Synthesis, characterization and multi-spectroscopic DNA interaction studies of a new platinum complex containing the drug metformin

Article abstract of DOI:10.1016/j.saa.2014.02.167

A new platinum(II) complex; [Pt(Met)(DMSO)Cl]Cl in which Met = metformin and DMSO: dimethylsulfoxide, was synthesized and characterized by ¹H NMR, IR, UV-Vis spectra, molar conductivity and computational methods. Binding interaction of this complex with calf thymus (CT) DNA has been investigated by using absorption, emission, circular dichroism, viscosity measurements, differential pulse voltammetry and cleavage studies by agarose gel electrophoresis. UV-Vis absorption studies showed hyperchromism. CD studies showed less perturbation on the base stacking and helicity bands in the CD spectrum of CT-DNA (B → C structural transition). In fluorimeteric studies, the Pt(II) complex can bind with DNA-NR complex and forms a new non-fluorescence adduct. The anodic peak current in the differential pulse voltammogram of the Pt(II) complex decreased gradually with the addition of DNA. Cleavage experiments showed that the Pt(II) complex does not induce any cleavage under the experimental setup. Finally all results indicated that Pt(II) complex interact with DNA via groove binding mode.

Full text of DOI:10.1016/j.saa.2014.02.167

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 128 (2014) 377–385
Contents lists available at ScienceDirect
Spectrochimica Acta Part A: Molecular and
Biomolecular Spectroscopy
journal homepage: www.elsevier.com/locate/saa
Synthesis, characterization and multi-spectroscopic DNA interaction
studies of a new platinum complex containing the drug metformin
⇑
Nahid Shahabadi , Leila Heidari
Department of Inorganic Chemistry, Faculty of Chemistry, Razi University, Kermanshah, Iran
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
ꢀ
ꢀ
ꢀ
A new platinum(II) complex
containing the drug metformin was
synthesized.
This study is expected to provide
greater insight into the use of other
drugs like cis-platin as anticancer.
All the results showed that the
complex was a DNA groove binder.
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 2 January 2014
Received in revised form 16 February 2014
Accepted 21 February 2014
Available online 12 March 2014
A new platinum(II) complex; [Pt(Met)(DMSO)Cl]Cl in which Met = metformin and DMSO: dimethylsulf-
oxide, was synthesized and characterized by H NMR, IR, UV–Vis spectra, molar conductivity and compu-
1
tational methods. Binding interaction of this complex with calf thymus (CT) DNA has been investigated
by using absorption, emission, circular dichroism, viscosity measurements, differential pulse voltamme-
try and cleavage studies by agarose gel electrophoresis. UV–Vis absorption studies showed hyperchro-
mism. CD studies showed less perturbation on the base stacking and helicity bands in the CD
spectrum of CT-DNA (B ? C structural transition). In fluorimeteric studies, the Pt(II) complex can bind
with DNA–NR complex and forms a new non-fluorescence adduct. The anodic peak current in the differ-
ential pulse voltammogram of the Pt(II) complex decreased gradually with the addition of DNA. Cleavage
experiments showed that the Pt(II) complex does not induce any cleavage under the experimental setup.
Finally all results indicated that Pt(II) complex interact with DNA via groove binding mode.
Ó 2014 Elsevier B.V. All rights reserved.
Keywords:
DNA
Metformin
Platinum complex
Spectroscopic methods
Groove binding mode
Introduction
residues as protein constituents, ATP and nucleic acids. Moreover,
the ligands can interact with biomolecules through coordinative or
The involvement of some metal ions in regulation of physiolog-
ical processes has stimulated the development of metal-based
therapeutics. The pharmaceutical use of complexes arises from
the fact that the positively charged metal centers are favored to
bind to negatively charged biomolecules such as the amino acid
hydrogen bonds as well as dipole dipole interactions [1]. Therefore,
the use of the metal complexes as therapeutics has become
increasingly important over the last years resulting in a variety
of interesting drugs used in fields as cancer, arthritis, ulcer, diabe-
tes, anemia and cardiovascular medicine [2].
Platinum-based drugs play an important role in clinical cancer
chemotherapy. The first platinum antitumor drug introduced in
the clinic was cis-diammine dichloro platinum(II). However, in
⇑
Corresponding author. Tel./fax: +98 831 8360795.
E-mail address: nahidshahabadi@yahoo.com (N. Shahabadi).
http://dx.doi.org/10.1016/j.saa.2014.02.167
386-1425/Ó 2014 Elsevier B.V. All rights reserved.
1

3
78
N. Shahabadi, L. Heidari / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 128 (2014) 377–385
spite of their high activity, the applications of cis-platin and similar
compounds are limited by the side effects. Therefore, the search for
soluble and less toxic analogues of cis-platin is one of the main
goals in the synthesis of new platinum(II) complexes [3–5].
Recently, new active monofunctional platinum complexes with
structural features that violate the ‘classical’ structure–activity
relationships have been described [6–9].
Transition metal complexes can bind to DNA via covalent inter-
actions, where a labile ligand of the complexes is replaced by a
nitrogen base of DNA such as guanine N7, and/or non-covalent
interactions which include intercalative, electrostatic and groove
50 mM of the Tris–HCl buffer at pH 7.4 dialyzing exhaustively
against the same buffer for 24 h. A solution of calf thymus DNA
gave a ratio of UV absorbance at 260 and 280 nm more than 1.8,
indicating that DNA was sufficiently free from protein. The stock
solutions were stored at 4 °C and used over no more than 4 days.
Synthesis of platinum complex
2 2
Synthesis of cis-[Pt(DMSO) Cl ] complex. The complex was prepared
according to the following procedure and operated in dark. K
2
[
4
PtCl ] (0.415 g; 0.001 mol) was mixed with water 4 mL. DMSO
(
0.242 g; 0.003 mol) was then added to the red aqueous solution
(
surface) binding of metal complexes along outside of DNA helix,
of tetrachloroplatinate(II). The resulting mixture was stirred for
2 h, at 20 °C. The yellow solid precipitate was filtered off, washed
towice with water, ethanol and diethylether, and then air-dried at
0 °C to afford yellow crystals (Yield:84%).
along major or minor groove [10]. These interactions have been
shown to disrupt replication and/or transcription culminating in
cellular death.
1
2
It is well known that biguanide derivatives exhibit both biolog-
ical and coordinative properties [11]. Interest in biguanide deriva-
tives arises from their well acclaimed medicinal values as
germicidal, bacteriostatic, hypoglycemic and anticarcinogenic
agents [12]. Also, biguanide and biguanide derivatives are extre-
mely powerful and formidable coordinating ligands which have
played important roles in elucidating many interesting aspects of
coordination chemistry [13,14].
Synthesis of [Pt(Met)(DMSO)Cl]Cl. A suspension of cis-[PtCl
2
(-
DMSO) ] (0.222 g; 0.53 mmol) in MeOH (15 mL) was treated with
2
the stoichiometric amount of metformin (0.09 g; 0.53 mmol) dis-
solved in 20 mL MeOH–Water (1:1) solution, stirred at room tem-
perature for 1 day and filtered. The filterate was left for
crystallization and after 3 days white crystals of the complex pre-
cipitated (Yield: 93.6%).
Among biguanide drugs, the most widely prescribed type II dia-
betes medications is N,N-dimethylbiguanide known as Metformin
[
15]. Beside decreasing the glucose level, the N,N-dimethylbigua-
Instrumentation
nide also acts as analgesic, antimalarial and antimetabolite for
organisms that inhibit the metabolism of folic acid [16,17]. Re-
cently, it was found that platinum(IV) complex with N,N-dimethyl-
biguanide shows antitumor activity [18]. Attempting to modulate
the biological activity of this derivative it was demonstrated that
some of their complexes possess antimicrobial activity and also
have an interesting thermal behavior [19]. In addition, one ap-
proach to accelerate the availability of new drugs is to reposition
drugs approved for other indications as anticancer agents. In our
laboratory we investigated the interaction of known drugs with
protein and DNA and the effects of metals on their interactions
in order to study the anticancer ability of them and highlighted
the role of metal centers on bioinorganic medicinal chemistry re-
searches. Interestingly, the mechanism of action and toxic side ef-
fects of purely organic drugs can be modulated in the presence of
metals.
In this paper, we report the preparation, characterization and
DNA binding studies of platinum(II) complex with metformin
hydrochloride (Met) as ligand. The binding properties on Pt(II)
complex to calf thymus DNA in physiological buffer (pH 7.4) was
investigated by multi-spectroscopic methods. The results showed
that spectroscopy techniques could provide a convenient way to
characterize both the binding mode and the interaction mecha-
nism of Pt(II) complex binding to DNA. We believe that the knowl-
edge gained from this study will be helpful to further understand
the binding mechanisms and can provide much fruitful informa-
tion for designing a new type of highly effective anti-cancer drugs.
The complex obtained was characterized by UV–Vis, FT-IR and
H NMR spectroscopy. The NMR spectra were recorded with a Bru-
1
ker Avance DPX 200 MHz (4.7 T-esla) spectrometer using d
6
-DMSO
solvent. IR spectra were obtained with an ABB BOMEM MB 104 FT-
ꢁ1
IR spectrometer using KBr discs for 400–4000 cm range. The ele-
mental analysis was performed using a Heraeus CHN elemental
analyzer.
Absorbance spectra were recorded using an HP spectrophotom-
eter (Agilent 8453) equipped with a thermostated bath (Huber
polysat cc1). Absorbance experiments for the DNA interaction of
Pt(II) complex were carried out by keeping the concentration of
ꢁ5
the complex constant (7.5 ꢂ 10 M) while varying DNA concen-
ꢁ
4
tration from 0 to 1.05 ꢂ 10 M (r
i
= [DNA]/[complex] = 0, 0.2, 0.6,
0.8 and 1.4). Absorbance values were recorded after each succes-
sive addition of DNA solution, followed by an incubation period
(2 h).
The UV–Vis absorption spectra of DNA–NR complex and the
mixture of different concentrations of Pt(II) complex and DNA–
NR complex were measured under the pH 7.4, Tris–HCl buffer,
respectively. The sample tube containing 2.0 mL mixture solution
ꢁ
5
ꢁ1
ꢁ5
ꢁ1
of 1.0 ꢂ 10 mol L
NR and 6 ꢂ 10 mol L
CT-DNA was al-
lowed to stand for 5 min at room temperature. The mixture solu-
ꢁ3
ꢁ1
tion was titrated by successive additions of 1 ꢂ 10 mol L
stock solutions of Pt(II) complex. For every addition, the mixture
solution was shaken and allowed to stand for 5 min at room
temperature.
For viscosity measurements, a viscometer (SCHOT AVS 450) was
used, kept at 25 °C by a constant temperature bath. Flow time was
measured with a digital stopwatch; the mean values of two repli-
Experimental
cated measurements were used to evaluate the viscosity (
samples. The data were reported as ( ) versus the [complex]/
DNA] ratio, where is the viscosity of the DNA solution alone.
g) of the
Materials and methods
g/g
0
[
g
0
Commercial pure chemicals such as Metformin, Dipotassium
tetrachloroplatinum(II), Methanol and DMSO were purchased from
Merck and Tris–HCl, calf thymus DNA (CT-DNA), pUC18 and Neu-
tral red (NR) were purchased from Sigma Co. Solutions were pre-
pared with double distilled water.
Viscosity values were calculated from the observed flow time of
CT-DNA containing solution corrected from the flow time of buffer
alone (t
0
) as
g
= (t ꢁ t
0 0
)/t .
CD measurements were recorded on a JASCO (J-810) spectropo-
larimeter by keeping the concentration of DNA constant
ꢁ5
Experiments were carried out in Tris–HCl buffer at pH 7.0. The
stock solution of CT-DNA was prepared by dissolving of DNA in
(8 ꢂ 10 M) while varying the platinum complex concentration
(r = [complex]/[DNA] = 0.0, 0.2 and 0.6).
i

N. Shahabadi, L. Heidari / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 128 (2014) 377–385
379
Fluorescence measurements were carried out with a JASCO
Table 1
Elemental analysis data for platinum(II) complex.
spectrofluorimeter (FP 6200). The competitive interaction between
the neutral red and platinum complex with DNA was carried out as
follows: fixed amounts of the neutral red and DNA were titrated by
successive additions of 1 ꢂ 10 mol L stock solutions of Pt(II)
complex. These samples were excited at 530 nm and emission
was observed between 530 and 700 nm.
[Pt(Met)(DMSO)Cl]Cl
Theoretical calculation
Found
%
%N
%H
C
15.22
14.79
3.59
15.38
14.91
3.64
ꢁ3
ꢁ1
The differential pulse voltammetry (DPV) measurements were
performed using an AUTOLAB model (PG STAT C), with a three-
electrode system: a 0.10-cm-diameter Glassy carbon (GC) disc as
working electrode, an Ag/AgCl electrode as reference electrode,
and a Pt wire as counter electrode. Electrochemical experiments
were carried out in a 25-mL voltammetric cell at room tempera-
ture. All potentials are referred to the Ag/AgCl reference. Their sur-
faces were freshly polished with 0.05 mm alumina prior to each
experiment and were rinsed using double distilled water between
each polishing step. The supporting electrolyte was 0.05 M of Tris–
HCl buffer solution (pH 7.4) which was prepared with double dis-
tilled water. The current–potential curves and experimental data
were recorded on software GPES.
tentatively attributed to the formation of the chelate ring on com-
plexation or to the activation of some IR inactive band as a result of
complexation [20]. The sharp absorption band at 1102 cm indi-
ꢁ1
cates S-bonded DMSO, as reported for this soft coordination center
[
21].
1H NMR spectra of Met and Pt(II) complex were recorded at
room temperature in d
shown in Table 3.
6
-DMSO as the solvent and the data are
The 1H NMR spectra of all N,N-dimethylbiguanidium deriva-
tives exhibit a pattern correlated with the isolated methyl protons
in the region 2.9–3.1 ppm. Additional resonance arise from the
amine and imines protons at d = 6.8 and d = 7.22 ppm, respectively.
For the gel electrophoresis experiments, pUC18 plasmid DNA
was treated with Pt complex in TAE buffer (pH = 8), and the con-
tents were incubated for 2 h at 37 °C. The samples (r = [Pt com-
i
1
The H NMR spectrum of the complex exhibits a pattern assignable
to methyl protons at 3.03 ppm. The shift of the broad peak of ligand
assigned to imine protons from 7.2 ppm to downfield (7.53 and
plex]/[DNA] = 0.0, 0.1, 0.3, 0.5, 0.7 and 0.9) were eletrophoresed
for 1 h at 80 V on 0.8% agarose gel in TAE buffer. Gel was stained
with ethidium bromide and photographed using UV illumination.
For dialysis we used dialysis bag for each solutions of cis-Pt-
DNA, Pt(II) com-DNA and Hoechst-DNA and then absorbance and
emission values were recorded after 24 h for inner solutions dialy-
sis bag.
9
N
.76 ppm) in the spectrum indicates their coordination through
and N atoms. The peak of imine proton at 9.76 ppm assigned
2
4
to a hydrogen bond with oxygen atom of DMSO group bonded to
platinum. The peak at 2.54 assigned to methyl protons of coordi-
nated DMSO in complex.
Electronic spectra of the complex and the metformin ligand
ꢁ4
were recorded in Tris–HCl buffer with 10 M/L. Met exhibits an
intense absorption band around 233 nm which is assigned to
ꢃ
p
?
p
transition of drug while in the electronic absorption spec-
Results and discussion
trum of platinum complex this band shifted to 228 nm. This band
shows a blue shift and decreasing for the absorbance intensity in
the complex (Fig. S1). Cis-[Pt(Cl) (DMSO) ] exhibit an absorption
2 2
band around 215 nm that absorption band of the platinum com-
plex shifted to higher wavelength as compared with cis-platin
Synthesis and characterization of platinum(II) complex
The complex [Pt(Met)(DMSO)Cl]Cl (Scheme 1) was prepared by
2 2
the reaction of cis-[PtCl (DMSO) ] and Met in aqueous methanol.
(
Fig. S2). New absorption band appeared around 316 nm can be
The computational methods, molar conductivity, elemental analy-
attributed to the ligand-to-metal charge transfer bands from the
electronic lone pairs of respective nitrogen to the metal ion [22].
The conductivity of metformin complex was measured in DMF
solvent at room temperature. The complex showed the molar con-
1
sis, H NMR, FT-IR, UV–Vis spectra assignments confirm the prep-
aration of the complex.
The data of elemental analysis are shown in Table 1.
In order to obtain information about the coordination mode of
the ligand, FT-IR spectra of the complex and the free ligand were
recorded and the data are shown in Table 2. In the complex spec-
ꢁ3
ꢁ1
2
ductance value for 10 mol/L concentration equal to 65
X
cm -
ꢁ1
mol . According to the literature data the complex is a 1:1
electrolyte as shown by its molar conductivity (
in DMF [23].
M
K ) measurements
ꢁ
1
trum a band at 1682 cm , assigned to m(C@N), shifted to higher
wave numbers as compare with ligand. These modifications indi-
The presence of chloride ion in metformin complex was con-
firmed by the addition an equimolar amount of silver nitrate to
cate the involvement of the N2 and N4 atoms in coordination
[
20]. It is interesting to note that a band of medium intensity ap-
ꢁ5
the complex (both 10 M) and leading to the appearance of white
ꢁ
1
pears at 1300 cm in the platinum complex. This band is not pres-
ent in the spectrum of the free metfromin. Hence, it could be
precipitate.
Computational method
The calculations on the structure of the complex [C
OPtS]Cl (1) were performed by the appropriate theoretical method.
The structure of [C 17ClN OPtS]Cl(1) was obtained by PM6 meth-
od [24]. The point group of [C 17ClN OPtS]Cl(1) is C . The QM/
6
H17ClN5-
6
H
5
6
H
5
1
PM6 method was utilized, because of platinum at the structure
of 1. Platinum was defined and identified by the Hamiltonians of
PM6. The calculations have performed by Spartan ‘10 package
[
24]. The results of the calculations were demonstrated that the
complex [C 17ClN OPtS]Cl (1) has a planar structure around the
‘Pt’’ (Fig. 1a). The DMSO molecule was oriented near to the
6
H
5
‘
N1AH17. An electrostatic attraction (hydrogen bond) was identi-
fied between oxygen in DMSO (O1) and H17 of the imine group
(O1AH17). The heat of formation and the dipole moment of this
Scheme 1. The structure of [Pt(Met)(DMSO)Cl]Cl.

3
80
N. Shahabadi, L. Heidari / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 128 (2014) 377–385
Table 2
FT-IR data for free ligand and its Pt complex (in cmꢁ1).
Compound
mas (NH
2
)
m
(NH)
m
s
(NH
2
)
t
(C@N)
m
(chelate ring)
t
(S-bonded DMSO)
Metformin
Pt Complex
3372
3433
3299
3310
3162
3197
1626
1682
–
1300
–
1102
Table 3
Selected 1H NMR spectral data (in ppm).
respectively. The signal of 1623 was related to the imine C3@N2
and C4@N1. The PM6 calculated signals at 322 and 792 cm have
ꢁ1
Compound
Metformin
NH
6.8
2
NH (amine) NH (imine) CH
3
(Met) CH
3
(DMSO)
determined for the stretching vibrations of Pt
respectively.
1 1 1 1
ACl and Pt AS ,
6.8
7.2
2.9
Pt Complex 7.28
6.75
7.53, 9.76
0.33, 2.56
3.03
0.13
2.54
D
0.48 ꢁ0.25
DNA binding studies
Dialysis
ꢁ1
configuration were ꢁ276.309 kJ mol
and 2.18 debye, respec-
Transition metal complexes can bind to DNA via covalent inter-
actions (irreversible binding), and/or non-covalent interactions
(reversible binding) which include intercalative, electrostatic and
groove (surface) binding of metal complexes along outside of
DNA helix, along major or minor groove. So, at first we determined
nature of the binding between Pt(II) complex and DNA. Competi-
tion dialysis is a powerful new tool for the discovery of ligands that
tively. The calculated IR spectrum for the complex [C
6
H17ClN5-
OPtS]Cl(1) was shown in Fig. 1b. The calculated IR data have
shown very good agreement with the experimental results for
the identified configuration. Study on the calculated vibration fre-
ꢁ1
quencies by PM6 method has shown that the signal of 2759 cm
ꢁ1
and 2683 cm were related to N2AH19 and N1AH17 stretching,
Fig. 1. (a) The most stable configuration of platinum complex. (b) The calculated IR spectrum for platinum complex.

N. Shahabadi, L. Heidari / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 128 (2014) 377–385
381
bind to nucleic acids. The method is based on firm thermodynamic
principles and is simple to implement [25]. In this study, we used
dialysis to discover the binding mode between the complex and
DNA. If the nature of binding between DNA and platinum complex
be covalent we expect that the absorbance or emission of the DNA–
Pt complex does not show a lot of change before and after dialysis
because the nature of the binding is irreversible and if the nature of
the binding be non-covalent, the absorbance or emission of the
DNA–Pt complex changes before and after dialysis according to
separating Pt complex from DNA because the nature of the binding
is reversible. For this aim, we studied absorbance/emission
helix ct-DNA [28]. Also a coordination bond of DNA base with plat-
inum can occur through replacement of the chloride ligand in the
complex, probably through aquation as has been observed with
cis-platin [29]. No significant shift in the UV absorption band
was observed, which implies that there is no change of the coordi-
nation environment of the metal ion [30], covalent/coordinate link-
age to the N(7) of nucleobase is strongly ruled out.
b
The intrinsic binding constant, K , of Pt(II)complex with calf
thymus DNA was calculated from Wolfe–Shimmer equation:
½
DNAꢄ
½DNAꢄ
1
¼
þ K
À
Á
e
a
ꢁ
e
f
e
b
ꢁ
e
f
eb
ꢁ
ef
b
changes of cis-platin (cis-Pt(NH
groove binder) and metformin platinum complex in this way.
Fig. S3 shows the absorbance of the cis-Pt(NH Cl –DNA complex
3 2 2
) Cl ) (covalent binder), Hoechst
(
where [DNA] is the concentration of DNA in base pairs, the apparent
3
)
2
2
absorption coefficient
a f b
e , e and e correspond to Aobsd/[Complex],
does not change before and after dialysis. Fig. S4 shows that the
emission of the Hoechst-DNA complex has a change after the dial-
ysis according to separating of Hoechst from DNA, because emis-
sion of that after dialysis decreases, this shows that a lot of the
Hoechst were separated from DNA. For the [Pt(Met)(DMSO)Cl]Cl
complex (Fig. S5) absorbance of the platinum complex–DNA before
and after dialysis show changes according to separating high
amount of the platinum complex from DNA. So, the nature of the
binding between the platinum complex and DNA is reversible
and belongs to non-covalent interaction.
the extinction coefficient for the free Pt(II) complex, for each addi-
tion of DNA to the Pt(II) complex and Pt(II) complex in the fully
bound form, respectively. In plot of [DNA]/(
e
a
ꢁ
f
e ) versus [DNA],
4
ꢁ1
K
b
(1.5 ꢂ 10 M ) is given by the ratio of slope to the intercept.
The observed binding constant is more in keeping with groove bind-
ing, as observed in the literature [31–34]. Our results illustrate that
platinum complex may bind to DNA via groove binding.
Effect of ionic strength on the spectrum of Pt(II) complex–DNA
Monitoring the change of ionic strength is an efficient method
to recognize the binding modes between molecules and DNA.
Increasing the concentration of cation will increase the complexa-
tion probability between the cation and DNA phosphate backbone.
Due to a competition for phosphate anion, the addition of the cat-
ion will weaken the surface-binding interactions which include
electrostatic interaction and hydrogen binding between DNA and
molecules [35]. In order to prove whether Pt(II) complex provide
an electrostatic binding or other kinds of binding with DNA, Pt(II)
complex solutions was titrated with NaCl in the absence and pres-
UV–Vis absorption studies
Monitoring the effect of adding increasing amounts of DNA on
the absorption spectrum of a metal complex is one of the most
widely used methods for determining overall binding constants.
‘
‘Hyperchromic’’ and ‘‘hypochromic’’ effects are the spectra fea-
tures of DNA with regard to its double helical structure.
As shown in Fig. 2, upon adding CT DNA to [Pt(Met)(DMSO)Cl]Cl
complex in Tris HCl buffer solution, the absorption band exhibits
hyperchromism without shifts in band position.
ꢁ5
ence of DNA (3 ꢂ 10 M). Pt(II) complex was titrated with NaCl
ꢁ
7
The Pt(II) complex can bind to the double-stranded DNA in dif-
ferent binding modes on the basis of their structure, charge and
type of ligand. As DNA double helix possesses many hydrogen
bonding sites which are accessible both in the minor and major
grooves, it is likely that the NH groups of complex forms hydrogen
bonds with N of adenine or O of thymine in the DNA, which may
contribute to the hyperchromism observed in the absorption spec-
tra. On the other hand, our Pt(II) complex which possesses methyl
groups of the DMSO, can bind to DNA by van der Waals interaction
between the methyl groups and the thymine methyl group [26].
The hyperchromic effect may also be due to the electrostatic inter-
action [27] between positively charged Pt(II)complex and the neg-
atively charged phosphate backbone at the periphery of the double
with the increasing concentration of the salt from 5 ꢂ 10 up to
ꢁ4
ꢁ1
1
.9 ꢂ 10
M . As it is elucidated from Fig. 3, the slopes of the
curves in the absence and presence of DNA are the same. It is de-
duced that NaCl has no effect on DNA binding with Pt(II) complex.
Therefore, the results indicated that Pt(II) complex does not give an
electrostatic or outside binding with DNA.
Effect of Pt(II) complex on the absorption spectrum of NR–DNA system
It has been shown that an absorption spectrum in the visible re-
gion of the NR solution is characterized by a band at 464 nm. This
absorption gradually decreased with the increasing concentration
of DNA, and a new band at 525 nm developed. An isosbestic point
at 497 nm provided evidence of the new NR–DNA complex forma-
tion (Fig. 4). In general, hypochromism and red shift are commonly
associated with molecular intercalation into the base stack of the
DNA [36]. These two observed spectral effects were ascribed to a
strong interaction between the electronic state of the intercalating
ꢁ
5
Fig. 2. Absorption spectra of [Pt(Met)(DMSO)Cl]Cl complex (7.5 ꢂ 10 M) in the
absence and presence of increasing amounts of CT-DNA: (r = 0.0, 0.2, 0.6, 0.8 and
.4).
i
Fig. 3. Effect of ionic strength on the absorbance spectrum of the Pt(II) complex in
the absence (j) and presence (Ç) of the DNA.
1

3
82
N. Shahabadi, L. Heidari / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 128 (2014) 377–385
Fig. 4. Absorption spectra of NR in the presence of DNA at different concentrations.
Conditions: [NR]: 3 ꢂ 10ꢁ mol L , [DNA] (ꢂ10 mol L ): 0.0, 1.5 and 3.6.
5
ꢁ1
ꢁ5
ꢁ1
Fig. 6. The changes of CD spectrum of 8 ꢂ 10ꢁ M DNA in the absence and presence
5
chromophore and that of the DNA bases. The strength of this elec-
tronic interaction is expected to decrease as the cube of the dis-
tance between the chromophore and the DNA bases [36]. The
observed large spectral hypochromism during the interaction of
the NR dye with the DNA strongly proposed that the distance be-
tween the intercalated NR dye and the DNA bases was small. The
effects of Pt(II) complex on the absorption spectrum of DNA–NR
system were measured. As shown in Fig. 5, the absorption of
DNA–NR complex decreases with the increasing concentration of
Pt(II) complex. This indicates that the complex ion cannot ex-
change with the intercalated NR, because the spectrum of the free
NR ligand is not restored.
of addition of Pt(II) complex.
including the conversion from a more B-like to a more C-like struc-
ture within the DNA molecule. These changes are indicative of a
non-intercalative mode of binding of the complex and offer sup-
port to the groove binding nature [39].
Fluorescence studies
No luminescence is observed for the complex; it is hard to mon-
itor the interaction of this complex with DNA by employing direct
fluorescence emission methods but actually possible by using a
fluorescence assay of the organic molecule probe. We used neutral
red probe for the interaction of platinum complex with DNA.
On addition of NR to DNA solution, it is found that the fluores-
cence intensity of the system is greatly enhanced. Thus, NR is a
sensitive probe to study the interaction between DNA and drug
molecules. When Pt(II) complex is added to the solution of DNA–
NR complex, the fluorescence intensity of DNA–NR complex de-
crease with the increasing concentration of Pt(II) complex. The
fluorescence spectra are shown in Fig. 7. There are three possible
reasons considering the decrease of the fluorescence intensity for
DNA–NR system upon addition of Pt(II) complex. Firstly, the bind-
ing between NR and Pt(II) complex might occur and the intensity of
DNA–NR complex decreases. Secondly, Pt(II) complex competes
against NR in binding with DNA and excludes the intercalated NR
from the duplex, so the concentration of NR binding with DNA re-
duces. This competition binding generally occurs between the two
intercalative binders. Thirdly, Pt(II) complex binding with DNA–NR
forms a new non-fluorescence complex, DNA–NR–Pt(II) complex,
which causes the fluorescence quenching of DNA–NR. In this
Circular dichroism spectroscopy
Circular dichroic spectral technique is useful in diagnosing
changes in DNA morphology during drug–DNA interactions as
the CD signals are quite sensitive to the mode of DNA interactions
with target molecules. A solution of CT-DNA exhibits a positive
band (275 nm) from base stacking interactions and a negative band
(
245 nm) from the right-handed helicity of DNA. These bands are
sensitive towards binding of any small molecule or drug and hence
changes in CD signals of DNA observed on interaction with drugs
may often be assigned to the corresponding changes in DNA struc-
ture [37]. Simple groove binding and electrostatic interaction of
small molecules show less or no perturbation on the base-stacking
and helicity bands, while intercalation enhances the intensities of
both the bands stabilizing the right-handed B conformation of CT
DNA as observed for the classical intercalator methylene blue [38].
The effect of the Pt complex on DNA structure is evaluated by
circular dichroism spectroscopy and the results were shown in
Fig. 6. As it is observed the intensities of both the negative and po-
sitive bands decrease (shifting to zero levels). This suggests that
the DNA binding of the complex induces conformational changes,
Fig. 7. Fluorescence spectra of the competition between Pt(II) complex and Neutral
ꢁ
5
ꢁ5
ꢁ4
Fig. 5. Effect of the Pt(II) complex on the absorption spectrum of DNA–NR. [NR]:
red. CPt(II) complex = 0, 3.37 ꢂ 10 , 6.5 ꢂ 10 , 1.11, 1.59, 2.15, 2.74 and 3.39 ꢂ 10
-
ꢁ
5
ꢁ1
ꢁ5
ꢁ1
ꢁ5
ꢁ1
ꢁ1
ꢁ6
ꢁ1
3
1
ꢂ 10 mol L ; [DNA]:6 ꢂ 10 mol L [Pt(II) complex] (ꢂ10 mol L ): 0.98,
mol L
for
curves
1–8,
C
Nuetral
red = 5.0 ꢂ 10 mol L
,
and
.45, 2.42, 4.28 and 7.8.
C
DNA = 4.97 ꢂ 10ꢁ mol L
5
ꢁ1
.

N. Shahabadi, L. Heidari / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 128 (2014) 377–385
383
experiment, it is found that Pt(II) complex does not react with NR
Table 5
Binding constants (K
f
) and number of binding sites (n) of the DNA-complex system.
when DNA is absence. So, the first reason is not true of interpreting
the decrease of DNA–NR intensity. In addition, the binding con-
stant of NR with DNA is greater than that of the platinum complex.
Pt(II) complex can hardly scramble for the binding sites on DNA
with NR. Besides, in the UV–Vis experiment, it has been already
indicated that the intercalated NR molecules are not restored with
addition of Pt(II) complex to DNA–NR system. So, the second rea-
son also seems not real reason. Then, the third reason above men-
tioned seems to be more reasonable, namely the decrease in the
fluorescence of the system is due to the quenching of DNA–NR
complex by the bound Pt(II) complex [40].
Quenching can occur by different mechanisms, which are usually
classified as dynamic and static quenching. Dynamic quenching re-
fers to a process in which the fluorophore and the quencher come
into contact during the transient existence of the excited state,
while static quenching refers to fluorophore–quencher complex for-
mation. In general, dynamic and static quenching can be distin-
guished by their differing dependence on temperature and
excited-state lifetime. Since in both cases the fluorescence intensity
is related to the concentration of the quencher, the quenched fluo-
rophore can serve as an indicator for the quenching agent [41].
Fluorescence quenching is described by the Stern–Volmer
equation:
T (K)
83
298
310
R2
n
log Kf
Kf
3
3
3
2
0.994
0.991
0.998
1.116
1.179
1.248
3.524
3.727
3.969
3.31 ꢂ 10
5.32 ꢂ 10
9.34 ꢂ 10
of entropy changes (
equation:
DS) can be determined from the van’t Hoff
D
H
DS
LnK ¼ ꢁ
þ
R
RT
where K is the binding constant at the corresponding temperature
and R is gas constant. The values of H and S were obtained from
the slope and intercept of the linear plot (van’t Hoff equation) based
D
D
on lnK versus 1/T. The free energy change (DG) was estimated from
following equation:
D
G ¼
D
H ꢁ T
DS
The values of
D
H, DS and DG between Pt(II) complex with DNA
are listed in Table 6. It can be seen that the negative value of
revealed the interaction process is spontaneous, the positive
DG
D
H
and
DS values indicated that hydrophobic forces play main roles
in the binding of Pt(II) complex to DNA.
F
F
0
¼
1 þ K
q
s
0
½Qꢄ ¼ 1 þ K
sv
½Qꢄ
Viscosity measurements
where F and F represent the fluorescence intensities in the absence
0
Optical photophysical probes generally provide necessary, but
not sufficient clues to support binding mode. The hydrodynamic
measurements are sensitive to the length change of DNA in ab-
sence and presence of foreign molecules. A classical intercalation
model demands that the DNA helix must lengthen as base pairs
are separated to accommodate the binding ligand, leading to in-
crease in DNA viscosity [49]. In contrast, a partial and/or non-clas-
sical intercalation ligand could bend (or kink) the DNA helix,
reduce its effective length and concomitantly its viscosity, while li-
gands that bind exclusively in the DNA grooves (e.g., netropsin,
distamycin), under the same conditions, typically cause less pro-
and in the presence of quencher, respectively. K is the fluorophore
quenching rate constant, Ksv is quenching constant, is the lifetime
of the fluorophore in the absence of a quencher (
is the concentration of quencher [42]. Dynamic and static quench-
ing can be distinguished by their different dependence on temper-
ature [43]. The results in Table 4 indicate that the probable
quenching mechanism of Pt(II) complex by CT-DNA involves static
quenching, because Ksv decreases with increasing temperature [44].
q
s
0
ꢁ
8
0
s = 10 ), and [Q]
Binding constant and the number of binding sites. The binding con-
stant (K ) and the binding stoichiometry (n) for the complex forma-
f
tion between platinum complex with DNA were measured using
the following equation [45]:
nounced changes (positive or negative) or no changes in DNA solu-
1/3
tion viscosity [50]. The values of relative specific viscosity (
here and
absence and in the presence of the Pt(II) complex were plotted
against r (r = [complex]/[DNA]) (Fig. 8).
g/g
0
)
,
g
0
g are the specific viscosity contributions of DNA in the
logðF
0
ꢁ FÞ=F ¼ log K
f
þ n log½Qꢄ
i
i
Here F
0
and F are the fluorescence intensities of the fluorophore
For the platinum complex at lower loading, the viscosity in-
creases and then fixes at higher loadings. In principle, this could
be explained by changes in conformation, flexibility or solvation
of the DNA molecule. The slow increase in viscosity is an indication
of groove binding [51].
in the absence and in the presence of different concentrations of
Q], respectively. The values of K and n are shown in Table 5.
[
f
Thermodynamic studies. The interaction forces between drug and
biomolecule may involve hydrophobic forces, electrostatic interac-
tions, van der Waals interactions, hydrogen bonds, etc. [46].
According to the data of enthalpy changes (
changes ( S), the model of interaction between drug and biomol-
ecule can be concluded [47]: (1) H > 0 and S > 0, hydrophobic
forces; (2) H < 0 and S < 0, van der Waals interactions and
hydrogen bonds; (3) H < 0 and S > 0, electrostatic interactions
48]. When there is little change of temperature, the enthalpy
Cleavage of pUC18 plasmid DNA by metformin ligand and its platinum
complex
DH) and entropy
D
It is known that DNA cleavage is controlled by relaxation of
supercoiled circular conformation of pUC18 DNA, nicked circular
and/or linear conformations. When electrophoresis is applied to
circular plasmid DNA, the relatively fast migration is related to
the intact supercoiled form (Form I). If one strand is cleaved, the
supercoil will be relaxed to produce a slower-moving nicked
D
D
D
D
D
D
[
change (DH) can be seen as a constant, and then its value and that
Table 4
Table 6
The quenching constants of platinum complex by CT-DNA at different temperatures.
Thermodynamic parameters for the binding of Pt(II) complex to CT-DNA.
R2
K
sv (L molꢁ1) ꢂ 10ꢁ3
K
(L molꢁ1) ꢂ 10ꢁ11
T (K)
D
G (kJ mol
ꢁ1
)
D
H (kJ mol
ꢁ1
)
D
S (J mol
ꢁ1 ꢁ1
K )
T (K)
q
283
298
310
0.980
0.981
0.987
1.463
1.445
1.381
1.463
1.445
1.381
283
298
310
ꢁ18.99
ꢁ21.46
ꢁ23.42
27.33
27.33
27.33
163.71
163.71
163.71

3
84
N. Shahabadi, L. Heidari / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 128 (2014) 377–385
[
56]. The binding constant for the platinum complex estimated
from equation below [57]:
ꢀ
ꢁ
ꢀ
ꢁ
1
I
log
¼ log K þ log
½
DNAꢄ
I
0
ꢁ I
0
where K is the binding constant, I and I are the peak currents of the
free guest and the complex, respectively. The binding constant with
3
a value of 1.86 ꢂ 10 was obtained from the intercept of the plot of
log (1/[DNA]) versus log (I/(I
0
ꢁ I)).
Conclusions
Fig. 8. Effect of increasing amounts of platinum complex on the viscosity of CT-DNA
ꢁ
5
(
5 ꢂ 10 M) in 50 mM in Tris–HCl buffer (r
i
= 0.0, 0.2, 0.4, 0.6, 0.9 and 1.0).
In this work we synthesized a Pt(II) complex containing the
drug metformin that bearing the metformin and DMSO ligand,
which is similar to cis-platin complex and expected to show anti-
cancer ability on the binding interaction with DNA and improve
water solubility of the complex and might influence better diffu-
sion through membranes. Different instrumental methods were
used to stndy the interaction mechanism of platinum complex
with CT-DNA and BSA. Finally, The results suggest that Pt(II) com-
plex interact with DNA via groove binding mode. Non-covalent
binding is typically preferred over covalent adduct formation keep-
ing the drug metabolism and toxic side effects in mind. However,
the high binding strength of covalent binders is a major advantage.
In comparison with metformin [58], this study showed that the
Pt center did not change the mode of binding but more experi-
ments especially in the cell lines are in progress.
conformation (Form II). If both strands are cleaved, a linear confor-
mation (Form III) will be generated which migrates in between
[
52,53]. The present study with agarose gel electrophoresis dem-
onstrating that compared with DNA control (Lane 1), the patterns
of Lanes 2–6, does not show any evidences of DNA cleavage (Fig. 9).
Differential pulse voltammetry (DPV)
DPV technique provides higher sensitivity and better peak res-
olution compared to CV for studying the electrochemical behavior
of biological systems [54]. Voltammogram of the platinum com-
plex in the absence and presence of DNA is shown in Fig. 10. In
the absence of CT-DNA, the voltammogram featured an anodic
peak (Epa 1.125 V). In this experiment it can be seen that the anodic
peak currents decreased gradually with the addition of DNA. The
decrease in current may be attributed to the diffusion of the com-
plex bound to the large, slowly diffusing DNA molecule. The de-
creases in the peak currents are ascribed to the stronger binding
between the complexes and DNA [55]. Also during DNA addition,
the anodic peak potential (Epa), showed positive shifts. These posi-
tive shifts are considered as evidences for hydrophobic interaction
that this result is in agreement with fluorescence measurements
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.saa.2014.02.167.
References
[
[
1] C. Xin Zhang, S.J. Lippard, Cur. Opin. Chem. Biol. 7 (2003) 481–489.
2] I. Bertini, H.B. Grey, E.I. Stiefel, J.S. Valentine, Biological Inorganic Chemistry.
Structure and Reactivity, University Science Books, Sausalito, California, 2007.
3] A. Abu-Surrah, M. Kettunen, Curr. Med. Chem. 13 (2006) 1337–1357.
4] J. Reedijk, J. Chem. Soc. Chem. Commun. 7 (1996) 801–806.
[
[
[
5] E.-J. Gao, L. Wang, M.-C. Zhu, L. Liu, W.-Z. Zhang, Eur. J. Med. Chem. 45 (2010)
311–316.
[
[
[
[
6] G.Y. Park, J.J. Wilson, Y. Song, S.J. Lippard, PNAS 109 (2012). 11987–1192.
7] K.S. Lovejoy, R.C. Todd, Sh Zhang, PNAS 105 (2008) 8902–8907.
8] N. Margiotta et al., J. Inorg. Biochem. 100 (2006) 1849–1857.
9] L. Steven Hollis, Alan R. Amundsen, Eric W. Stern, J. Med. Chem. 32 (1) (1989)
1
28–136.
10] Q.L. Zhang, J.G. Liu, H. Chao, G.Q. Xue, L.N. Ji, J. Inorg. Biochem. 83 (2001) 49–
5.
[
5
[
[
11] P. Hubberstey, U. Suksangpanya, Struct. Bond. 111 (2004) 33–83.
12] G. Patrinoiu, L. Patron, O. Carp, N. Stanica, J. Therm. Anal. Cal. 72 (2003) 489–
495.
Fig. 9. Gel electrophoresis of pUC18 in the presence of increasing amounts of Pt(II)
complex (r = 0, 0.1, 0.3, 0.5,0.7 and 0.9).
i
[
[
[
[
[
[
13] P. Rây, Chem. Rev. 61 (1961) 313–359.
14] A. Syamal, J. Sci. Ind. Res. 37 (1978) 654–661.
15] A.J.J. Wood, C.J. Bailey, R.C. Turner, New England J. Med. 334 (2003) 574–579.
16] P. Pignard, Ann. Biol. Clin. 20 (1962) 325–333.
17] G. Siest, F. Roos, J.J. Gabou, Bull. Soc. Pharm. Nancy 58 (1963) 29–38.
18] F. Bentefrit, G. Morgant, B. Viossat, S. Leonce, N. Guilbaud, A. Pierre, G. Atassi,
D. Nguyen-Huy, J. Inorg. Biochem. 68 (1997) 53–59.
[
[
19] R. Olar, M. Badea, E. Cristurean, V. Lazar, R. Cernat, C. Balotescu, J. Therm. Anal.
Cal. 80 (2005) 451–455.
20] P.V. Babykutty, C.P. Prabhakaran, R. Anantaraman, C.G.R. Nair, J. Inorg. Nucl.
Chem. 36 (1974) 3685–3688.
[
[
21] E. Szlyk, Polyhedron 21 (2002) 2001–2007.
22] A. Foziah Al-Saif, Moamen S. Refat, J. Therm. Anal. Calorim. 3 (2012) 2459–
2468.
[
[
23] W.J. Geary, Coord. Chem. Rev. 7 (1971) 81–122.
24] Spatran ’10-Quantum Mechanics Program: (PC/x86) 1.1.0v4, Wavefunction
Inc., USA, 2011.
[
[
25] J.B. Chaires, Curr. Med. Chem. Anticancer Agents 5 (4) (2005) 339–352.
26] M.T. Carter, M. Rodriguez, A.J.J. Bard, J. Am. Chem. Soc. 111 (1989) 8901–8911.
Fig. 10. Differential pulse voltammogram of 1 mM platinum complex on a polished
glassy carbon electrode in the absence and presence of different concentrations of
DNA.
[27] H. Gonzalez-Diaz, A. Sanchez-Gonzalez, Y. Gonzalez-Diaz, J. Inorg. Biochem.
100 (2006) 1290–1297.
[28] Y.N. Xiao, C.X. Zhan, J. Appl. Polym. Sci. 84 (2002) 887–893.

N. Shahabadi, L. Heidari / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 128 (2014) 377–385
385
[
[
29] S.E. Sherman, D. Gibson, A.H.J. Wang, S.J. Lippard, J. Am. Chem. Soc. 110 (1988)
368–7381.
30] T. Hirohama, Y. Kuranuki, E. Ebina, T. Sugizaki, H. Arii, M. Chikira, P.T. Selvi, M.
Palaniandavar, J. Inorg. Biochem. 99 (2005) 1205–1219.
[44] P.B. Kandagal, S.M.T. Shaikh, D.H. Manjunatha, J. Seetharamappa, B.S. Nagaralli,
J. Photochem. Photobiol. 189 (2007) 121–127.
[45] M. Jiang, M.X. Xie, D. Zheng, Y. Liu, X.Y. Li, X. Chen, J. Mol. Struct. 692 (2004)
71–80.
7
[
[
31] F. Arjmand, F. Sayeed, J. Mol. Struct. 965 (2010) 14–22.
32] S. Sharma, S. Singh, K.M. Chandra, D.S. Pandey, J. Inorg. Biochem. 99 (2005)
[46] D.P. Ross, S. Subramanian, J. Biochem. 20 (1981) 3096–3102.
[47] D.B. Nike, P.N. Moorty, K.I. Priyadarsini, Chem. Phys. Lett. 168 (1990) 533–538.
[48] N. Shahabadi, S. Hadidi, Spectrochim. Acta Part A Mol. Biomol. Spectrosc. 96
(2012) 278–283.
[49] J.B. Chaires, Biopolymer. (Nucleic Acid Sci.) 44 (1998) 201–215.
[50] F. Arjmand, M. Aziz, Eur. J. Med. Chem. 44 (2009) 834–844.
[51] P.X. Xi, Z. Xu, X.H. Liu, F.J. Chen, Z.Z. Zeng, X.W. Zhang, Y. Liu, J. Fluoresc. 19
(2009) 63–72.
[52] S. Parveen, F. Arjmand, Spectrochim. Acta Part A 85 (2012) 53–60.
[53] P.R. Reddy, K.S. Rao, B. Satyanarayana, Tetrahedron Lett. 47 (2006) 7311–7315.
[54] S. Shankara, K. Umesha, S. Jaldappagari, J. Electroanal. Chem. 636 (2009) 93–
100.
4
58–466.
[
[
33] Q.L. Zhang, J.G. Liu, J. Liu, G.Q. Xue, J. Inorg. Biochem. 85 (2001) 291–296.
34] X.L. Wang, X.L. Hong, Y.J. Liu, L.F. Tan, L.N. Ji, J. Inorg. Biochem. 98 (2004) 1143–
1
150.
[
[
[
[
[
[
35] S. Kashanian, J.EzzatiNazhad Dolatabadi, Food Chem. 116 (2009) 743–747.
36] C.Z. Huang, Y.F. Li, P. Feng, Talanta 55 (2001) 321–328.
37] P. Lincoln, E. Tuite, B. Norden, J. Am. Chem. Soc. 119 (1997) 1454–1455.
38] B. Norden, F. Tjerneld, Biopolymers 21 (1982) 1713–1734.
39] P.U. Maheswari, M. Palaniandavar, J. Inorg. Biochem. 98 (2004) 219–230.
40] B. Shuyun, Z. Hanqi, Q. Chunyu, S. Ying, L. Chunming, Spectrochim. Acta Part A
6
9 (2008) 123–129.
[55] Y. Fei, G. Lu, G. Fan, Y. Wu, Anal. Sci. 25 (2009) 1333–1338.
[56] A. Bard, M. Cater, M. Rodoriguez, J. Am. Chem. Soc. 111 (1989) 8901–8911.
[57] Q. Feng, N. Li, Y. Jiang, Anal. Chim. Acta 344 (1997) 97–104.
[58] N. Shahabadi, L. Heidari, Spectrochim. Acta A 97 (2012) 406–410.
[
[
41] M.R. Eftink, C.A. Ghiron, Anal. Biochem. 114 (1981) 199–227.
42] Y. Sun, S. Bi, D. Song, C. Qiao, D. Mu, H. Zhang, Sens. Actuators B 129 (2008)
7
99–810.
[
43] F.L. Cui, J. Fan, J.P. Li, Z.D. Hu, Bioorg. Med. Chem. 12 (2004) 151–157.