Synthesis and biological activities of bisnaphthalimido polyamines derivatives: Cytotoxicity, DNA binding, DNA damage and drug localization in breast cancer MCF 7 cells

Article abstract of DOI:10.1016/j.bcp.2004.09.020

New bisoxynaphthalimidopolyamines (BNIPOPut, BNIPOSpd and BNIPOSpm) were synthesized. Their cytotoxic properties were evaluated against breast cancer MCF 7 cells and compared with bisnaphthalimidopolyamines BNIPSpd and BNIPSpm. Among the bisoxynaphthalimi

Full text of DOI:10.1016/j.bcp.2004.09.020

Biochemical Pharmacology 69 (2005) 19–27
www.elsevier.com/locate/biochempharm
Synthesis and biological activities of bisnaphthalimido polyamines
derivatives: cytotoxicity, DNA binding, DNA damage and drug
localization in breast cancer MCF 7 cells
Anne-Marie Dancea, Lynda Raltona, Zoe Fullera, Lesley Milneb, Susan Duthieb,
Charles S. Bestwickb, Paul Kong Thoo Lina,
*
aThe Robert Gordon University, School of Life Sciences, St. Andrew Street, Aberdeen AB251HG, Scotland, UK
bThe Rowett Research Institute, Greenburn Road, Aberdeen AB299SB, Scotland, UK
Received 9 June 2004; accepted 8 September 2004
Abstract
New bisoxynaphthalimidopolyamines (BNIPOPut, BNIPOSpd and BNIPOSpm) were synthesized. Their cytotoxic properties were
evaluated against breast cancer MCF 7 cells and compared with bisnaphthalimidopolyamines BNIPSpd and BNIPSpm. Among the
bisoxynaphthalimido polyamines, BNIPOSpm and BNIPOSpd exhibited cytotoxic activity with an IC50 of 29.55 and 27.22 mM,
respectively, while BNIPOPut failed to exert significant cytotoxicity after 48-h drug exposure. DNA binding was determined by midpoint
of thermal denaturation (Tm) measurement, ethidium bromide displacement and DNA gel mobility. Both BNIPOSpm and BNIPOSpd
exhibited strong binding affinities with DNA. BNIPOPut had the least effect. The results were compared with other cytotoxic
bisnaphthalimido compounds (BNIPSpm and BNIPSpd) previously reported by us. Using the single cell gel electrophoresis assay, it
was found that BNIPSpm and BNIPSpd caused substantial DNA damage to MCF 7 treated cells while BNIPOSpm showed no significant
effect over a range of drug concentrations after 4-h drug exposure. However, after 12-h drug exposure, BNIPOSpm had induced significant
DNA damage similar to that of BNIPSpm (after 4-h drug exposure). Fluorescence microscopic analysis revealed that at 1 mM drug
concentration and after 6-h drug exposure, both BNIPSpm and BNIPSpd were located within the cell while the presence of BNIPOSpm,
was not observed. Therefore, we conclude that BNIPSpd, BNIPSpm and BNIPOSpm induce DNA damage consistent with their rate of
uptake into the cells.
# 2004 Elsevier Inc. All rights reserved.
Keywords: Bisnaphthalimides; Synthesis; Cytotoxicity; DNA binding and damage; Single cell gel electrophoresis assay; Fluorescence microscopy
1. Introduction
and as tools to understand a number of biological processes
[2].
Studies on molecules interacting with DNA are rapidly
generating intense interest in many research laboratories
[1,2]. Some of them have found applications as anticancer
agents [3], as important dyes used in molecular biology [4]
Naphthalimides and bisnaphthalimides are well-known
cytotoxic DNA intercalating agents and have shown pro-
mise as potential anti-cancer agents [5]. Recently the
bisnaphthalimide, elinafide (Fig. 1) has gone into phase
I clinical trials in Europe and the USA [6]. NMR spectro-
scopy, suggests that elinafide bisintercalates to specific
DNA sequence with the propylene diamine linker lying
in the major groove of the DNA molecule [7]. Although
bisnaphthalimides have shown to have higher anticancer
activity than the mononaphthalimides, they are however
very insoluble and can render testing difficult [8].
In a previous report, we described the synthesis and
cytotoxic properties of bisaminopropylnaphthalimido
Abbreviations: BNIPPut, bisnaphthalimidopropylputrescine; BNIPSpd,
bisnaphthalimidopropylspermidine; BNIPSpm, bisnaphthalimidopropyl-
spermine; BNIPOPut, bisnaphthalimidooxypropylputrescine; BNIPOSpd,
bisnaphthalimidooxypropylspermidine; BNIPOSpm, bisnaphthalimidoox-
ypropylspermine; DMF, dimethylformamide; THF, tetrahydrofuran; MTT,
3-(4,5-diemthylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; Tm, mid-
point of thermal denaturation
* Corresponding author. Tel.: +44 1224 262818; fax: +44 1224 262828.
E-mail address: p.kong@rgu.ac.uk (P.K.T. Lin).
0006-2952/$ – see front matter # 2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.bcp.2004.09.020

20
A.-M. Dance et al. / Biochemical Pharmacology 69 (2005) 19–27
Fig. 1. Chemical structures of bisnaphthalimides.
polyamines (BNIPSpd, BNIPSpm in Fig. 1) containing
more than two nitrogen atoms in the linker chain with
improved solubility without unduly compromising their
biological activity [9,10]. We reason that the higher the
number of heteroatoms present in the linker chain, the
higher the solubility of these compounds. Furthermore,
they strongly stabilized DNA duplexes and were localized
preferentially inside the nucleus [10]. Previous studies in
our laboratory demonstrated that compound BNIPSpd
might take part in the initiation of apoptosis in MCF 7
breast and HL60 Leukemia cancer cells [11,12].
the timing of uptake and localisation to the extent of DNA
damage and cytotoxicity.
2. Materials and methods
2.1. Materials
Cell lines were obtained from ECACC. All reagents
were purchased from Aldrich, Fluka and Lancaster and
were used without purification. TLC was performed on
Kieselgel plates (Merck) 60 F254 in chloroform: methanol
(97:3 or 99:1). Column chromatography was done with
silica gel 60, 230–400 meshes using chloroform and
methanol as eluent. FAB-mass spectra were obtained on
a VG Analytical AutoSpec (25 kV) spectrometer; EC/CI
spectra were performed on a Micromass Quatro II (low
resolution) or a VG Analytical ZAB-E instrument (accu-
In this paper, we have extended our previous work [9,12]
by synthesizing three new bisaminooxypropylnaphthali-
mido polyamines derivatives, BNIPOPut, BNIPOSpd and
BNIPOSpm (Fig. 1) in which oxygen atoms are introduced
in the a-position with respect to the naphthalimido ring.
The rationale in synthesising BNIPOPut, BNIPOSpd and
BNIPOSpm is to show that such modification would
enhance the solubility of those compounds in aqueous
media without affecting their biological activity. Their
cytotoxicities against MCF 7 cells were evaluated and
their DNA binding affinities were studied using UV, fluor-
escence and gel mobility measurements. The results
obtained were compared with BNIPSpd and BNIPSpm
(Fig. 1). BNIPPut containing two nitrogen atoms in its
linker chain could not be used in our study because of its
high insolubility. We have also for the first time studied the
extent of DNA damage induced by BNIPSpd, BNIPSpm
and BNIPOSpm in drug treated MCF 7. Using fluorescence
microscopy, we found previously that BNIPSpd was
located in the cell nucleus after 8-h drug exposure [10].
In this paper, we have also extended the microscopic
studies with MCF 7 cells exposed to BNIPSpd, BNIPSpm
and BNIPOSpm over differing time scales in order to relate
1
rate mass). H and 13C NMR spectra were recorded on a
JEOL JNM-EX90 FT NMR spectrometer.
BNIPPut, BNIPSpd and BNIPSpm were synthesised
according to our methods previously reported [9,13].
2.2. Synthesis of 3-phthalimidooxy-1-propanol, 2
N-hydroxyphthalimide (9.2 g, 0.0565 mol) was dis-
solved in DMF (50 mL) followed by the addition of 3-
bromopropanol (7 g, 0.0504 mol) and triethylamine (5.7 g,
0.565 mol). The solution was left overnight at 85 8C. The
solvent was removed under vacuo and the residue dis-
solved in chloroform (50 mL). The organic phase was
washed with (i) water and (ii) saturated bicarbonate solu-
tion. The chloroform layer was dried with anhydrous
sodium sulphate and removal of solvent gave 2 (7.91 g,

A.-M. Dance et al. / Biochemical Pharmacology 69 (2005) 19–27
21
71%) as a waxy solid. 1H NMR (CDCl3): d = 7.70–8.00 (m,
4H, phthalimido protons), 4.50 (t, 2H, O–CH2), 4.10 (t, 2H,
–CH2–O), 2.50 (broad, 1H, OH), 2.10 (p, 2H, 2H, –CH2–).
13C NMR (CDCl3): d = 163.94 (C=O), 134.74, 129.15,
123.78 (aromatic C), 75.78 (OCH2), 59.28 (CH2O),
31.10 (CH2). Compound 2 was used in the next step
without further purification.
(CDCl3): d = 8.70–7.35 (m, 6H, aromatic protons), 4.45 (t,
2H, CH2), 4.35 (t, 2H, CH2), 2.50 (s, 3H, CH3), 2.25 (p, 2H,
CH2). 13C NMR (CDCl3): d = 161.30 (C=O), 145.10–
123.10 (aromatic carbons), 73.10, 67.90, 28.70
(3Â CH2), 22.10 (CH3). LRMS (FAB): Calcd. for
C12H19NO6S 425.09, found: 426 [MH]+.
2.6. General N-alkylation reaction (Fig. 2)
2.3. Synthesis of 3-aminooxy-1-propanol, 3
Mesitylated polyamines [9,12] (0.651 mmol) were dis-
solved in anhydrous DMF (13.5 mL) followed by the
Compound 2 (7.91 g, 0.0358 mol) was dissolved in
absolute ethanol followed by the addition of hydrazine
hydrate (1.32 g, 41.2 mmol). The solution was left stirring
at room temperature overnight. The resulting white pre-
cipitate was filtered off. The filtrate was evaporated to
dryness and the residue was resuspended in dichloro-
methane (25 mL). The solution was filtered and the filtrate
was reduced under vacuo to give 3 (2.33 g, 64.4%) as a
thick oil. 1H NMR (CDCl3): d = 4.20 (s, broad, 2H, ONH2),
3.50–4.00 (m, 4H, OCH2, CH2O), 1.70–2.00 (p, 2H, –
CH2–). 13C NMR (CDCl3): d = 75.00 (OCH2), 61.78
(CH2O), 32.71 (–CH2–). LRMS (FAB): Calcd. for
C3H9NO2 91, found: 92 [MH]+.
addition of
5 (0.13 mmol) and cesium carbonate
(1.06 g). The solution was left at 85 8C and completion
of the reaction was monitored by thin layer chromatogra-
phy. DMF was removed under vacuo and the residue was
dissolved in chloroform and extracted with (i) water and
(ii) saturated bicarbonate solution. The crude product was
purified by column chromatography.
2.7. General deprotection reaction (Fig. 2)
The fully protected polyamine derivatives (0.222 mmol)
were dissolved in anhydrous dichloromethane (10 mL)
followed by the addition of hydrobromic acid/glacial acetic
acid (1 mL). The solution was left stirring at room tem-
perature for 24 h. The yellow precipitate formed, was
filtered off and washed with dichloromethane, ethylacetate
and ether.
2.4. Synthesis of 3-(N-naphthalimidooxy)-
propan-1-ol, 4
Naphthalic anhydride (6.34 g, 0.032 mol) was dissolved
in DMF (50 mL) followed by the addition of aminooxy-
propanol 3 (2.33 g, 0.0256 mol) and DBU (4.87 g,
0.032 mol). The solution was left stirring at 85 8C. The
solvent DMF was removed under reduced pressure and the
resulting residue was dissolved in dichloromethane
(100 mL). The organic layer was washed with (i) water
and (ii) saturated bicarbonate solution. After drying with
anhydrous sodium sulphate and removal of the solvent,
gave a crude solid, which was subjected to column chro-
matography to give pure product 4 as a white powder
(1.97 g, 26.4%). NMR (CDCl3): d = 8.65–7.80 (m, 6H,
aromatic protons), 4.45 (t, 2H, –N–OCH2), 4.05 (t, 2H,
2.8. Synthesis of BNIPOPut
The reaction between N,N-dimesitylputrescine
6
with O-tosylpropyloxynaphthalimide 5 gave BNIPOPut
as yellow solid (48% yield). 1H NMR (DMSO-d6):
d = 8.62 (s, broad, –+NH2), 8.43–8.36 (m, aromatic hydro-
gens), 7.82–7.77 (m, aromatic hydrogens), 4.29–4.25 (t,
O–CH2–), 3.27–3.25 (m, N–CH2–), 3.10 (m, broad, N–
CH2–), 2.10–2.07 (m, –CH2–), 1.79 (m, broad, –CH2–).13C
NMR (DMSO-d6): d = 160.71 (C=O), 135.21–121.83
(aromatic carbons), 75.00–22.00 (10Â CH2). LRMS
(FAB): Calcd. for C34H34N4O6 2HBr 756.66, found: 426
[M–2HBr]+.
CH2–O–), 3.40 (s, broad, 1H, OH), 2.10 (p, 2H, CH2). 13
C
NMR (CDCl3): d = 161.70 (C=O), 135.70–122.90 (aro-
matic carbons), 74.90, 59.90, 30.90 (3Â CH2). LRMS
(FAB): Calcd. for C15H13NO4 271.08, found: 272 [MH]+.
2.9. Synthesis of BNIPOSpd
2.5. Synthesis of O-tosylpropyloxynaphthalimide, 5
The reaction between N,N,N-trimesitylspermidine 7
with O-tosylpropyloxynaphthalimide 5 gave BNIPOSpd
as yellow solid (72% yield). 1H NMR (DMSO-d6): d = 8.72
(s, broad, –+NH2), 8.62 (s, broad, –+NH2), 8.58–8.50 (m,
aromatic hydrogens), 7.95–7.85 (m, aromatic hydrogens),
4.35–4.25 (m, O–CH2–), 3.35–3.25 (m, N–CH2–), 3.15–
2.90 (m, N–CH2–), 2.20–2.00 (m, –CH2–). 13C NMR
(DMSO-d6): d = 161.00 (C=O), 135.50–122.80 (aromatic
carbons), 74.20–22.70 (13Â CH2). LRMS (FAB): Calcd.
for C37H41N5O6 3HBr 894.32, found: 654 [M–3HBr]+.
Compound 4 (1.97 g, 7.26 mmol) was dissolved in
anhydrous CH2Cl2 (8 mL) followed by the addition of
tosyl chloride (6.09 g, 0.032 mol) and triethylamine
(4.0 mL). The solution was left overnight at 50 8C. After
cooling the CH2Cl2 solution was washed with (i) water and
(ii) saturated bicarbonate solution. After drying and the
removal of the solvent, a dark residue was obtained which
was then subjected to column chromatography. The main
1
fraction gave 5 (1.02 g, 33%) as a buff solid. H NMR

22
A.-M. Dance et al. / Biochemical Pharmacology 69 (2005) 19–27
2.10. Synthesis of BNIPOSpm
change in midpoint of the thermal denaturation curves of
DNA.
The reaction between N,N,N,N-tetramesitylspermine 8
with O-tosylpropyloxynaphthalimide 5 gave BNIPOSpm
as yellow solid (68% yield). 1H NMR (DMSO-d6): d = 8.64
(s, broad, –+NH2), 8.48–8.36 (m, aromatic hydrogens),
7.82–7.77 (m, aromatic hydrogens), 4.29 (m, O–CH2–),
3.27 (m, N–CH2–), 3.10 (m, N–CH2–), 2.10–2.07 (m, –
CH2–), 1.79 (m, –CH2–). 13C NMR (DMSO-d6):
d = 161.07 (C=O), 135.58–122.20 (aromatic carbons),
74.20–23.12 (16Â CH2). LRMS (FAB): Calcd. for
C40H48N6O6 4HBr 1032.36, found: 709 [M–4HBr]+.
2.13. Fluorescence-binding studies
Binding studies were carried out by competitive dis-
placement fluorometric assay with DNA-bound ethidium.
Calf thymus DNA (Sigma) was used without further pur-
ification. C50 values are defined as the drug concentration
(mM), which generate a 50% decrease in the fluorescence
of the bound ethidium/DNA.
Test solutions were made by adding 160 mL of com-
pound in 0.01 M phosphate buffer (25 mM Na2HPO4,
15 mM NaH2PO4, and 1.25 mM EDTA) to a fixed volume
of calf thymus DNA (20 mL of 16.6 mM) and EB (20 mL of
20 mM) in 96-microplates wells. All measurements were
taken at room temperature using a Perkin-Elmer LS 55
luminescence spectrometer (excitation at 546 nm; emis-
sion at 595 nm).
2.11. Cytotoxic studies-MTT assay
MCF 7 cells were maintained in Earle’s minimum
essential medium (Sigma), supplemented with 10% (v/v)
foetal calf serum (Labtech Int.), 2 mM L-glutamine
(Sigma), 1% (v/v) non-essential amino acids (Sigma),
100 IU mLÀ1 penicillin and 100 mg mLÀ1 streptomycin
(Sigma). Exponentially growing cells were plated at
2 Â 104 cells cmÀ2 into 96-well plates and incubated for
24 h before the addition of drugs. Stock solutions of
compounds were initially dissolved in 20% (v/v) DMSO
and further diluted with fresh complete medium.
2.14. Gel mobility assay
The interaction of the drugs with DNA was assessed on
the basis of agarose gel mobility of the supercoiled plasmid
pBR 322 (Biolabs Inc., N 3033 S). The plasmid (1 ng) was
mixed with the polyamines derivatives (10 mM) and incu-
bated at 37 8C for 30 min. Following the addition of the
loading buffer (50% (v/v) glycerol, 0.25% w/v bromophe-
nol blue), the sample was loaded onto 1% (w/v) agarose gel
(made in TAE buffer). The electrophoresis was run in
1Â TAE buffer (40 mM tris buffer (pH 7.6), 20 mM acetic
acid, 1 mM EDTA) on a horizontal gel system (BioRad
Subcell1 GT, USA). This was powered by a BioRad 300
power pack. The electrophoresis was run for 4 h at 40 mA
and was then visualised under UV light.
The growth-inhibitory effects of the compounds were
measured using standard tetrazolium MTT assay [15].
After 24 and 48 h of incubation at 37 8C, the medium
was removed and 200 mL of MTT reagent (1 mg/mL) in
serum free medium was added to each well. The plates
were incubated at 37 8C for 4 h. At the end of the incuba-
tion period, the medium was removed and pure DMSO
(200 mL) was added to each well. The metabolized MTT
product dissolved in DMSO was quantified by reading the
absorbance at 560 nm on a micro plate reader (Dynex
Technologies, USA). IC50 values are defined, as the drug
concentrations required to reduce the absorbance by 50%
of the control values. The IC50 values were calculated from
the equation of the logarithmic line determined by fitting
the best line (Microsoft Excel) to the curve formed from
the data. The IC50 value was obtained from the equation for
y = 50 (50% value).
2.15. Single cell gel electrophoresis assay
For the determination of DNA strand breakage, expo-
nentially growing cells were trypsinised and plated at a
density of 1 Â 106 cells per 24-well plate and allowed to
attach for 24 h before incubating with different concentra-
tions of test compound. Concentrations of BNIPSpd were
incubated for 4 h and BNIPOSpm were incubated for 4 and
12 h.
2.12. Thermal denaturation studies
Optical thermal denaturation experiments were per-
formed in stoppered quartz cuvettes using a Perkin-Elmer
Lambda 2 UV–vis spectrophotometer fitted with a Peltier
circulating heating/cooling temperature programmer and
water supply. Calf thymus DNA (Sigma) working solutions
(200 mM) were prepared by dissolving DNA (5 mg) in
100 mL of 0.01 M phosphate buffer (25 mM Na2HPO4,
15 mM NaH2PO4, and 1.25 mM EDTA). The DNA-drug
solutions were prepared by the addition of the drug (20%
DMSO/water) to give the desired final concentration. The
binding affinity (i.e. Tm) was measured by determining the
After the desired incubation, the medium was aspirated
off and cells were rinsed twice with EBSS (Earle’s
balanced salt solution) before trypsinising. Contents of
each well were collected before centrifuging at
2000 rpm for 5 min. Cells were then resuspended in fresh
full medium before counting the cells using a haemocyt-
ometer. The cells were centrifuged again at 1500 rpm for
3 min. The supernatant was then quickly removed
before adding 1 mL MEM supplemented with 20% FCS.
Two hundred microlitres of this suspension was then
added to 1 mL of cooled PBS and left for 5 min. This

A.-M. Dance et al. / Biochemical Pharmacology 69 (2005) 19–27
23
was then centrifuged at 1500 rpm for 3 min. Cells
3. Results
were resuspended in 1% (w/v) LMP agarose before
being put on a microscope slide previously coated with
1% (w/v) normal melting point agarose. Microscope
slides were placed into cooled slide boxes with lysis
solution (pH 10) and left overnight at 4 8C. Slides were
then placed in an electrophoresis tank covered with elec-
trophoresis solution (NaOH, EDTA, pH 13, 4 8C) for
40 min to allow DNA unwinding before electrophoresis
at 25 V for 30 min at 4 8C. Slides were then placed in
neutralising buffer (pH 7.5, 0.4 M tris) at 4 8C for 5 min.
This was repeated twice more. Slides were dried at 37 8C
before scoring. After drying, 30 mL LMP was added to
each gel and allowed to harden before adding 20 mL DAPI
(1 mg/mL).
Cells were scored for DNA damage visually. One
hundred comets were scored per slide (2 gels per slide)
based on the amount of fluorescence in the nucleoid
(comet) tail and given a value of 0, 1, 2, 3 or 4 (were
undamaged cells are given a score of 0 and maximum
damaged cells a score of 4). The total score therefore
ranged from 0 to 400. This method of classification
has been extensively validated using computerised
image analysis. The single cell gel electrophoresis assay
was performed three times with four replicates per sample
[14].
3.1. Synthesis of bisoxynaphthalimidopolyamines
The synthesis of compounds BNIPOPut, BNIPOSpd
and BNIPOSpm is depicted in Fig. 2. The common
starting material is O-tosylpropyloxynaphthalimide 5
prepared from 3-bromopropan-1-ol 1. To introduce the
aminooxy functionality, 3-bromopropan-1-ol was treated
with N-hydroxyphthalimide in DMF and in the presence
of triethylamine to afford hydroxypropyloxyphthalimide
2. Removal of the phthalimido-protecting group with
hydrazine hydrate at room temperature yielded aminoox-
ypropanol 3. Treatment of 3 with naphthalic anhydride
under basic conditions gave hydroxypropyloxynaphtha-
limide 4, which upon reaction with tosyl chloride in
anhydrous dichloromethane gave 5. The mesitylated
polyamines 6–8 were synthesized according to our pre-
vious report [9,12] and they were treated with 5 in
anhydrous DMF and in the presence of cesium carbonate
to give the fully protected bisnaphthalimidooxy polya-
mines. Subsequent removal of the mesityl groups with
glacial acetic acid/hydrobromic acid afforded BNIPOPut,
BNIPOSpd and BNIPOSpm in 48, 72 and 68% yields,
respectively. BNIPOPut, BNIPOSpd and BNIPOSpm all
showed good solubility (20% DMSO/water) and we were
able to use them in all the biological assays described in
this paper.
2.16. Fluorescence microscopy
For the determination of cell morphology after treat-
ment, exponentially growing cells were trypsinised and
plated at a density of 3.6 Â 104 cells/cm2 onto sterile
coverslips in 60 mm culture dishes. Microscopic analysis
was carried out a minimum of three times with each sample
having four replicates. Cells were allowed to attach for
24 h before incubating with 1 mM of test compound for
different time intervals. After the desired incubation, the
medium was aspirated and cultures were rinsed twice
with PBS. The coverslips were then mounted onto
microscope slides using 90% (v/v) glycerol. To confirm
cellular uptake of the compounds, fluorescence micro-
scopy was repeated with the cells at 4 8C rather than
37 8C. Cells were incubated overnight at 37 8C for attach-
ment onto coverslips. They were then incubated at 4 8C for
2 h before the addition of 1 mM of test compound for 6
and 12 h. Coverslips were then mounted onto slides as
described above.
3.2. Cytotoxic studies
The cytotoxicity of polyamine derivatives was assayed
against MCF 7 cells (a breast cancer cell line). Cells
(2 Â 104) were incubated in the growth media at 37 8C
overnight in the presence of the polyamine derivatives at
varied concentrations (0.1–50 mM) for 24 and 48 h. The
degree of cell viability was analyzed by the MTTassay [15]
(Table 1).
When comparing cytotoxity after 24-h drug exposure
using the concentration response curves obtained from the
MTT assays, similar curves were observed with BNIPSpm
and BNIPOSpd. BNIPSpd demonstrated the sharpest
decrease in absorbance (see Fig. 3). Increasing compound
exposure led to higher cell death except in the case of
BNIPOPut. After 48-h drug exposure, BNIPOSpm and
BNIPOSpd gave similar IC50 of 29.55 Æ 8.01 mM and
27.22 Æ 6.57 mM, respectively.
It is interesting to note that when BNIPOSpm is exposed
to cells for 72 h, an increase in cytotoxicity was observed
with an IC50 value of 10.0 mM. However both BNIPOSpm
and BNIPOSpd were less toxic when compared with
BNIPSpm and BNIPSpd with IC50 of 5.5 Æ 0.5 mM and
0.73 Æ 0.17 mM, respectively. The order of toxicity of
polyamine derivatives on MCF 7 cells after 48-h drug
exposure was BNIPSpd > BNIPSpm > BNIPOSpd > B-
BNIPOSpm > BNIPOPut; therefore, the extent of toxicity
2.17. Statistics
A minimum of three independent repeat experi-
ments was conducted for each set of analyses. All experi-
ments comprised three internal replicates. Data is pre-
sented as means Æ S.D. and where appropriate data
was compared by student’s t-test (Sigma Stat, SPSS
software).

24
A.-M. Dance et al. / Biochemical Pharmacology 69 (2005) 19–27
Fig. 2. Reagents and conditions: (i) N-hydroxyphthalimide in DMF, Et3N, 85 8C, overnight; (ii) H2NNH2 in abs. ethanol, RT, overnight; (iii) naphthalic
anhydride in DMF, DBU, 85 8C, overnight; (iv) tosyl-Cl, CH2Cl2, Et3N, 50 8C, overnight; (v) mesitylated polyamines, DMF, 85 8C, overnight; (vi) HBr/glacial
CH3CO2H in CH2Cl2, RT, 24 h.
depends significantly on the nature of the polyamine chain
linking the two naphthalimido rings.
POSpm and BNIPOPut. Fluorescence of ethidium bromide
is markedly enhanced when bound to DNA [16,17]. C50
values (drug concentrations necessary to reduce the fluor-
escence of initially DNA-bound ethidium by 50%) of each
compound was measured using 1.66 mM calf thymus DNA
and 2 mM ethidium bromide at 25 8C. The order of binding
strength to DNA is BNIPSpd > BNIPSpm > BNI-
BNIPOSpd > BNIPOSpm > BNIPOPut (Table 2).
3.3. DNA-binding properties
The determination of melting point, i.e. midpoint of
thermal denaturation (Tm) of calf thymus DNA duplexes,
measured in presence and absence of above compounds
and the results are shown in Table 2.
3.4. Gel mobility assay
The polyamine derivatives, at a concentration of
10 mM, all increased the Tm of calf thymus DNA duplexes
indicating a DNA–drug interaction.
Intercalation of a drug between the base pair results in
physical unwinding of the DNA which may be detected by
retardation of the electrophoretic migration of supercoiled
DNA [18] Fig. 4.
The fluorescent ethidium bromide displacement assay
(Table 2) was performed to evaluate DNA-binding affinity
of compounds BNIPSpd, BNIPSpm, BNIPOSpd, BNI-
All the compounds except BNIPOPut caused DNA
unwinding. BNIPOSpd was found to intercalate into
DNA more strongly than the other drugs. BNIPSpm,
BNIPSpd and BNIPOSpm exhibit no significant difference
in their binding strength to DNA, agreeing with the Tm
values.
Table 1
Cytotoxicity of polyamines derivatives against MCF 7 cells
Compoundsa
b
IC50 (mM)
24 h
48 h
5.5 Æ 0.5
0.73 Æ 0.17
>50
BNIPSpm
BNIPSpd
BNIPOPut
BNIPOSpm
BNIPOSpd
a
13.32 Æ 0.63
1.50 Æ 0.42
>50
3.5. DNA damage
>50
29.55 Æ 8.01
DNA single strand breakage was measured after expo-
sure to three bisnaphthalimido compounds BNIPSpm,
BNIPSpd and BNIPOSpm. This assay is routinely used
to study the extent DNA damage in drug treated cells.
BNIPSpd and BNIPSpm were chosen since being the
most active compounds (IC50 = 0.73 Æ 0.17 mM and
32.12 Æ 6.14
27.22 Æ 6.57
MCF 7 cells were incubated in the presence or absence of the indicated
compounds.
b
IC50 is defined as the concentration of compound required to reduce
absorbance measured at 560 nm by 50%. Data are means Æ S.D. of three
replicates.

A.-M. Dance et al. / Biochemical Pharmacology 69 (2005) 19–27
25
Fig. 3. Dose–response curves. Absorbance determined by the MTT assay and data obtained after treating MCF 7 cells with varying concentrations of
compounds (0.01–50 mM) for 24 h. Data are means Æ S.D. of six replicates.
5.5 Æ 0.5 mM, respectively) and the results obtained, were
compared with the moderately active BNIPOSpm
(IC50 = 29.55 Æ 8.01 mM). After 4-h exposure to drug
concentrations from 0.01 to 100 mM, BNIPSpm and
BNIPSpd were found (Fig. 5) to cause significant DNA
damage at 0.1 mM, reaching a maximum effect at 100 mM
BNIPOSpm only produced a low, non-dose dependent
level of DNA damage from 20 mM (data not shown).
However, when BNIPOSpm was exposed to cells for
12 h, significant and extensive dose dependent DNA
damage was observed. The extent of DNA damage is
similar to BNIPSpm when the latter was exposed to cells
for 4 h as illustrated in Fig. 5.
was intense. However with BNIPOSpm, no fluorescence
was observed. After 12-h drug exposure, all the three drugs
showed intense fluorescence properties as shown in Fig. 6.
When the above experiments were performed at 4 8C, no
fluorescence was observed in all cases.
4. Discussion
Three bisoxynaphthalimido polyamine derivatives
(BNIPOPut, BNIPOSpd and BNIPOSpm) were synthe-
sized and were found to have good solubility properties.
It would appear that the presence of several heteroatoms in
the linker chain enhances the solubility of bisnaphthali-
mido compounds. However, the presence of oxygen atoms
in the linker chains tends to cause a decrease in cytotoxic
activity.
3.6. Intracellular localisation
MCF 7 cells treated with 1 mM of BNIPSpm, BNIPSpd
and BNIPOSpm were visualized under a fluorescence
microscope after 1, 6 and 24 h. With BNIPSpm and
BNIPSpd, some fluorescence can be observed after 2 h
while no fluorescence could be detected with BNIPOSpm.
After 6 h, the fluorescence with BNIPSpm and BNIPSpd
It is interesting to note that BNIPOSpm and BNIPOSpd
intercalate as effectively to calf thymus DNA as BNIPSpd
Table 2
Effect of drug treatment on the thermal denaturation (Tm) of calf thymus
DNA and ethidium bromide displacement bound to calf thymus DNA (C50
values)
Compoundsa
Tm (8C)b
C50 (mM)c
Calf thymus DNA alone
BNIPSpm 10 mM
BNIPSpd 10 mM
BNIPOPut 10 mM
BNIPOSpm 10 mM
BNIPOSpd 10 mM
a
80.66 Æ 0.53
92.21 Æ 0.56
95.20 Æ 1.59
86.89 Æ 2.45
93.22 Æ 1.84
97.11 Æ 0.39
–
3.93 Æ 0.12
1.21 Æ 0.06
>40
17.33 Æ 3.7
6.03 Æ 0.49
Calf thymus DNA was incubated in the presence or absence of
compounds for 5 min.
b
Fig. 4. Effect of oxa-polyamines on the mobility of the supercoiled pBR
322. All lanes contained 1 mg of pBR 322. Lane M, molecular marker; lane
C, pBR 322 only; lane 1, BNIPSpm (10 mM); lane 2, BNIPSpd (10 mM);
lane 3, BNIPOPut (10 mM); lane 4, BNIPOSpm (10 mM); lane 5, BNI-
POSpd (10 mM).
Tm is temperature (8C) when half of the DNA is denaturated.
C50 is drug concentrations required to reduce the fluorescence of
c
initially DNA-bound ethidium by 50%. Data are means Æ S.D. of three
replicates.

26
A.-M. Dance et al. / Biochemical Pharmacology 69 (2005) 19–27
drug exposure, BNIPOSpm inflicted more DNA damage
when compared with BNIPSpd and yet the IC50 value of
BNIPOSpm is >50 mM after 24-h drug exposure. This
may suggest the participation of many other intricate
factors that are involved in inducing DNA damage and
cytotoxicity.
The subtle difference between the structures of
BNIPSpm and BNIPOSpm (Fig. 1) revealed large differ-
ences in their cytotoxicity against MCF 7 cells. Because all
the compounds discussed in this paper are polycations, we
reason that like other polyamine analogues [19,20], these
compounds could be transported through the cell by an
active transport system, e.g. the polyamine transporter. It is
worth mentioning that little is known about the character-
ization of the polyamine transporter [21]. This theory is
further supported by the apparent absence of drug uptake as
indicated by the lack of fluorescence at low temperature of
4 8C (data not shown).
Fig. 5. DNA damage (strand breakage) determined by Single Cell Gel
Electrophoresis. Data obtained after treating MCF 7 cells with BNIPSpm,
BNIPSpd for 4 h and BNIPOSpm for 12 h at 5 mM drug concentration.
H2O2 (200 mM) was used as a positive control. Data are means Æ S.D. of
three replicates, *P < 0.001 vs. negative control.
as demonstrated by the Tm value and the gel mobility assay.
However, BNIPSpd being the most active compound
showed the highest efficiency in displacing intercalating
agent, ethidium bromide from DNA and also caused
maximum DNA damage at low drug concentration. The
low toxicity of BNIPOSpm and BNIPOSpd may be
explained by their less effective uptake compared to
BNIPSpm and BNIPSpd. This is supported that by the
microscopic observations of no fluorescence in BNI-
POSpm treated cells for 6 h (Fig. 6). In contrast, BNIPSpm
and BNIPSpd showed high fluorescence and toxicity.
It is important to note that the cytoxicities of BNIPSpm,
BNIPSpd and BNIPOSpm do not correlate directly with
the data obtained on DNA damage. For example after 12-h
We conclude that the introduction of oxygen atom in the
a-position of the naphthalimido ring like in the case of
BNIPOPut, BNIPOSpd and BNIPOSpm, may have
resulted in the compounds not being recognized by the
polyamine transporter. Therefore the bisoxynaphthalimi-
dopolyamines BNIPOPut, BNIPOSpd and BNIPOSpm
entered the cells perhaps by some other routes, e.g. very
slow passive diffusion resulting in their delayed and
reduced cytotoxicities. We are currently investigating
the mechanism of how these compounds are being trans-
ported in cancer cells. Furthermore, while the ability of
these compounds to bind to DNA may play a role in their
activity, there are also other factors contributing to their
biological properties.
Fig. 6. Fluorescence microscopy images of MCF 7 cells treated with 1 mM BNIPSpd, BNIPSpm and BNIPOSpm for 6 and 12 h. Cells (1 Â 106) were grown on
sterile coverslips (22 mm  22 mm) and allowed to adhere for 48 h. Medium was removed and fresh medium added with the drug to a final concentration of
1 mM and incubated for the desired time point then washed twice with PBS and put on slide using 90% glycerol as mountant. Images are representative of eight
repeat experiments each comprising three internal replicates.

A.-M. Dance et al. / Biochemical Pharmacology 69 (2005) 19–27
27
[9] Kong Thoo Lin P, Pavlov V. The synthesis and in vitro cytotoxic
studies of novel bis-naphthalimidopropyl polyamine derivatives.
Bioorg Med Chem Lett 2000;10:1609–12.
Acknowledgements
The authors wish to thank The Robert Gordon Univer-
sity, The Scottish Executive Environment and Rural
Affairs Department (SEERAD) and the Royal Society of
Chemistry for financial support and The Centre of Mass
Spectrometry at the University of Wales, Swansea, for
mass spectroscopic analyses.
[10] Pavlov V, Kong Thoo Lin P, Rodilla V. Cytotoxicity, DNA binding and
localization of bis-naphthalimidopropyl polyamine derivatives. Chem
Biol Interact 2001;137:15–24.
[11] Pavlov V, Kong Thoo Lin P, Rodilla V. Morphological changes in
MCF-7 human breast cancer cell lines in response to bis-naphthali-
midopropylspermidine-treatment. Biotechnol Biotechnol Equip
2002;16:118–23.
[12] Kong Thoo Lin P, Dance AM, Bestwick C, Milne L. The biological
activities of new polyamine derivatives as potential therapeutic agents.
Biochem Soc Trans 2003;31:407–10.
References
[13] Kong Thoo Lin P, Wardell S. Synthesis and structure of 1,12-bis-(N-
naphthalamido)-4,9-bis-(2,4,6-trimethylbenzenesulfonyl)-4,9-diaza-
dodecane. J Crystallogr Spectrosc 1998;28:683–7.
[1] Reddy BS, Sondhi SM, Lown JW. Synthetic DNA minor groove-
binding drugs. Pharmacol Ther 1999;84:1–111.
[14] Duthie SJ, Ma A, Ross MA, Collins AR. Antioxidant supplementation
decreases oxidative damage in human lymphocytes. Cancer Res
1996;56:1291–5.
[2] Neidle S, Pearl LH, Skelly JV. DNA structure and perturbation by drug
binding. Biochem J 1987;2143:1–13.
[3] Brana MF, Cacho M, Gradillas B, De Pascual-Teresa B, Ramos A.
Intercalators as anticancer drugs. Curr Pharm Des 2001;7:1745–80.
[4] Rye HS, Yue S, Wemmer DE, Quesada MA, Haugland RP, Mathies
RA, et al. Stable fluorescent complexes of double-stranded DNA with
bis-intercalating asymmetric cyanine dyes: properties and applica-
tions. Nucleic Acids Res 1992;20:2803–12.
[15] Mossman T. Rapid colorimetric assay for cellular growth and survival:
application to proliferation and cytotoxicity assays. J Immunol Meth-
ods 1983;65:55–63.
[16] Le Pecq JB, Paoletti C. A fluorescent complex between ethidium
bromide and nucleic acids. J Mol Biol 1967;27:87–106.
[17] McConnaughie AW, Jenkins TC. Novel acridine-triazenes as proto-
type combilexins: synthesis, DNA binding, and biological activity. J
Med Chem 1995;38:3488–501.
[5] Brana MF, Ramos A. Naphthalimides as anti-cancer agents: synthesis
and biological activity. Curr Med Chem Anti-Canc Agents 2001;1:
237–55.
[18] Lerman LS. Structural considerations in the interactions of DNA and
acridine. J Mol Biol 1961;3:18–30.
[6] Bousquet PF, Brana MF, Conlon D, Fitzgerald KM, Perron D,
Cocchiaro C, et al. Preclinical evaluation of LU 7955: a novel bis-
naphthalimide with potent antitumor activity. Cancer Res 1995;55:
1176–80.
[19] Burns MR, Carlson CL, Vanderwerf SM, Ziemer JR, Weeks RS, Cai F,
et al. Amino acid/spermine conjugates: polyamine amides as potent
spermidine uptake inhibitors. J Med Chem 2001;44:3632–44.
[20] Huber M, Pelletier JG, Torossian K, Dionne, Gamache PI, Charest-
Gaudreault R, et al. 2,2-Dithiobis(N-ethyl-spermine-5-carboxamide)
is a high affinity, membrane-impermeant antagonist of the mammalian
polyamine transport system. J Biol Chem 1996;271:27556–63.
[21] Wang C, Delcross JG, Cannon L, Konate F, Carias H, Biggerstaff J, et
al. Defining the molecular requirements for the selective delivery of
polyamine conjugates into cells containing active polyamine trans-
porters. J Med Chem 2003;46:5129–30.
[7] Gallego J, Reid BR. Solution structure and dynamics of a complex
between DNA and the antitumor bisnaphthalimide LU-79553: inter-
calated ring flipping on the millisecond time scale. Biochemistry
1999;38:15104–15.
[8] Brana MF, Castello JM, Moran M, Perez de Vega MJ, Qian X-D,
Romerdahl CR, et al. Bis-naphthalimides 2. Synthesis and biological
activity of 5,6-acenaphthalimidoalkyl-18-naphthalimidoalkyl amines.
Eur J Med Chem 1995;30:235–9.