
the yield was unchanged in comparison with the report by Braña
et al.10
To the best of our knowledge, so far no studies on the use of the
Pd-based catalyst Suzuki reaction into the naphthalimide system
were reported. Several papers13,15,16 described perylene bisimides
derivatives which are a similar structure to naphthalimide, have
been reported, while our study was conducting. The yields in our
2.3. DNA-binding studies
amonafide, therefore, the DNA-binding of the 5-aromatic
substituted naphthalimides derivatives was also studied. 4a bear-
Table 2
Preparation of 4a–g by Suzuki reaction of 3 with various substituted phenylboric acid and the antiproliferative activity against HeLa and P388D1 cell line in vitro
Kb (105 Mꢀ1
)
Kapp
b
b
c
Compound
R
Yield (%)
Cytotoxicity (IC50
P388D1
,
lM)
HeLa
5.70 0.09
HeLa
6.04 0.09
4a
4b
4c
4d
48
51
42
40
0.68 0.08
2.19 0.10
10.84 0.12
10.76 0.11
12.36 0.11
11.76 0.10
0.3175
0.3134
0.4335
0.4324
CH3
F
45.81 0.13
36.82 0.11
5.09 0.08
4.51 0.08
Not determined
Not determined
Not determined
Not determined
CF3
F
4r
4f
43
5.30 0.11
4.70 0.08
0.98 0.12
0.38 0.09
5.54 0.09
4.72 0.08
12.78 0.09
14.85 0.11
0.4341
1.0402
F
NO2
40
55
S
4g
3.10 0.09
6.02 0.09
0.68 0.11
0.68 0.08
3.70 0.08
6.45 0.09
13.21 0.07
1.05 0.07
0.5323
0.0559
NH2
Amonafide
a
b
c
Reaction conditions: compound 3 (0.5 mmol), Pd(PPh3)4 (0.025 mmol), K3PO4 (0.75 mmol), and toluene (3 mL).
Kb Scatchard binding constants which was calculated according to the fluorescence quenching technique.
Kapp apparent binding constants which was calculated according to the equation Kapp = ([EB]/[compound]50%) ꢁ KEB. where Kapp is the apparent binding constant of the
naphthalimides, KEB is the DNA-binding constant of EB, [EB] is the EB concentrations of the EB–DNA system and [compound]50% are the naphthalimides concentrations at 50%
fluorescence.

964
L. Xie et al. / Bioorg. Med. Chem. 19 (2011) 961–967
(c)
(a)
(b)
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
-0.1
1.2
1.1
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1.1
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
250
300
350
400
250
300
350
400
250
300
350
400
450
500
Wavelength (nm)
Wavelength (nm)
Wavelength (nm)
Figure 2. Absorption spectral changes of CT-DNA ([DNA] = 50 lM) in the presence of 4a (a), 4g (b) or amonafide (c), which was added in ratio R (R = [compound]/[DNA]) was
0, 0.02, 0.06, 0.1, 0.15, 0.2, 0.3, 0.4, and 0.5 in Tris–HCl buffer (30 mM, pH 7.5), respectively.
ing phenyl which is a typical substituted group and 4g was only
heteroacryl substituted derivative in the new naphthalimides, thus
4a and 4g were selected as study object.
A plot of rb/rf versus rb was constructed using linear regression
at the origin and the Scatchard binding constant was evaluated
from the slope of the plot (Fig. 3a and b, Table 2). In most cases,
compounds strongly binding to DNA are high cytotoxic agent,
whereas, the results revealed that there is no same relationship
between DNA-binding and cytotoxicity. Because the Kb value of
azonafide has not been reported, the DNA-binding between the
new 5-aromatic substituted naphthalimides and the azonafide, a
fused phenylnaphthalimides, can not be compared. Typical binding
constant between organic compound and DNA usually range from
104 to 106 Mꢀ1, in contrary to the guess that minimal intercalation
mode, these compounds may be strong DNA-intercalator, thus the
detailed action between DNA and the naphthalimides need further
study.
Primarily, the absorption spectra of CT-DNA in the absence and
presence of 4a were measured and showed in Figure 2a. As the
concentration of 4a was increased, the hyperchromic phenomenon
of the DNA solution was observed, which was attributed to absorp-
tion superposition of CT-DNA and 4a, meanwhile no significant
wavelength shift was seen. The absorption spectra of CT-DNA in
the absence and presence of 4g or amonafide (Fig. 2b or c) are sim-
ilar to the absorption spectra of CT-DNA-4a. These spectral charac-
teristics suggested that there are some similar interactions
between the compounds and DNA.
Next, the fluorescent properties (Fig. 3) were performed and
Scatchard binding constants (Kb) were calculated according to
the fluorescence quenching technique18,19 (Table 2). The emission
intensities of 4a and 4g decreased with increasing the amount of
In order to obtain further information regarding the DNA-
binding properties, the competitive binding between ethidium
bromide (EB) and these naphthalimides were carried out and
CT-DNA as most of the intercalators did20–23 and the wavelength
of 4g appeared blue shift (Fig. 3b). The blue shift would tend to
indicating that the 4g enters CT-DNA-stacking region with lower
polarity rather than the bulk solution of CT-DNA24 and implying
that the 4g more easily inserts into CT-DNA than 4a. Based on
the Scatchard method,18,19 two following Eqs. (1) and (2) are gen-
erated, where Co is the total concentration of bond and free drug,
CDNA is the free DNA concentration, and Fo and F1 are the fluores-
cence intensity before and after the intercalation of the drug into
DNA.
‘apparent’ equilibrium constants (Kapp
)
were calculated,
25,26
respectively (Table 2). EB emits weak fluorescence in aqueous
or buffer solution, whereas, displays a dramatic enhancement
of its emission intensity when it intercalate into DNA. If a mol-
ecule intercalates into DNA, it leads to a decrease in the binding
sites of DNA available for EB, which in turn decreases the fluo-
rescence intensity of the EB–DNA system. The addition of these
naphthalimides to EB–DNA system caused appreciable reduction
in emission intensity and indicated that these naphthalimides
compete with ethidium bromide in binding to DNA (no show)
25,26
by intercalation. The apparent equilibrium constants (Kapp
)
rb ¼ CoðFo ꢀ F1Þ=ðCDNAFoÞ
rb=rf ¼ CoðFo ꢀ F1Þ=ðCDNAF1Þ
ð1Þ
ð2Þ
was calculated from the competition experiments using Eq. (3)
and presented in Table 2.
(a)
(b)
1000
250
200
150
100
50
800
600
400
200
0
0
400
450
500
550
600
650
450
500
550
600
650
Wavelength (nm)
Wavelength (nm)
Figure 3. Fluorescence spectrum before and after interaction of 4a (a) or 4g (b) and CT-DNA (calf thymus DNA) in Tris–HCl buffer (30 mM, pH 7.5), and inserts show
Scatchard plot of binding of 4a or 4g. The concentration of 4a is 15, 10, 5, 3 and 1 M contained 50 M DNA, respectively. The concentration of 4 g is 25, 20, 15, 10, 7.5, 5, 3,
and 1 M contained 50 M DNA, respectively.
l
l
l
l

L. Xie et al. / Bioorg. Med. Chem. 19 (2011) 961–967
965
1.25
1.20
1.15
1.10
1.05
1.00
10
5
DNA
4a
4g
amonafide
4a
4g
0
-5
-10
0.00
0.02
0.04
0.06
0.08
0.10
200
300
400
500
600
[compound]/[DNA]
Wavelength (mdeg)
Figure 5. Effect of 4a (j), 4g (N), and amonafide (d) on the relative viscosities of
Figure 4. CD spectra of CT-DNA in the absence and presence of 4a and 4g at
CT-DNA at 25 ( 0.1) °C. [DNA] = 100 lM. g is the viscosity of DNA in the presence of
concentration of DNA 100
buffer (pH 7.0).
l
M, the concentration of 4a or 4g is 10
lM, in Tris–HCl
the compounds and g0 is the viscosity of DNA in the absence of the compounds.
versus the increase in concentration of 4a, 4g, and amonafide,
especially, the plot of 4g approach to a line, which predicated these
compounds bind to DNA via intercalation.
Kapp ¼ ð½EBꢂ=½compoundꢂ50%Þ ꢁ KEB
ð3Þ
Where Kapp is the apparent binding constant of the naphthali-
mides, KEB is the DNA-binding constant of EB, [EB] is the EB
concentrations of the EB–DNA system and [compound]50% are the
naphthalimides concentrations at 50% fluorescence. The DNA-
binding constants of EB reported vary considerably due to the exper-
imental condition, thereby, the Kapp of the naphthalimides were
showed in a number ꢁ KEB. While fluorescence intensity of EB–
DNA decrease 50% in the presence of a fluorescent molecule, the ra-
tio of concentration between the molecule and DNA is still less than
100 which is supposed that the molecule possesses the same inter-
action model like EB does.27 In our experiment, EB and DNA concen-
3. Conclusion
It is generally accepted that an efficient approach to enhance
the antitumor activity of naphthalimides was to fuse one or more
aromatic heterocycles with the naphthalene core. Therefore, this
research may provide some new suggestions for the design of no-
vel antitumor agents based on naphthalimides. The study demon-
strated that the Pd-catalyzed Suzuki coupling reaction can serve as
a valuable tool for the functionalization of naphthalimide system
and provided the replacements for amonafide. The replacements
not only exhibited improved antitumor activity over amonafide,
but also can avoid the side effects.
trations are 1.26 and 50 lM, respectively, according to the Kapp
equation, the all ratio of [compound]50%/[DNA] were less than 1,
thus the naphthalimides bind to DNA via intercalation. Usually,
the KEB is 107,28 and the Scatchard binding constant Kb and the
apparent binding constant Kapp are parallel.
CD (circular dichroism) is a very powerful technique to monitor
the conformational state of the DNA double helix in solution. Later,
the CD spectra of 4a-DNA and 4g-DNA were recorded and dis-
played in Figure 4, where the 275 nm band due to base stacking
and 248 nm band due to right-handed helicity. The increases in
the intensity of the positive band and the decrease in the intensity
of the negative band of 4a were observed, which was consistent
with the B to A-like conformational change,29 while the increase
in intensity of both the positive and negative bands of 4g were
4. Experimental
All the solvents are of analytic grade. 1H NMR and 13C NMR spec-
tra were measured on a Bruker AV-400 spectrometer with chemical
shifts reported in ppm (in CDCl3, TMS as internal standard). Melting
points were determined by using an X-6 micro-melting point appa-
ratus and are uncorrected. Column chromatography was performed
using silica gel 200–300 mesh. IR spectra were obtained using a
Nicolet 470 FT-IR instrument. High-resolution mass spectra were
obtained on a HP 1100 LC–MS spectrometer.
exhibited, which was typical of intercalation involving p-stacking
and stabilization of the right-handed B form of CT-DNA,30 after
the compounds were incubated with CT-DNA. A negative induced
CD (ICD) signal of 4g was found, which was ascribed to that 4g
intercalated into DNA with its long axis parallel to the base-pair,29
while ICD signal was not detected in 4a. ICD of the intercalated
compound is dependent on the orientation of the transition mo-
ment inside the intercalation site, their lateral displacement rela-
tive to the helix axis and the type of base pairs forming the
intercalation site. The difference of ICD signals between 4a and
4g additionally supported the notion that the compound 4a and
4g intercalated into DNA in different fashion.
Spectroscopic data are necessary, but not sufficient to support a
binding mode. Viscosity measurement is regarded as a reliable tool
to determine the binding model in solution in the absence of crys-
tallographic structural data and NMR data.20 Therefore, the viscos-
ity of DNA solution versus the concentration of 4a, 4g, and
amonafide were studied and were illustrated in Figure 5. Intercala-
tion of a molecule into DNA results in a lengthening, unwinding
and stiffening of the helix which increases the viscosity of the solu-
tion.20,22 The increase in viscosity of DNA solution was observed
4.1. Synthesis
4.1.1. 5-Bromo-2-[2-(dimethylamino)ethyl]-1H-benzo[de]iso-
quinoline-1,3(2H)-dione (3)
The compound 2 was prepared according to the report.31 The
compound 2 (277 mg, 1.0 mmol) and N,N-dimethylethyldiamine
(92 mg, 1.0 mmol) were refluxed in EtOH (15 mL) for 2 h to give
intermediate 3.
Yield 92%, an light brown solid, mp 77.2–78.6 °C; 1H NMR
(400 MHz CDCl3), d (ppm), 8.65 (s, 1H), 8.59 (dd, 1H, J = 1.0 and
7.2 Hz), 8.35 (s, 1H), 8.11 (dd, 1H, J = 0.7 and 8.3 Hz), 7.77 (t, 1H,
J = 7.8 Hz), 4.34 (t, 2H, J = 6.9 Hz), 2.70 (t, 2H, J = 6.9 Hz), 2.39
(s, 6H); MS (APCI) m/z 348 (M+1).
4.1.2. General procedure for the new naphthalimides (4a–g)
The intermediate 3 (174 mg, 0.5 mmol), Pd(PPh3)4 (12.5 mg,
0.01 mmol), K3PO4 (230.3 mg, 0.75 mmol) and aromatic boronic
acid with substituted groups (0.75 mmol) in dry toluene (3 mL)
were mixed, stirred and reflux for 28 h under nitrogen. The crude

966
L. Xie et al. / Bioorg. Med. Chem. 19 (2011) 961–967
products were concentrated under vacuum, and then were purified
by chromatography on silica gel with a solution of CH2Cl2 and
MeOH as eluent to give desired product.10
4.1.8. 5-(30-Nitro-phenyl)-2-[2-(dimethylamino)ethyl]-1H-benzo-
[de]isoquinoline-1,3(2H)-dione (4f)
Yield 40%, a pale yellow solid, mp 188.2–183.6 °C; 1H NMR
(400 MHz CDCl3), d (ppm), 8.87 (s, 1H), 8.64–8.63 (m, 2H), 8.45
(s, 1H), 8.33–8.30 (m, 2H), 8.11 (d, 1H, J = 8.0 Hz), 7.83 (t, 1H,
J = 8.0 Hz), 7.74 (t, 1H, J = 8.0 Hz), 4.39 (t, 2H, J = 6.8 Hz), 2.74 (t,
2H, J = 6.8 Hz), 2.41 (s, 6H); 13C NMR (100 MHz CDCl3), d (ppm),
165.2, 149.0, 141.0, 137.4, 134.2, 133.3, 132.1, 131.7, 130.3,
130.0, 127.8, 123.7, 123.1, 122.7, 122.2, 56.9, 49.9, 45.6, 38.1,
29.7, 22.7; IR (KBr cmꢀ1), 3333, 2933, 1695, 1658, 1385; HRMS
(ESI) m/z (M+H)+ calcd for C22H20N3O4 390.1430, found: 390.1434.
4.1.3. 5-Phenyl-2-[2-(dimethylamino)ethyl]-1H-benzo[de]iso-
quinoline-1,3(2H)-dione (4a)
Yield 48%, a white solid, mp 126.4–128.2 °C; 1H NMR (400 MHz
CDCl3), d (ppm), 8.87 (s, 1H), 8.58 (dd, 1H, J = 0.8 and 6.0 Hz), 8.38
(s, 1H), 8.25 (dd, 1H, J = 0.8 and 6.0 Hz), 7.78–7.75 (m, 3H), 7.53 (t,
2H, J = 6.0 Hz), 7.47–7.44 (m, 1H), 4.38 (t, 2H, J = 7.0 Hz), 2.73 (t,
2H, J = 7.0 Hz), 2.41 (s, 6H); 13C NMR (100 MHz CDCl3), d (ppm),
164.2, 140.1, 139.3, 134.1, 132.2, 131.3, 131.0, 130.8, 129.2,
128.4, 127.5, 127.4, 127.3, 123.2, 122.6, 57.0, 45.6, 38.0; IR (KBr
cmꢀ1), 3444, 2948, 1688, 1654; HRMS (EI) m/z (M+H)+ calcd for
4.1.9. 5-Thiophene-2-[2-(dimethylamino)ethyl]-1H-benzo[de]-
isoquinoline-1,3(2H)-dione (4g)
C
22H20N2O2 344.1525; found 344.1526.
Yield 55%, a yellow solid, mp 110.5–112.2 °C; 1H NMR (400 MHz
CDCl3), d (ppm), 8.87 (s, 1H), 8.55 (d, 1H, J = 7.2 Hz), 8.36 (s, 1H),
8.21 (d, 1H, J = 7.2 Hz), 7.76 (t, 1H, J = 7.6 Hz), 7.58 (dd, 1H, J = 0.8
and J = 4.0 Hz), 7.43 (dd, 1H, J = 0.8 and J = 4.0 Hz), 7.20–7.18 (m,
1H), 4.40 (t, 2H, J = 6.8 Hz), 2.78 (t, 2H, J = 6.8 Hz), 2.45 (s, 6H);
13C NMR (100 MHz CDCl3), d (ppm), 164.1, 142.3, 140.6, 136.9,
133.9, 133.4, 132.2, 130.9, 129.4, 129.3, 128.6, 127.6, 127.3,
126.4, 124.9, 123.2, 122.6, 56.8, 45.5, 37.8; IR (KBr cmꢀ1), 3348,
3103, 2370, 1691, 1654; HRMS (ESI) m/z (M+H)+ calcd for
4.1.4. 5-(40-Methyl-phenyl)-2-[2-(dimethylamino)ethyl]-1H-
benzo[de]isoquinoline-1,3(2H)-dione (4b)
Yield 51%, a white solid, mp 219.8–220.9 °C; 1H NMR (400 MHz
CDCl3), d (ppm), 8.87 (s, 1H), 8.57 (d, 1H, J = 8.0 Hz), 8.37 (s, 1H),
8.25 (d, 1H, J = 8.0 Hz), 7.79 (t, 1H, J = 7.2 Hz), 7.68 (d, 2H,
J = 8.0 Hz), 7.35 (d, 2H, J = 8.0 Hz), 4.41 (t, 2H, J = 6.8 Hz), 2.79 (t,
2H, J = 6.8 Hz), 2.45 (s, 9H); 13C NMR (100 MHz CDCl3), d (ppm),
164.3, 140.0, 138.4, 136.3, 134.1, 132.2, 130.9, 130.8, 130.0,
129.9, 127.2, 123.0, 122.5, 56.8, 45.4, 37.8, 21.2; IR (KBr cmꢀ1),
3370, 2940, 1694, 1663; HRMS (ESI) m/z (M+H)+ calcd for
C20H19N2O2S 351.1123, found: 351.1144.
4.2. Evaluation of in vitro cell proliferation by means of the MTT
colorimetric assay
C23H23N2O2 359.1735, found: 359.1736.
4.1.5. 5-(40-1F-Phenyl)-2-[2-(dimethylamino)ethyl]-1H-benzo-
[de]isoquinoline-1,3(2H)-dione (4c)
HeLa (human cervical carcinoma) and P388D1 (murine macro-
phage) cells were seeded into 96-well microculture plates at a den-
sity of 2.5 ꢁ 104 cells/well in 180
lL, maintained in 1640 complete
Yield 42%, a white solid, mp 115.1–116.3 °C; 1H NMR (400 MHz
CDCl3), d (ppm), 8.82 (d, 1H, J = 1.6 Hz), 8.59 (t, 1H, J = 6.8 Hz), 8.34
(d, 1H, J = 1.6 Hz), 8.27–8.25 (m, 1H), 7.80–7.72 (m, 3H), 7.25–7.21
(m, 2H), 4.43 (t, 2H, J = 6.8 Hz), 2.80 (t, 2H, J = 6.8 Hz), 2.51 (s, 6H);
13C NMR (100 MHz CDCl3), d (ppm), 164.5, 162.7, 162.1, 139.3,
135.6, 134.3, 132.4, 131.6, 130.8, 129.4, 129.3, 127.5, 123.4,
122.8, 116.5, 116.3, 56.8, 45.4, 36.6; 19F NMR d (ppm), ꢀ114.2; IR
(KBr cmꢀ1), 2925, 2854, 1747, 1697, 1662, 1172; HRMS (ESI) m/z
(M+H)+ calcd for C22H20N2O2F 363.1509, found: 363.1522.
medium and incubated at 37 °C in a 5% CO2 atmosphere for 24 h.
The D-hank’s buffer was used as a negative control and amonafide
as a positive control. After cells were exposed to compounds in
20
were washed twice with PBS (phosphate-buffered saline, pH 7.4),
and then 100 L/well of MTT reagent (3-(4,5-dimethylthiazol-2-
yl)-2,5-diphenyltetrazolium bromide, 0.5 g/ L in PBS; Sigma) in
lL/well at concentrations from 100 to 0.01 lM for 48 h, the cells
l
l
l
1640 complete medium were added. The plates were returned to
the incubator for 4 h. Subsequently, DMSO was added as solvent.
Absorbance was determined at 492 nm (with a reference at
630 nm) with a microplate reader. All experiments were per-
formed at least three times, and the average of the percentage
absorbance was plotted against concentration. The concentration
of drug required to inhibit 50% of cell growth (IC50) was then
calculated for each compound.
4.1.6. 5-(40-Trifluomethyl-phenyl)-2-[2-(dimethylamino)ethyl]-
1H-benzo[de]isoquinoline-1,3(2H)-dione (4d)
Yield 40%, a white solid, mp 144.7–146.1 °C; 1H NMR (400 MHz
CDCl3), d (ppm), 8.79 (d, 1H, J = 6.4 Hz), 8.52 (t, 1H, J = 8.0 Hz), 8.32
(d, 1H, J = 7.2 Hz), 8.22–8.14 (m, 1H), 7.74–7.67 (m, 2H), 7.47 (t, 1H,
J = 7.2 Hz), 7.39 (t, 1H, J = 7.2 Hz), 7.20 (s, 1H), 4.38 (t, 2H,
J = 6.8 Hz), 2.82 (t, 2H, J = 6.8 Hz), 2.45 (s, 6H); 13C NMR
(100 MHz CDCl3), d (ppm), 173.4, 173.0, 164.4, 139.4, 134.3,
132.3, 131.9, 131.3, 129.4, 128.6, 127.9, 127.9, 126.3, 123.3,
122.8, 56.7, 45.2, 37.6; 19F NMR d (ppm), ꢀ63.0; IR (KBr cmꢀ1),
2925, 2854, 1746, 1697, 1663, 1171; HRMS (ESI) m/z (M+H)+ calcd
for C23H20N2O2F3, 413.1477 found: 413.1469.
4.3. DNA-binding studies
UV–vis absorption spectra were recorded on a PGENERAL
TU-1901 UV–vis spectrophotometer and fluorescent spectra were
measured on a Hitachi F-4500 luminescence spectrophotometer.
Calf-thymus (CT) DNA was purchased from the Sino–American
Biotechnology Company. Solution of CT-DNA in Tris–HCl buffer
(30 mM, pH 7.5) gave a ratio of UV absorbance at 260 and 280 nm
of 1.8–1.9:1, indicating that the DNA was sufficiently free from pro-
tein. The concentration of CT-DNA was determined by its absorption
intensity at 260 nm witha known molar absorption coefficient value
4.1.7. 5-(30,40-2F-Phenyl)-2-[2-(dimethylamino)ethyl]-1H-benzo-
[de]isoquinoline-1,3(2H)-dione (4e)
Yield 43%, a white solid, mp 115.1–116.3 °C; 1H NMR (400 MHz
CDCl3), d (ppm), 8.78 (d, 1H, J = 2.0 Hz), 8.61 (d, 1H, J = 7.2 Hz), 8.32
(d, 1H, J = 2.0 Hz), 8.26 (d, 1H, J = 8.0 Hz), 7.80 (t, 1H, J = 8.0 Hz),
7.61–7.56 (m, 1H), 7.53–7.49 (m, 1H), 7.37–7.30 (m, 1H), 4.37 (t,
2H, J = 6.8 Hz), 2.69 (t, 2H, J = 6.8 Hz), 2.37 (s, 6H); 13C NMR
(100 MHz CDCl3), d (ppm), 164.0, 151.9, 149.4, 137.9, 136.4,
136.3, 134.0, 132.1, 131.3, 131.1, 130.1, 127.5, 123.5,122.7, 118.2,
118.0, 116.5, 116.4, 57.0, 45.8, 38.3; 19F NMR d (ppm), ꢀ136.4,
ꢀ138.1; IR (KBr cmꢀ1), 2940, 1703, 1654, 1142; HRMS (ESI) m/z
(M+H)+ calcd for C22H19N2O2F2 381.1415, found: 381.1396.
of 6600 Mꢀ1 cmꢀ1
.
4.3.1. UV–vis absorption spectra studies
The titration absorption spectra studies were performed by
keeping constant the concentration of DNA while varying the
concentration of compound at room temperature. Initially, solu-
tions of the blank buffer were placed in the reference and sam-
ple cuvettes (1 cm path length), respectively, and then the first

L. Xie et al. / Bioorg. Med. Chem. 19 (2011) 961–967
967
spectrum was recorded in the range 200–600 nm. During the
Acknowledgment
titration, aliquots of buffered compound solution were added
and the solutions were mixed by repeated inversion. After mix-
ing for 10 min, the absorption spectra were recorded. The titra-
tion processes were repeated until there was no change in the
spectra for at least four titrations indicating binding saturation
had been achieved.
This work was financially supported by the Key New Drug
Creation and Manufacturing Program (2009ZX09103-102), the
China 111 Project (Grant B07023), and the Scientific Research
Foundation of the Huaqiao University (09BS406).
Supplementary data
4.3.2. Fluorescent spectra and EB competitive studies
The two groups of samples for experiments were prepared, one
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2010.11.055. These data
include MOL files and InChiKeys of the most important compounds
described in this article.
at a constant DNA concentration of 50
lM and at concentrations
compounds ranging from 1 M to 25
l
lM in Tris–HCl (30 mM, pH
7.5), and the other having the same concentration of compound
but absence of DNA as control. All the above solutions were
ultrasonic shaken for 1 day at 25 °C in the dark. Fluorescence
wavelength and intensity area of the samples were measured by
the excitation wavelength at 350 nm.
DNA and EB were dissolved in buffer Tris–HCl (30 mM, pH 7.5)
at the concentrations of 50 and 1.26 lM, respectively, and then the
fluorescent spectra was performed. Different amounts of naphthal-
imides were added to EB bound with CT-DNA, mixed and emission
spectra of EB were measured using excitation wavelength 518 nm.
Before examining the fluorescent properties of EB, it was checked if
the used naphthalimides did not quench the EB fluorescence. The
plot of fluorescent intensity of EB–DNA system versus concentra-
tion of the naphthalimides were constructed at the Origin
software, the [compound]50% were obtained and the apparent
binding Kapp constant were calculated from Eq. (3).
References and notes
1. Braña, M. F.; Ramos, A. Curr. Med. Chem. Anticancer Agents 2001, 1, 237.
2. Braña, M. F.; Cacho, M.; Gradillas, A.; de Pascual-Teresa, B.; Ramos, A. Curr.
Pharm. Des. 2001, 7, 1745.
3. Ingrassia, L.; Lefranc, F.; Kiss, R.; Mijatovic, T. Curr. Med. Chem. 2009, 16, 1192.
4. Lv, M.; Xu, H. Curr. Med. Chem. 2009, 16, 4797.
5. Ratain, M. J.; Mick, R.; Berezin, F.; Janisch, L.; Schilsky, R. L.; Williams, S. F.;
Smiddy, J. Clin. Pharmacol. Ther. 1991, 50, 573.
6. Ratain, M. J.; Rosner, G.; Allen, S. L.; Costanza, M.; Van Echo, D. A.; Henderson, I.
C.; Schilsky, R. L. J. Clin. Oncol. 1995, 13, 741.
7. Quaquebeke, E. V.; Mahieu, T.; Dumont, P.; Dewelle, J.; Ribaucour, F.; Simon, G.;
Sauvage, S.; Gaussin, J. F.; Tuti, J.; Yazidi, M. E.; Vynckt, F. V.; Mijatovic, T.;
Lefranc, F.; Darro, F.; Kiss, R. J. Med. Chem. 2007, 50, 4122.
8. Xie, L.; Xu, Y.; Wang, F.; Liu, J.; Qian, X.; Cui, J. Bioorg. Med. Chem. 2009, 16, 804.
9. Braña, M. F.; Gradillas, A.; Gómez, A.; Acero, N.; Llinares, F.; Muñoz-Mingarro,
D.; Abradelo, C.; Rey-Stolle, F.; Yuste, M.; Campos, J.; Gallo, M. Á.; Espinosa, A. J.
Med. Chem. 2004, 47, 2236.
10. Atwell, G. J.; Baguley, B. C.; Denny, W. A. J. Med. Chem. 1989, 32, 396.
11. Miyaura, N.; Suzuki, A. Chem. Rev. 1995, 95, 2457.
12. Harvey, R. G. Curr. Org. Chem. 2004, 8, 303.
Kapp ¼ ð½EBꢂ=½compoundꢂ50%Þ ꢁ KEB
ð3Þ
13. Qiu, W.; Chen, S.; Sun, X.; Liu, Y.; Zhu, D. Org. Lett. 2006, 8, 867.
14. Martin, R.; Buchwald, S. L. Acc. Chem. Res. 2008, 41, 1461.
15. Bhosale, S. V.; Kalyankar, M. B.; Bhosale, S. V.; Langford, S. J.; Reid, E. F.; Hogan,
C. F. New J. Chem. 2009, 33, 2409.
16. Jiang, W.; Li, Y.; Yue, W.; Zhen, Y.; Qu, J.; Wang, Z. Org. Lett. 2010, 12, 228.
17. Kuroda, M.; Mimaki, Y.; Sashida, Y.; Hirano, T.; Oka, K.; Dobashi, A.; Li, H.;
Harada, N. Tetrahedron 1997, 53, 11549.
4.3.3. Circular dichroism spectra studies
The CD (circular dichroism) spectra were scanned with a J-
810 spectrophotometer (Jasco, Japan) using a 1-cm path quartz
cell and subtracted from the spectrum of Tris–HCl buffer alone.
The CD spectra were recorded at the compound concentration
of 10 lM and DNA concentration of 100 lM, in the region
200–600 nm.
18. Scatchard, G. Ann. N.Y. Acad. Sci. 1949, 51, 660.
19. Gupta, M.; Ali, R. J. Biochem. 1984, 95, 1253.
20. Ihmels, H.; Otto, D. Top. Curr. Chem. 2005, 258, 161.
21. Long, E. C.; Barton, J. K. Acc. Chem. Res. 1990, 23, 271.
22. Palchaudhuri, R.; Hergenrother, P. J. Curr. Opin. Biotechnol. 2007, 18, 497.
23. Wheate, N. J.; Brodie, C. R.; Collins, J. G.; Kemp, S.; Aldrich-Wright, J. R. Mini-
Rev. Med. Chem. 2007, 7, 627.
24. Yang, X.; Liu, W.; Jin, W.; Shen, G.; Yu, R. Spectrochim. Acta, Part A 1999, 55,
2719.
25. Yegorova, A.; Karasyov, A.; Duerkop, A.; Ukrainets, I.; Antonovich, V.
Spectrochim. Acta, Part A 2005, 61, 109.
26. Han, M.; Duan, Z.; Hao, Q.; Zheng, S.; Wang, K. J. Phys. C. 2007, 111, 16577.
27. Wu, F.; Xie, F.; Wu, Y.; Hong, J. J. Fluoresc. 2008, 18, 175.
28. Antonini, I.; Polucci, P.; Magnano, A.; Sparapani, S.; Martelli, S. J. Med. Chem.
2004, 47, 5244.
29. Zhang, Z.; Yang, Y.; Zhang, D.; Wang, Y.; Qian, X.; Liu, F. Bioorg. Med. Chem.
2006, 14, 6962.
30. Maheswari, P. U.; Palaniandavar, M. J. Inorg. Biochem. 2004, 98, 219.
31. Rule, H. G.; Thompson, S. B. J. Chem. Soc. 1937, 1761.
32. Basili, S.; Bergen, A.; Dall’Acqua, F.; Faccio, A.; Ranzhan, A.; Ihmels, H.; Moro, S.;
Viola, G. Biochemistry 2007, 46, 12721.
33. Li, F.; Cui, J.; Guo, L.; Qian, X.; Ren, W.; Wang, K.; Liu, F. Bioorg. Med. Chem. 2007,
15, 5114.
4.3.4. Viscosity experiments
Calf-thymus DNA was dissolved in Tris–HCl buffer (30 mM, pH
7.5) and left at 4 °C overnight. It was treated in an ultrasonic bath
for 10 min, and the solution was filtered through a PVDF mem-
brane filter (pore size of 0.45
the concentration of CT-DNA was 100
l
m) to remove insoluble material,
l
M.31 Viscometric titrations
were performed with an Ubbelodhe viscometer immersed in a
thermostated bath maintained 25 ( 0.1) °C. The flow times were
measured with a digital stopwatch, each sample was measured
three times, and an average flow time was calculated. Data are pre-
sented as (g/g ) g is the viscosity
0 1/3 versus [complex]/[DNA], where
of DNA in the presence of complex and g0 is the viscosity of DNA
alone. Viscosity values were calculated from the observed flowing
time of DNA-containing solutions (t) corrected for that of the buffer
alone (t0),
g
= (t ꢀ t0).32,33