Crystal structures, antioxidation and DNA binding properties of Dy(III) complexes with Schiff-base ligands derived from 8-hydroxyquinoline-2-carboxaldehyde and four aroylhydrazines

Article abstract of DOI:10.1016/j.ejmech.2009.09.015

X-ray crystal and other structural analyses indicate that Dy(III) and every ligand can form a binuclear Dy(III) complex with nine-coordination and 1:1 metal-to-ligand stoichiometry at the Dy(III) center. All the ligands and Dy(III) complexes can bind to C

Full text of DOI:10.1016/j.ejmech.2009.09.015

European Journal of Medicinal Chemistry 44 (2009) 5080–5089
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Crystal structures, antioxidation and DNA binding properties of Dy(III)
complexes with Schiff-base ligands derived from 8-hydroxyquinoline-2-
carboxaldehyde and four aroylhydrazines
,
,
Yong-chun Liu a b, Zheng-yin Yang a
*
a College of Chemistry and Chemical Engineering, State Key Laboratory of Applied Organic Chemistry, Lanzhou University, Lanzhou 730000, PR China
b College of Chemistry and Chemical Engineering, Longdong University, Qingyang, Gansu 745000, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
X-ray crystal and other structural analyses indicate that Dy(III) and every ligand can form a binuclear
Dy(III) complex with nine-coordination and 1:1 metal-to-ligand stoichiometry at the Dy(III) center. All
the ligands and Dy(III) complexes can bind to Calf thymus DNA through intercalations with the binding
Received 1 January 2009
Received in revised form
13 August 2009
Accepted 10 September 2009
Available online 16 September 2009
constants at the order of magnitude 105–107
M
ꢀ1, but Dy(III) complexes present stronger affinities to
DNA than ligands. All the ligands and Dy(III) complexes can be used as potential anticancer drugs. All the
ligands and Dy(III) complexes have strong scavenging effects for hydroxyl radicals and superoxide
radicals but complex containing active phenolic hydroxyl group shows stronger scavenging effects for
hydroxyl radicals and complex containing N-heteroaromatic substituent shows stronger scavenging
effects for superoxide radicals.
Keywords:
Rare earths
Schiff bases
Ó 2009 Elsevier Masson SAS. All rights reserved.
X-ray crystallography
Calf thymus DNA binding properties
Antioxidation
1. Introduction
the binding [13–15]. A number of metal chelates, as agents for
mediation of strand scission of duplex DNA and as chemothera-
The chemistry of quinoline and its derivatives has attracted
special interest due to their therapeutic properties. Quinoline sul-
phonamides have been used in the treatment of cancer, tubercu-
losis, diabetes, malaria, and convulsion [1,2]. Apart from the
magnetic and photophysical properties, the bioactivities of
lanthanides such as antimicrobial, antitumor, antivirus, anticoag-
ulant action, enhancing NK and macrophage cell activities,
prevention from arteriosclerosis, etc., have been explored in recent
decades [3–10]. In addition, Schiff bases are able to inhibit the
growth of several animal tumors, and some metal chelates have
shown good antitumor activities against animal tumors [11,12]. So,
well-designed organic ligands enable a fine turning of special
properties of the metal ions.
peutic agents, have been used as probes of DNA structure in solu-
tion [16–18]. Albrecht and coworkers reported that both the crystal
structures of [YLNO3)(DMF)2]2Cl2$2(DMF) and [LaL(NO3)(-
MeOH)2]2(NO3)2 with nine-coordination have central planar four-
membered (LaO)2 and (YbO)2 rings, respectively, where ligand L–H
is 2-[(8-hydroxyquinolinyl)methylene]hydrazinecarboxamide and
acts as a tetradentate ligand binding to yttrium(III) and lanthanu-
m(III) [4]. Such structures may have strong affinities of binding to
DNA through intercalations. In this paper, four Schiff-base ligands,
which were structurally similar to the ligand L, were prepared from
8-hydroxyquinoline-2-carboxaldehyde with four aroylhydrazines
to form their Dy(III) complexes and to investigate the DNA binding
properties.
DNA is an important cellular receptor, many chemicals exert
their antitumor effects by binding to DNA thereby changing the
replication of DNA and inhibiting the growth of the tumor cells,
which is the basis of designing new and more efficient antitumor
drugs and their effectiveness depends on the mode and affinity of
On the other hand, an excess of activated oxygen species in the
ꢁ
forms of superoxide anion (Oꢀ2 ) and hydroxyl radical (OHꢁ),
generated by normal metabolic processes, may cause various
diseases such as carcinogenesis, drug-associated toxicity, inflam-
mation, atherogenesis, and aging in aerobic organisms [19–21].
Although the naturally occurring antioxidants can scavenge free
radicals in the body, they have been confined by their low effec-
tiveness even though they are considered to be active in elimi-
nating reactive oxygen and controlling toxic effects. The potential
* Corresponding author. Tel.: þ86 931 8913515; fax: þ86 931 8912582.
E-mail address: yangzy@lzu.edu.cn (Z.-y. Yang).
0223-5234/$ – see front matter Ó 2009 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2009.09.015

Y.-c. Liu, Z.-y. Yang / European Journal of Medicinal Chemistry 44 (2009) 5080–5089
5081
value of antioxidants has prompted investigators to search for the
cooperative effects of metal complexes and natural compounds for
improving antioxidant activity and cytotoxicity [22]. It has been
recently demonstrated that some minor groove binders for DNA are
effective inhibitors of the formation of a DNA/TBP complex or
topoisomerases [23–25]. Adding a reactive entity endowed with
oxidative properties should improve the efficiency of inhibitors.
The antioxidation properties of the ligands and Dy(III) complexes
were investigated in this paper. Furthermore, the substituent
effects of these compounds on antioxidation and DNA binding
properties were investigated further.
membered (DyO)2 ring formed by the two Dy atoms and the
phenolic oxygen atoms. Dimerization of this monomeric unit
occurs through the phenolate oxygen atoms leading to a central
four-membered (DyO)2-ring with a Dy/Dy separation of 4.027 Å.
At the dimerization site, a ‘‘set off’’ of the two ‘‘DyL1-planes’’ by
1.637 Å takes place. This crystal structure of binuclear [DyL1(-
NO3)(DMF)2]2 complex with a 1:1 metal-to-ligand stoichiometry
and nine-coordination is similar to that of [YL(NO3)-
(DMF)2]2Cl2$2(DMF) or [LaL(NO3)(MeOH)2]2(NO3)2, where ligand
L–H is 2-[(8-hydroxyquinolinyl)methylene]hydrazinecarboxamide
and acts as a monad tetradentate ligand binding to yttrium(III) or
lanthanum(III) through the phenolate oxygen atom, nitrogen atom
of quinolinato unit and the C]N group, C]O group of the
semicarbazone side chain [4]. However, there are two marked
differences between them. One is that the ‘‘set off’’ of the two
‘‘DyL1-planes’’ by 1.637 Å takes place while the ‘‘set off’’ of the two
‘‘YL-planes’’ and ‘‘LaL-planes’’ by approximately 2 Å takes place.
Another is that O]C–NH– group of the benzoylhydrazine side
chain has enolized and deprotonated into ꢀO–C]N– group after
the formation of [DyL1(NO3)(DMF)2]2 complex, where the C–Oꢀ
single bond length is 1.294(4) Å and the N]C double bond length is
1.334(4) Å. Whereas the O]C–NH– group of the semicarbazone
side chains has not enolized in [YL(NO3)(DMF)2]2Cl2$2(DMF) or
[LaL(NO3)(MeOH)2]2(NO3)2 complex. The difference in the depro-
tonization and enolization may well be due to the fact that trie-
thylamine was added into the reaction mixtures to deprotonate the
phenolic hydroxyl substituent of 8-hydroxyquinolinato unit during
the formation of the Dy(III) complex. It is the enolization and
deprotonation of O]C–NH– group changing into ꢀO–C]N– that
[DyL1(NO3)(DMF)2]2 is of neutral charge and non-electrolyte, but
both of [YL(NO3)(DMF)2]2Cl2$2(DMF) and [LaL(NO3)(MeOH)2]2
(NO3)2 are electrolytes. The enolization and deprotonation will
afford an efficient route for investigators to design favorable
molecules well.
2. Chemistry
Four Schiff-base ligands, 8-hydroxyquinoline-2-carbox-
aldehyde-(benzoyl)hydrazone (1a, H2L1), 8-hydroxyquinoline-
2-carboxaldehyde-(20-hydroxybenzoyl)hydrazone (1b, H2L2),
8-hydroxyquinoline-2-carboxaldehyde-(40-hydroxybenzoyl)hy-
drazone (1c, H2L3) and 8-hydroxyquinoline-2-carboxaldehyde-
(isonicotinyl)hydrazone (1d, H2L4) were prepared from equimolar
amounts of 8-hydroxyquinoline-2-carboxaldehyde and benzoyl-
hydrazine, 2-hydroxybenzoylhydrazine, 4-hydroxybenzoylhy-
drazine, and isonicotinylhydrazine, respectively, whose structures
were determined by IR spectroscopy, 1H NMR and ESI-MS. The
synthetic routes for ligands are presented in Scheme 1. Then, their
Dy(III) complexes (2a–d) were easily prepared from these ligands
and equimolar amounts of Dy(NO3)$6H2O, respectively.
3. Results and discussion
3.1. Crystal structure analyses of the Dy(III) complexes
The orange transparent, X-ray quality crystals of complex 2a and
2b were obtained by vapor diffusion of diethyl ether into DMF
solution of the powder complex at room temperature for 2 weeks,
respectively. Crystal data and structure refinements for the X-ray
structural analyses are presented in Table 1. Selected bond lengths
and angles of the metal complexes are presented in Tables S1 and
S2 (see Supplementary data Tables S1–S2).
3.1.2. The crystal structure of complex [DyL2(NO3)(DMF)2]2
The ORTEP illustration (30% probability ellipsoids) of complex
2b in Fig. 1B shows that the complex composition is of [DyL2(-
NO3)(DMF)2]2. The crystal of a binuclear [DyL2(NO3)(DMF)2]2
complex with a 1:1 metal-to-ligand stoichiometry and nine-coor-
dination is much similar to that of [DyL1(NO3)(DMF)2]2. Both the
complexes are almost isostructural. The center of symmetry is at 1,
1, and 0. It is located in the middle of the four-membered (DyO)2
ring. Dimerization of this monomeric unit also occurs through the
phenolate oxygen atoms leading to a central four-membered
(DyO)2-ring with a Dy/Dy separation of 4.005 Å. At the dimer-
ization site, a ‘‘set off’’ of the two ‘‘DyL2-planes’’ by 1.640 Å takes
place. The O]C–NH– group of the 2-hydroxylbenzoyhydrazine side
chain has also enolized and deprotonated into ꢀO–C]N– after the
formation of [DyL2(NO3)(DMF)2]2 complex, where the ꢀO–C and
N]C bond lengths are 1.273(3) and 1.327(3) Å, respectively. On the
other hand, the 2-hydroxyl substituent linking with benzoylhy-
drazine has not take part in binding to Dy(III), largely due to the
steric effect and long distance (5.217 Å) between the 2-hydroxyl
substituent and Dy(III), but it may form an intramolecular hydrogen
bond, a stabilizing six-membered ring, with adjacent nitrogen atom
(1.843 Å) of the same side chain.
3.1.1. The crystal structure of complex [DyL1(NO3)(DMF)2]2
The ORTEP illustration (30% probability ellipsoids) of complex
2a in Fig. 1A shows that the complex composition is of [DyL1(-
NO3)(DMF)2]2. Ligand 1a acts as a dibasic tetradentate ligand,
binding to Dy(III) through the phenolate oxygen atom, nitrogen
atom of quinolinato unit and the C]N group, ꢀO–C]N– group
(enolized and deprotonated from O]C–NH–) of the benzoylhy-
drazine side chain. In addition, one DMF molecule is binding
orthogonally to the ligand-plane from one side to the metal ion,
while another DMF and nitrate (bidentate) are binding from the
other. The center of symmetry according to the crystallographic
coordinate is at 1/2, 0, 1/2. It is located in the middle of the four-
5
4
3
6
7
O
20
17
N
O
19
18
R2
N
N
N
11
H
16
ethanol
OH
OH
R2
+
R1
R3
95
%
3.2. Structure analyses of the powder metal complexes
O
H2L1 (1a): R1=H, R2=C, R3=H
NH2
3.2.1. Elemental analysis and molar conductance
H2L2 (1b): R1=OH, R2=C, R3=H
H2L3 (1c): R1=H, R2=C, R3=OH
H2L4 (1d): R1=H, R2 R3=N
N
H
All the Dy(III) complexes are of orange powders, stable in air,
and soluble in DMF and DMSO, but slightly soluble in methanol,
ethanol, acetonitrile, ethyl acetate and acetone, THF and CHCl3. The
melting points of all the Dy(III) complexes exceed 300 ꢂC.
R3
R1
Scheme 1. The synthetic routes for ligands (1a–d).

5082
Y.-c. Liu, Z.-y. Yang / European Journal of Medicinal Chemistry 44 (2009) 5080–5089
Table 1
Crystal data and structure refinement of metal complexes.
Complex
[DyL1(NO3)(DMF)2]2
[DyL2(NO3)(DMF)2]2
Empirical formula
Formula weight
Crystal system
Space group
Unit cell dimensions
C46H50N12O14Dy2
1319.98
Monoclinic
P21/c
C46H50N12O16Dy2
1351.98
Monoclinic
P21/c
a ¼ 11.444(3) Å,
b ¼ 18.796(5) Å,
c ¼ 12.377(3) Å,
273(2) K
a
b
g
¼ 90.00ꢂ
a ¼ 11.6129(6) Å,
a
¼ 90.00ꢂ
¼ 92.316(4)ꢂ
¼ 90.00ꢂ
b ¼ 18.0596(10) Å,
b
¼ 95.2960(10)ꢂ
c ¼ 12.7482(7) Å,
273(2) K
g
¼ 90.00ꢂ
T
Volume
Z
Density (calculated)
Absorption coefficient
F(000)
2660.2(12) Å3
2
2662.2(2) Å3
2
1.648 mg/m3
2.861 mmꢀ1
1308
1.687 mg/m3
2.863 mmꢀ1
1340
Crystal size
Crystal description
0.22 ꢃ 0.20 ꢃ 0.18 mm3
Orange block
0.18 ꢃ 0.15 ꢃ 0.13 mm3
Orange block
q
Range for data collection
1.78–26.49ꢂ
1.76–28.35ꢂ
Index ranges
ꢀ14 ꢄ h ꢄ 14, ꢀ23 ꢄ k ꢄ 18, ꢀ15 ꢄ l ꢄ 15
ꢀ15 ꢄ h ꢄ 14, ꢀ20 ꢄ k ꢄ 24, ꢀ15 ꢄ l ꢄ 14
Reflections collected
Independent reflections
Absorption correction
Max. and min. transmission
Refinement method
Data/restraints/parameters
Goodness-of-fit on F2
14,615
15,968
5488 (Rint ¼ 0.0281)
Semi-empirical from equivalents
0.597 and 0.539
6399 (Rint ¼ 0.0195)
Semi-empirical from equivalents
0.688 and 0.601
Full-matrix least-squares on F2
5488/0/338
Full-matrix least-squares on F2
6399/0/348
1.010
1.124
Final R indices [I > 2
R indices (all data)
Largest diff. peak and hole
s
(I)]
R1 ¼ 0.0234, wR2 ¼ 0.0462
R1 ¼ 0.0392, wR2 ¼ 0.0504
0.601 and ꢀ0.379 eÅꢀ3
R1 ¼ 0.0229, wR2 ¼ 0.0498
R1 ¼ 0.0363, wR2 ¼ 0.0536
0.600 and ꢀ0.522 eÅꢀ3
Elemental analyses indicate that all the Dy(III) complexes are of 1:1
metal-to-ligand (stoichiometry) complexes, and the data of molar
conductance of the Dy(III) complexes in DMF solutions indicate that
all of them act as non-electrolytes [26].
3.2.2. Infrared spectrum study
The characteristic IR spectrum bands (nmax/cmꢀ1) of all the
ligands showed 3576–3320vs assigned to
assigned to (CO) of the carbonyl groups of aroylhydrazine side
chains and 1632–1602 assigned to (CN) of azomethines, whereas
1706s assigned to (CO) of the formyl disappeared. Moreover, 3318–
3139br and 1288–1267s should be assigned to (OH) and (C–OH) of
the phenolic hydroxyl substituents, respectively, and 1581–1532
should be assigned to (CN) of pyridines.
n(NH); 1682–1643s
n
n
n
n
n
n
Carefully compared with the characteristic IR bands of ligands
(data are listed in Table S3), it comes to the conclusion that: (1)
3406–3399br, 936–922w and 642–612w can be assigned to n(OH) of
H2O, rr (H2O) and rw (H2O), respectively, indicating that there are
coordinated water molecules participating in the Dy(III) complexes
[27,28]. (2) 1104–1102 can be assigned to n(C–OM), indicating that
the binding of metal ion to every ligand through an O–M linkage
takes place [29]. (3) The new bands of 3216s and 1256s can be,
respectively, assigned to n(OH) and n(C–OH) of the free phenolic
hydroxyl substituent of 2-hydroxybenzoylhydrazine side chain of
2b complex, and the new bands of 3193s and 1289s also can be,
respectively, assigned to
hydroxyl substituent of 4-hydroxybenzoylhydrazine side chain of
2c complex. (4) 1682–1643s assigned to (CO) and 3576–3320vs
assigned to (NH) of aroylhydrazine side chains of ligands have
n(OH) and n(C–OH) of the phenolic
n
n
disappeared, indicating that they participate in the Dy(III)
complexes with the groups of O]C–NH– of aroylhydrazine side
chains enolized and deprotonated into ꢀO–C]N– as proved by the
above crystal structural analyses. (5) 1634–1598 assigned to n(CN)
of azomethines of the Dy(III) complexes has shifted by 34–38 cmꢀ1
in comparison with bands of ligands, indicating that the nitrogen
atoms of azomethines participate in the complexes. (6) 1565–1548
Fig. 1. The ORTEP illustration (30% probability ellipsoids) of [DyL1(NO3)(DMF)2]2 (A)
and [DyL2(NO3)(DMF)2]2 (B) complexes.
assigned to n(CN) of pyridines of the Dy(III) complexes has shifted

Y.-c. Liu, Z.-y. Yang / European Journal of Medicinal Chemistry 44 (2009) 5080–5089
5083
by 31–36 cmꢀ1 in comparison with bands of ligands, indicating that
the nitrogen atoms of pyridines also participate in the complexes.
However, the new band of 1592 can be assigned to (CN) of free
pyridine of isonicotinylhydrazine side chain of 2d complex. (7)
585–563w assigned to (MO) and 492–489w assigned to (MN) of
A
1d
1c
1b
1a
1.4
1.3
1.2
1.1
1.0
n
n
n
the Dy(III) complexes further indicate that oxygen atoms and
nitrogen atoms participate in Dy(III) complexes. (8) All the Dy(III)
complexes show 1494–1491 (n1), 1336–1309 (n4), 1070–1058 (n2),
843–839 (n3), 761–740 (n5), and Dn
cating that nitrate ions bidentately participate in the Dy(III)
complexes [30].
(
n1
–
n4) ¼ 184–158 cmꢀ1, indi-
Additionally, the ESI-MS data show that the m/z ([M þ H]þ, DMF
solution) are 1321.2, 1353.2, 1353.2 and 1323.2 for 2a, 2b, 2c and 2d
complexes, respectively, indicating that the four coordinated water
molecules for every powder Dy(III) complex can be replaced by four
DMF molecules when dissolved in DMF solution. However, the
results of elemental analyses, molar conductance, IR and ESI-MS
data indicate that all the powder metal complexes are structurally
similar to each other and their compositions are of [DyL1–
4(NO3)(H2O)2]2.
0.0
0.1
0.2
Cligand /CDNA
0.3
0.4
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1.0
B
2d
2c
2b
2a
3.3. DNA binding properties
3.3.1. Viscosity titration measurements
Viscosity measurements are very sensitive to changes in the
length of DNA, as viscosity is proportional to L3 for rod-like DNA of
length L. Viscosity titration measurements were carried out to
clarify the interaction modes between the investigated compounds
and CT-DNA. Intercalation involves the insertion of a planar mole-
cule between DNA base pairs, which results in a decrease in the
DNA helical twist and lengthening of the DNA, therefore inter-
calators cause the unwinding and lengthening of DNA helix as base
pairs become separated to accommodate the binding compound
[31,32]. Whereas, agents bound to DNA through groove binding do
not alter the relative viscosity of DNA, and agents electrostatically
bound to DNA will bend or kink the DNA helix, reducing its effective
length and its viscosity, concomitantly [33,34]. The effects of
ligands and Dy(III) complexes on the viscosities of CT-DNA are
shown in Fig. 2. With the ratios of the investigated compounds to
DNA increase, the relative viscosities of DNA increase steadily,
indicating that there exist intercalations between all the ligands
and Dy(III) complexes with DNA helix. In addition, the relative
viscosities of DNA increase with the order of 1a > 1b > 1c > 1d for
ligands, the order of 2a > 2b > 2c > 2d for Dy(III) complexes, and
the orders of 2a > 1a, 2b > 1b, 2c > 1c and 2d > 1d. These orders
suggest the extents of the unwinding and lengthening of DNA helix
by compounds and the affinities of compounds binding to DNA,
which may be due to the key roles of substituent effects and the
larger coplanar structures of Dy(III) complexes than those of
ligands. Intercalation has been traditionally associated with mole-
cules containing fused bi/tricyclic ring structures [35], so it is logical
that all the large coplanar Dy(III) complexes containing fused
multiple cyclic ring structures and all the ligands containing fused
bicyclic ring structures can bind to DNA through intercalations.
0.0
0.1
0.2
0.3
0.4
Ccomplex /CDNA
Fig. 2. Effects of increasing amounts of the investigated compounds on the relative
viscosity of CT-DNA in 5 mmol Tris–HCl buffer solution (pH 7.20) containing 50 mmol
NaCl at 25.00 ꢅ 0.01 ꢂC. Plots of (A) and (B) represent the ligands–CT-DNA and Dy(III)
complexes–CT-DNA systems, respectively. The concentration of CT-DNA was 50
(bps).
mM
spectra of Dy(III) complexes have two types of absorption bands at
lmax in the regions of 326–332 nm (
3
¼ 2.94–4.49 ꢃ 104 dm3 molꢀ1
cmꢀ1
)
and 372–379 nm
(3
¼ 2.99–3.73 ꢃ104 dm3 molꢀ1 cmꢀ1),
which can be, respectively, assigned to
larger conjugated organic molecules and
p–p* transitions of the
p–p* of the C]N–N]C
groups coupled with charge transfers from ligands to metal ions
(L / Dy3þ) [27,28]. The band shifts of lmax and the changes of
3
for
complexes in comparison with ligands indicate the formations of
the Dy(III) complexes.
Upon successive additions of CT-DNA, the UV–Vis absorption
bands of ligands 1a, 1b, 1c and 1d show a progressive hypo-
chromism of 34.3% at 295 nm, 30.1% at 294 nm, 22.5% at 300 nm
and 8.4% at 290 nm by approximately saturated titration end points
at CDNA:Cligand ¼ 1.4–2.2:1, respectively, with a 1, 3, 1 and 0 nm red
shifts of absorption bands in the region of 290–300 nm. Ligands 1a,
1b, 1c and 1d show another progressive hypochromism of 11.1% at
323 nm, 18.1% at 329 nm, 4.8% at 326 nm and 1.0% at 325 nm,
respectively, with a 1, 3, 1 and 0 nm blue shifts in the region of 323–
329 nm. Similarly, upon successive additions of CT-DNA, the UV–
Vis absorption bands of metal complexs 2a, 2b and 2c show
a progressive hypochromism of 40.6% at 326 nm, 23.3% at 327 nm
and 22.2% at 332 nm by approximately saturated titration end
points at CDNA:Ccomplex ¼ 1.0–1.4:1, respectively, but all of them
have no shift of absorption bands. Complexs 2a, 2b and 2c show
another progressive hypochromism of 38.5% at 373 nm with 1 nm
blue shift, 23.8% at 377 nm with 5 nm red shift and 20.1% at 379 nm
with no band shift, respectively. Complex 2d shows two types of
3.3.2. Ultraviolet–visible (UV–Vis) spectroscopy study
The UV–Vis absorption spectra of the investigated compounds
in the absence and in the presence of the CT-DNA were obtained in
DMF:Tris–HCl buffer (5 mmol, pH 7.20) containing 50 mmol NaCl of
1:100 solutions, respectively (data are listed in Table S4). The UV–
Vis spectra of ligands have two types of absorption bands at lmax in
the regions of 290–300 nm (
3
¼ 2.86–3.55 ꢃ104 dm3 molꢀ1 cmꢀ1
)
and 323–329 nm (
3
¼ 1.78–2.36 ꢃ 104 dm3 molꢀ1 cmꢀ1), which can
be assigned to
p–
p
* transitions within the organic molecules, and
p–p* of the C]N and C]O groups, respectively. While the UV–Vis

5084
Y.-c. Liu, Z.-y. Yang / European Journal of Medicinal Chemistry 44 (2009) 5080–5089
slightly unsteady hypochromisms of 0.45% at 327 nm with 1 nm
blue shift and 0.72% at 372 nm with 1 nm red shift by an approxi-
mately saturated titration end point at CDNA:Ccomplex ¼ 1.4:1. In
addition, isosbestic points at 342–357 nm for ligands and at 404–
423 nm for Dy(III) complexes are observed, indicating that the
reaction between every investigated compound and DNA takes
places by an equilibrium.
2.4
2.2
2.0
1.8
1.6
1.4
1.2
1.0
A
(b)
(c)
(d)
(a)
Absorption titration can monitor the interaction of a compound
with DNA. The obvious hypochromism and red shift are usually
characterized by the noncovalently intercalative binding of
compound to DNA helix, due to the strong stacking interaction
between the aromatic chromophore of the compound and base
pairs of DNA [36,37]. However, the intercalation between
a compound and DNA helix cannot be excluded only by no or small
red shift of UV–Vis absorption bands. In fact, some groove binders
of Hoechst 33258 family can also present red shifts or even blue
shifts of absorption bands when they bind to DNA helix by groove
binding modes, especially for multiple binders [38,39]. After all,
hydrodynamic measurements that are sensitive to length change of
DNA (i.e., viscosity and sedimentation) are regarded as the least
ambiguous and the most critical criterions for binding modes in
solution in the absence of crystallographic structural data [40,41].
On the other hand, the magnitude of hypochromism is parallel
to the intercalative strength and the affinity of a compound binding
to DNA [42]. The appreciable hypochromisms of ligands and Dy(III)
complexes intercalating to DNA present the order of
1a > 1b > 1c > 1d and the order of 2a > 2b > 2c > 2d, which are in
good agreement with the orders of viscosity titration results. Here,
the substituent effects may play key roles in the interactions.
Besides the same structural units of these Dy(III) complexes, as for
complex 2a, the phenyl substituent may be more accessible to DNA
0
10
20
30
[Q] / µM
40
50
60
B
2.4
2.0
1.6
1.2
(a)
(c)
(b)
(d)
0
5
10 15 20 25 30 35 40 45
[Q] / µM
helix and much favorable of forming
p–p stacking interaction
between the aromatic chromophore of the complex and the base
pairs of DNA than 2-hydroxyphenyl and 4-hydroxyphenyl sub-
situents of 2b and 2c complexes. For complex 2b, 2-hydroxy
locating in benzoylhydrazine side chain may form an intra-
molecular hydrogen bond, a stabilizing six-membered ring that is
coplanar to one of DyL2-planes, with adjacent nitrogen atom of the
same side chain, which will induce smaller steric hindrances than
exposed 4-hydroxy locating in benzoylhydrazine side chain for
complex 2c. As for complex 2d, N atom of aromatic sextet of the
pyridine ring of isonicotinylhydrazine side chain has an exposed
and non-hybridized p orbital containing long pair electrons, which
Fig. 3. Stern–Volmer plots of Fo/F versus [Q] for ligands (A) and Dy(III) complexes (B).
Tests were performed in the conditions of 5 mmol Tris–HCl buffer containing 50 mmol
NaCl at 298 K. CDNA ¼ 4
m
M (nucleotides), CEtBr ¼ 0.32 M; lex ¼ 525 nm, lem ¼ 587 nm.
m
Lines of (a), (b), (c) and (d) in plot (A) for ligand 1c, 1b, 1a and 1d, respectively, while
lines of (a), (b), (c) and (d) in plot (B) for complex 2c, 2b, 2a and 2d, respectively.
(2.0 ꢃ 1010
M
ꢀ1 sꢀ1), the maximum quenching rate constant of
bimolecular diffusion collision, which are indicative of static types
of quenching mechanisms arisen from the formations of dark
complexes between the fluorophores and quenching agents
[43,44]. The loss of fluorescence intensity at the maximum wave-
length indicates the displacement of EtBr from EtBr–DNA complex
by a compound and the intercalative binding between the
compound with DNA [34]. The EtBr–DNA quenching results indi-
cate that most of the EtBr molecules have been displaced from
EtBr–DNA complex by every quencher at the approximately satu-
rated end point. Thus, it is reasonable that there exist intercalations
between DNA and these investigated compounds.
may result in a stronger electronic repulsion and hinder the
p–p
stacking interaction. Moreover, the aggregation of self-stacked
molecules of 2d may occur in the absence of DNA, even in an excess
of conjugate versus DNA bps, which induces the possibility of an
association/dissociation equilibrium in the reaction solution,
concomitantly, induces a slightly unsteady and small hypochrom-
ism of UV–Vis absorption [39].
Moreover, the Stern–Volmer dynamic quenching constants can
also be interpreted as binding affinities of the complexation reac-
tions [45,46]. The data of KSV present the order of 1c > 1b > 1a > 1d
for ligands, the order of 2c > 2b > 2a > 2d for complexes, and the
orders of 2a > 1a, 2b > 1b, 2c > 1c, 2d > 1d, which indicate the
abilities of displacement of EtBr from EtBr–DNA systems by
compounds and the binding affinities between compounds and
DNA. However, the orders are not slightly in good agreement with
the viscosity titration and UV–Vis spectroscopy study results. Here,
the phenolic hydroxy groups that can bind to nucleotides or/and
the sugar–phosphate backbone of DNA through hydrogen bonds
may play certain roles in the EtBr–DNA quenching tests. However,
the other weak interactions such as hydrophobic force, Van der
Waals force and electrostatic force (pH at 7.20) may not be
excluded. In other words, the interaction mechanism is not only
3.3.3. EtBr–DNA quenching assay
The fluorescence emission intensity of EtBr–DNA system
decreased dramatically upon the increasing amounts of every
ligand and Dy(III) complex. Stern–Volmer equation was used to
determine the fluorescent quenching mechanism [31]. Plots of Fo/F
versus [Q] are shown in Fig. 3 and the quenching data collected and
calculated from the good linear relationship when P < 0.05 are
listed in Table 2. As shown, the data of KSV are 1.405–
3.016 ꢃ 104 Mꢀ1 for ligands and 3.547–13.06 ꢃ 104 Mꢀ1 for Dy(III)
complexes, accordingly, the data of Kq calculated are 0.7806–
1.676 ꢃ 1013 Mꢀ1 sꢀ1 for ligands and 1.971–7.256 ꢃ 1013
M
ꢀ1 sꢀ1 for
Dy(III) complexes, respectively, when the value of so is taken as
1.8 ꢃ 10ꢀ9 s [31]. All of the current values of Kq for ligands and
Dy(III) complexes are much greater than that of Kq(max)

Y.-c. Liu, Z.-y. Yang / European Journal of Medicinal Chemistry 44 (2009) 5080–5089
5085
Table 2
ꢁ
Parameters of Kb, KSV, Kq, CF50, IC50 (OHꢁ and Oꢀ2 ) for ligands and the Dy(III) complexes.
ꢁ
Compounds Kb ꢃ 106 Mꢀ1
1/na
KSV ꢃ 104
M
ꢀ1 (R)
Kq ꢃ 1013
M
ꢀ1 sꢀ1 CF50 ꢃ 10ꢀ5 M
IC50
(
m
M) for OHꢁ (R)
IC50 (m
M) for Oꢀ2 (R)
b
c
c
(Ccompound/CDNA, nucleotides
)
1a
1b
1c
1d
2a
2b
2c
2d
0.2148 ꢅ 0.0205
0.9295 ꢅ 0.1315
0.7599 ꢅ 0.0867
0.1329 ꢅ 0.0180
5.687 ꢅ 1.084
0.081 2.086 ꢅ 0.014 (0.9997) 1.159
0.28
4.818 (12.05)
3.895 (9.738)
3.229 (8.073)
6.630 (16.58)
2.200 (5.500)
1.061 (2.653)
0.8465 (2.116)
2.832 (7.080)
14.66 ꢅ 0.495 (0.9937)
7.716 ꢅ 0.230 (0.9940)
11.38 ꢅ 0.441 (0.9902)
76.10 ꢅ 0.372 (0.9909)
13.80 ꢅ 0.303 (0.9914)
5.815 ꢅ 0.133 (0.9925)
4.687 ꢅ 0.086 (0.9948)
18.75 ꢅ 0.423 (0.9795)
6.831 ꢅ 0.219 (0.9954)
4.308 ꢅ 0.174 (0.9892)
4.096 ꢅ 0.112 (0.9955)
5.131 ꢅ 0.258 (0.9838)
7.971 ꢅ 0.517 (0.9768)
29.20 ꢅ 1.54 (0.9930)
16.17 ꢅ 1.37 (0.9663)
5.646 ꢅ 0.437 (0.9525)
2.352 ꢅ 0.018 (0.9998) 1.307
0.066 3.016 ꢅ 0.027 (0.9997) 1.676
0.092 1.405 ꢅ 0.013 (0.9995) 0.7806
0.14
0.15
0.18
0.35
4.588 ꢅ 0.083 (0.9990) 2.549
9.697 ꢅ 0.232 (0.9977) 5.387
13.06 ꢅ 0.211 (0.9990) 7.256
3.547 ꢅ 0.021 (0.9957) 1.971
13.41 ꢅ 4.760
26.94 ꢅ 3.540
0.5783 ꢅ 0.0959
a
The data of 1/n represent moles of compound/mol of base pair of DNA.
b
c
CF50 represents the molar concentration of the tested compound that causes a 50% loss in the fluorescence intensity of EtBr–DNA system.
IC50 value was calculated from regression line of the log of the tested compound concentration versus the scavenging effect (%) of the compound. R represents the linear
correlation coefficient.
determined by complex formation but also by certain weak inter-
actions [47].
which are consistent with the orders of EtBr–DNA quenching
results. Although it is a weaker binding for complex 2d to DNA, they
are stronger bindings for complexes 2c, 2b and 2a to DNA in
comparison with the classical intercalator (EtBr–DNA,
Kb ¼ 3.0 ꢃ 106 Mꢀ1 in 5 mmol Tris–HCl/50 mmol NaCl buffer,
pH ¼ 7.2), indicating that these three Dy(III) complexes can bind to
DNA effectively [53,54].
More importantly, DNA intercalators have been used extensively
as antitumor, antineoplastic, antimalarial, antibiotic, and antifungal
agents [31]. There is a criterion for screening out antitumor drugs
from others by EtBr–DNA fluorescent tracer method, i.e.,
a compound can be used as potential antitumor drug if it can cause
a 50% loss of EtBr–DNA fluorescence intensity by fluorescent
titrations before the molar concentration ratio of the compound to
DNA (nucleotides) does not overrun 100:1 [48]. CF50 value is
introduced to denote the molar concentration of a compound that
causes a 50% loss in the fluorescence intensity of EtBr–DNA system.
According to the data of CF50 and the molar ratios of compounds to
DNA shown in Table 2, it is interesting that at CF50, all the molar
concentration ratios of the investigated compounds to DNA are
largely under 100:1, indicating that all these ligands and Dy(III)
complexes can be used as potential antitumor drugs and the anti-
tumor activities of Dy(III) complexes are more excellent than those
of ligands. However, their pharmacodynamical and pharmacolog-
ical properties should be further studied in vivo.
3.4. Antioxidation properties
3.4.1. Hydroxyl radical scavenging activity
Fig. 4A and B shows the plots of hydroxyl radical scavenging
effect (%) for ligands and Dy(III) complexes, respectively, which are
concentration-dependant. As shown in Table 2, the values of IC50 of
ligands for OHꢁ are 7.716 ꢅ 0.230–76.10 ꢅ 0.372
mM with the order
of 1b < 1c < 1a < 1d, while the values of IC50 of Dy(III) complexes
for OHꢁ are 5.815 ꢅ 0.133–18.75 ꢅ 0.423
mM with the order of
2c < 2b < 2a < 2d. These orders of IC50 are opposite to the abilities
of scavenging effects for OHꢁ. It is marked that the hydroxyl radical
scavenging effects of Dy(III) complexes are much higher than those
of their ligands, possibly in that the larger conjugated metal
complexes can react with HOꢁ to form larger stable macromolecular
radicals than ligands [55]. Moreover, ligands 1b, 1c and their Dy(III)
complexes show higher abilities of scavenging effects for OHꢁ than
other ligands and Dy(III) complexes, possibly due to the key roles of
functional groups, –OH, which can react with HOꢁ to form stable
macromolecular radicals by the typical H-abstraction reaction [55].
Furthermore, for OHꢁ, there are two types of antioxidation mech-
anisms, in which one presents suppression of the generation of OHꢁ,
and another presents scavenging of the OHꢁ generated [55]. OHꢁ
production, detected by ethylene formation from methional, has
been investigated in plasma, lymph and synovial fluid in the
previous study [56]. In the presence of iron–EDTA as a catalyst,
addition of either H2O2 or xanthine and xanthine oxidase give rise
to OHꢁ formation that in most cases is not superoxide-dependent. In
the absence of iron–EDTA, the reaction is hardly detectable, the rate
3.3.4. Fluorescence spectroscopy study
When exited at lex ¼ 321–325 nm, ligands showed the fluores-
cence maximum wavelengths at lem ¼ 429–444 nm, and exited at
lex ¼ 333–334 nm the Dy(III) complexes showed the fluorescence
maximum wavelengths at lem ¼ 426–433 nm, respectively. Upon
additions of DNA, the fluorescence emission intensity of every
investigated compound grew steadily. Although the emission
enhancement cannot be regarded as a rigid criterion for binding
mode, it is related to the extent to which the compound gets into
a hydrophobic environment inside DNA and avoids the effect of
solvent water molecules. To compare quantitatively the affinities of
these compounds bound to DNA, the intrinsic binding constants Kb
can be obtained by the fluorescence titration methods and
Scatchard equation [34,49]. Scatchard plot should be a straight line
for a simple binding reaction [50]. Because of the significant
neighbour exclusion property of DNA binding to intercalating
agents, the Scatchard plot of r/Cf versus r usually presents a devia-
tion from linearity [51]. As shown in Fig. S1, every plot of r/Cf versus
r for ligands and Dy(III) complexes shows a deviation from linearity,
so the binding constant was obtained by McGhee and von Hippel
model [51,52]. The data of binding constants (Kb) and the moles of
compound bound per mol of base pair of DNA (1/n) are shown in
Table 2. It is clear that the data of binding constants (Kb) present the
order of 1b > 1c > 1a > 1d for ligands, which is slightly different
from the order of EtBr–DNA quenching result. But the data of
binding constants (Kb) present the orders of 2a > 1a, 2b > 1b,
2c > 1c, 2d > 1d, and the order of 2c > 2b > 2a > 2d for complexes,
being less than 5% of that observed with 1 mM iron–EDTA added. In
the present study, the chelation between phenolic hydroxyl group
and carbonyl group of 2-hydroxybenzoylhydrazine side chain for
ligand 1b with free Fe2þ in iron–EDTA reaction system may make
the concentration of free Fe2þ much lower so that the catalysis
becomes very poor and the OHꢁ formation has been suppressed,
thus, the inhibitive effect of 1b detected for OHꢁ is higher than those
of other ligands. However, after formation of Dy(III) complex, the
chelation between 1b and free Fe2þ may be destroyed with the
formation of intramolecular hydrogen bonds for 2b, so the hydroxyl
radical scavenging effect (%) of 2b is slightly lower than that of
complex 2c.

5086
Y.-c. Liu, Z.-y. Yang / European Journal of Medicinal Chemistry 44 (2009) 5080–5089
100
80
60
40
20
0
A
C
60
40
1d
1c
1b
1a
1d
20
1c
1b
1a
0
0
10
20
30
40
0
2
4
6
8
[Q] / µM
[Q] / µM
70
60
50
40
30
20
10
0
B
D
60
40
20
0
2d
2d
2c
2b
2a
2c
2b
2a
0
2
4
6
8
0
2
4
6
8
[Q] / µM
[Q] / µM
Fig. 4. Plots of antioxidation properties for ligands and Dy(III) complexes. (A) and (B) represent the hydroxyl radical scavenging effect (%) for ligands and Dy(III) complexes,
respectively. (C) and (D) represent the superoxide radical scavenging effect (%) for ligands and Dy(III) complexes, respectively.
3.4.2. Superoxide radical scavenging activity
leading to a central four-membered (DyO)2-ring. It is the key role of
enolization and deprotonation that the dimeric centronucleus of
every Dy(III) complex is of neutral charge, which will afford an
efficient route for investigators to well design favorable molecules.
In addition, all the ligands and Dy(III) complexes can bind to CT-DNA
Fig. 4C and D shows the plots of superoxide radical scavenging
effects (%) for ligands and Dy(III) complexes, respectively, which are
also concentration-dependant, but both the plots slightly blend
ꢁ
together. As shown in Table 2, the values of IC50 of ligands for O2ꢀ
are 4.096 ꢅ 0.112–6.831 ꢅ0.219
m
M with no significant difference
through intercalations with the binding constants at 105–107 Mꢀ1
,
ꢁ
from each other, but the values of IC50 of Dy(III) complexes for O2ꢀ
but Dy(III) complexes present stronger affinities to DNA than
ligands. 2a, 2b and 2c complexes can bind to DNA effectively. All the
ligands and Dy(III) complexes can be used as potential anticancer
drugs but the antitumor activities of Dy(III) complexes are more
excellent than those of ligands. However, their pharmacodynamical
and pharmacological properties should be further studied in vivo.
On the other hand, all the ligands and Dy(III) complexes have
strong abilities of antioxidation but Dy(III) complexes show
stronger scavenging effects for OHꢁ than ligands. Dy(III) complexes
or ligands containing active phenolic hydroxy groups present
stronger abilities of scavenging effects for OHꢁ than others, but
are 5.646 ꢅ 0.437–29.20 ꢅ 1.54
mM with a notably different order of
2d < 2a < 2c < 2b. These results suggest that complex containing
N-heteroaromatic substituent shows stronger scavenging effects
ꢁ
for Oꢀ2 , and that there are different mechanisms between scav-
enging and inhibiting OHꢁ and O2ꢀ , which should be further studied.
ꢁ
It is reported that the value of IC50 of ascorbic acid (Vc), a stan-
dard agent for non-enzymatic reaction, for OHꢁ is 1.537 mg cmꢀ3
(8.727 mmol), and the scavenging effect of Vc for Oꢀ2 ꢁ is only 25% at
1.75 mg cmꢀ3 (9.94 mmol) in vitro [57]. It is pronounced that all the
ligands and the Dy(III) complexes investigated here have much
stronger scavenging abilities for OHꢁ and O2ꢀ than ascorbic acid
complex containing N-heteroaromatic substituent shows stronger
ꢁ
ꢁ
(Vc). Endowed with antioxidative properties, these DNA binders
may be effective inhibitors of the formation of a DNA/TBP complex
topoisomerases [23–25].
scavenging effects for O2ꢀ
.
5. Experimental protocols
4. Conclusion
5.1. Materials
The Dy(III) complexes are prepared from Dy(NO3)3$6H2O and
Schiff-base ligands derived from 8-hydroxyquinoline-2-carbox-
aldehyde with four aroylhydrazines, respectively. X-ray crystals and
other structural analyses show that Dy(III) and every ligand can
form a binuclear Dy(III) complex with a 1:1 metal-to-ligand stoi-
chiometry and nine-coordination at the Dy(III) center. Every ligand
acts as a dibasic tetradentate ligand, binding to Dy(III) through the
phenolate oxygen atom, nitrogen atom of quinolinato unit and the
C]N group, ꢀO–C]N– group (enolized and deprotonated from
O]C–NH– group) of the aroylhydrazine side chain. Dimerization of
this monomeric unit occurs through the phenolate oxygen atoms
Calf thymus DNA (CT-DNA) and ethidium bromide (EtBr) were
obtained from Sigma–Aldrich Biotech. Co., Ltd. 8-Hydroxyquino-
line-2-carboxaldehyde was obtained form J&K Chemical Co., Ltd. All
the stock solutions (1.0 mmol) of the investigated compounds were
prepared by dissolving the powder materials into appropriate
amounts of DMF solutions, respectively. Deionized double distilled
water and analytical grade reagents were used throughout.
CT-DNA stock solution was prepared by dissolving the solid
material in 5 mmol Tris–HCl buffer (pH 7.20) containing 50 mmol
NaCl. Then, the solution was kept over 48 h at 4 ꢂC. The resulting
somewhat viscous solution was clear and particle-free. The solution

Y.-c. Liu, Z.-y. Yang / European Journal of Medicinal Chemistry 44 (2009) 5080–5089
5087
of CT-DNA in Tris–HCl buffer gave a ratio of UV–Vis absorbance at
r ¼ Cb=CDNA
(5)
260–280 nm of about 1.8–1.9, indicating that the CT-DNA was
sufficiently free of protein. The CT-DNA concentration in terms of
base pair Lꢀ1 was determined spectrophotometrically by employ-
where Ct is the total molar concentration of compound; Cb is the
molar concentration of compound bound for DNA; F is the observed
fluorescence emission intensity at a given DNA concentration CDNA
(nucleotides); Fo is the fluorescence emission intensity in the
absence of DNA; and Fmax is the maximum fluorescence emission
intensity of the compound totally bound for DNA at a titration end
point.
ing an extinction coefficient of
e
¼ 13200 Mꢀ1 cmꢀ1 (base pair)ꢀ1 at
260 nm. The CT-DNA concentration in terms of nucleotide Lꢀ1 was
also determined spectrophotometrically by employing an extinc-
tion coefficient of 6600 Mꢀ1 cmꢀ1 (nucleotide)ꢀ1 at 260 nm [58].
The stock solution was stored at ꢀ20 ꢂC until it was used. Working
standard solution of CT-DNA was obtained by appropriate dilution
of the stock solution in 5 mmol Tris–HCl buffer (pH 7.20) containing
50 mmol NaCl. EtBr was dissolved in 5 mmol Tris–HCl buffer (pH
7.20) and its concentration was determined assuming a molar
extinction coefficient of 5600 L molꢀ1 cmꢀ1 at 480 nm [31].
The binding constants were also obtained by McGhee and von
Hippel model [51,52]:
ꢀ
ꢁ
nꢀ1
r
Cf
1 ꢀ nr
¼ Kbð1 ꢀ nrÞ
(6)
1 ꢀ ðn ꢀ 1Þr
where Kb is the intrinsic binding constant and n is the exclusion
parameter in DNA base pairs. The experimental parameters Kb and
n were adjusted to produce curves that gave, by inspection, the
most satisfactory fits to the experimental data.
5.2. Methods
The melting points of the compounds were determined on an
XT4-100X microscopic melting point apparatus (Beijing, China).
Elemental analyses of C, N and H were carried out on an
Elemental Vario EL analyzer. The metal ion content was deter-
mined by complexo-metric titration with EDTA after destruction
of the complex in the conventional manner. The IR spectra were
recorded on a Nicolet Nexus 670 FT-IR spectrometer using KBr
disc in the 4000–400 cmꢀ1 region. 1H NMR spectra were recorded
on a Bruker Avance DRX 200-MHz spectrometer with tetrame-
thylsilane (TMS) as an internal standard. ESI-MS (ESI-Trap/Mass)
EtBr–DNA quenching assay was performed as reported in
a literature with a slight amendment [59]. DNA (4.0
tides) solution was added incrementally to 0.32 M EtBr solution,
until the rise in the fluorescence (lex ¼ 496 nm, lem ¼ 596 nm)
attained saturation. Then, small aliquots of concentrated
mM, nucleo-
m
a
compound solutions (1.0 mmol) were added till the drop in fluo-
rescence intensity (lex ¼ 525 nm, lem ¼ 587 nm) reached a constant
value. The experiments were measured after 5 min at a constant
room temperature, 298 K. Stern–Volmer equation was used to
determine the fluorescent quenching mechanisms [31]:
spectra were recorded on
spectrophotometer.
a
Bruker esquire6000 mass
Viscosity titration experiments were carried on an Ubbelodhe
Fo=F ¼ 1 þ Kqso½Qꢆ ¼ 1 þ KSV½Qꢆ
where Fo and F are the fluorescence intensities in the absence and in
the presence of a compound at [Q] concentration, respectively; KSV
is the Stern–Volmer dynamic quenching constant; Kq is the
(7)
viscometer in
a
thermostated water-bath maintained at
25.00 ꢅ 0.01 ꢂC. Titrations were performed for an investigated
compound that was introduced into DNA solution (50 mM, bps)
1/3
present in the viscometer. Data were presented as (
h/
ho)
versus
quenching rate constant of bimolecular diffusion collision;
lifetime of free EtBr.
so is the
the ratio of the compound to DNA, where is the viscosity of DNA
h
in the presence of the compound corrected from the solvent effect,
and ho is the viscosity of DNA alone. Relative viscosities for DNA in
either the presence or absence of compound were calculated from
the following relation:
The hydroxyl radicals in aqueous media were generated through
the Fenton-type reaction [56,60]. The 5 cm3 reaction mixtures
contained 2.0 cm3 of 100 mmol phosphate buffer (pH ¼ 7.4),
1.0 cm3 of 0.10 mmol aqueous safranin, 1 cm3 of 1.0 mmol aqueous
EDTA–Fe(II), 1 cm3 of 3% aqueous H2O2, and a series of quantita-
tively microadding solutions of the tested compound. The sample
without the tested compound was used as the control. The reaction
mixtures were incubated at 37 ꢂC for 60 min in a water-bath.
Absorbance at 520 nm was measured and the solvent effect was
corrected throughout. The scavenging effect for OHꢁ was calculated
from the following expression:
h
¼ ðt ꢀ toÞ=to
(1)
where t is the observed flow time of the DNA containing solution,
and to is the flow time of buffer [31,33].
Ultraviolet–visible (UV–Vis) spectra were obtained using a Per-
kinElmer Lambda UV–Vis spectrophotometer.
Fluorescence spectra were recorded using RF-5301PC spectro-
fluorophotometer (Shimadzu, Japan) with a 1 cm quartz cell. Both
the excitation and emission bandwidths were 10 nm. All the
experiments were measured after 5 min at a constant room
temperature, 298 K. The intrinsic binding constants Kb could be
obtained by the fluorescence titration methods and Scatchard
equation [49]:
Asample ꢀ Ablank
Acontrol ꢀ Ablank
Scavenging effect ð%Þ ¼
ꢃ 100
(8)
where Asample is the absorbance of the sample in the presence of the
tested compound, Ablank is the absorbance of the blank in the
absence of the tested compound and Acontrol is the absorbance in
the absence of the tested compound and EDTA–Fe(II) [34,61].
The superoxide radicals were produced by the MET–VitB2–NBT
system [34,61]. The solutions of MET (methionine), VitB2 (vitamin
B2), and NBT (nitroblue tetrazolium) were prepared with deionized
double distilled water at avoiding light, respectively. The 5 cm3
reaction mixtures contained 2.5 cm3 of 100 mmol phosphate buffer
(pH 7.8), 1.0 cm3 of 50 mmol MET, 1.0 cm3 of 0.23 mmol NBT,
r=Cf ¼ nKb ꢀ rKb
(2)
where r is the moles of compound bound per mole nucleotides of
DNA; Cf is the molar concentration of free compound; n is the
number of binding sites or the maximum number of compound
bound per nucleotide, and Kb is the association or binding constant.
Cf and r could be calculated according to the following equations
[34]:
0.50 cm3 of 33
mM VitB2, and a series of quantitatively microadding
solutions of the tested compound. After incubated at 30 ꢂC for
10 min in a water-bath and then illuminated with a fluorescent
lamp (4000 lux), the absorbance of the sample was measured at
Cf ¼ Ct ꢀ Cb
(3)
(4)
Cb ¼ CtðF ꢀ FÞo=ðFmax ꢀ FoÞ

5088
Y.-c. Liu, Z.-y. Yang / European Journal of Medicinal Chemistry 44 (2009) 5080–5089
560 nm and the solvent effect was corrected throughout. The
sample reaction mixtures without the tested compound were used
as the control. The scavenging effect for Oꢀ2 ꢁ was calculated from the
following expression:
isonicotinylhydrazine as the preparation of liþgand 1a. Yield: 71.0%;
m.p. ¼162–164 ꢂC; ESI-MS m/z 293.2 [M þ H] ; 1H NMR (DMSO-d6,
200 MHz, TMS)
d: 8.813 (d, J ¼ 5.2 Hz, 2 H, 17,19-CH), 8.657 (s, 1 H,
11-CH]N), 8.372 (d, J ¼ 8.8 Hz, 1 H, 4-CH), 8.129 (d, J ¼ 8.8 Hz, 1 H,
3-CH), 7.861 (d, J ¼ 5.2 Hz, 2 H, 16,20-CH), 7.528–7.409 (m, 2 H, 5,6-
CH), 7.148 (d, J ¼ 6.4 Hz, 1 H, 7-CH); IR (KBr): (cmꢀ1): 3576, 3193,
Ao ꢀ Ai
Scavenging effect ð%Þ ¼
ꢃ 100
(9)
Ao
1663, 1613, 1557, 1271; UV–Vis (DMF/H2O): lmax
(e
) ¼ 290 (28,600),
325 nm (17,800 Mꢀ1 cmꢀ1).
where Ai is the absorbance in the presence of the tested compound,
Ao is the absorbance in the absence of the tested compound.
The data for antioxidation presented as means ꢅ SD of three
determinations and followed by Student’s t-test. Differences were
considered to be statistically significant if P < 0.05. IC50 value was
introduced to denote the molar concentration of the tested
compound which caused a 50% inhibitory or scavenging effect on
hydroxyl radicals or superoxide radicals.
5.4. Preparation of the Dy(III) complexes
Complex 2a was prepared by refluxing and stirring equimolar
amounts of a 40 cm3 methanol solution of ligand 1a (0.058 g,
0.2 mmol) and Dy(NO3)$6H2O on a water-bath. After refluxed for
30 min, triethylamine (0.020 g, 0.2 mmol) was added into the
reaction mixtures drop wise to deprotonate the phenolic hydroxyl
substituent of 8-hydroxyquinolinato unit. Then, the mixtures were
refluxed and stirred continuously for 8 h. After cooling to room
temperature, the precipitate was centrifugalized, washed with
methanol and dried in vacuum over 48 h to give an orange powder.
Similarly, 2b, 2c and 2d complexes were prepared from equimolar
amounts of Dy(NO3)$6H2O and 1b, 1c and 1d, respectively.
5.3. Preparation of ligands
5.3.1. 8-hydroxyquinoline-2-carboxaldehyde-(benzoyl)hydrazone
(1a)
Ligand 1a was prepared by refluxing and stirring a 10 cm3
ethanol solution of 8-hydroxyquinoline-2-carboxaldehyde (0.519 g,
3 mmol) and a 10 cm3 90% ethanol aqueous solution of benzoyl-
hydrazine (0.408 g, 3 mmol) for 8 h. After cooling to room
temperature, the precipitate was filtered, recrystallized from 80%
methanol aqueous solution and dried in vacuum over 48 h to give
a pale yellow powder. Yield: 74.7% (0.656 g); m.p. ¼ 221 ꢂC; ESI-MS
5.4.1. Complex 2a
Yield: 88.3% (0.097 g); anal. calcd. for C34H30N8O14Dy2: C 37.09,
H 2.73, N 10.18, Dy 29.54; found: C 37.25, H 2.72, N 10.11, Dy 29.44;
ESI-MS m/z 1321.2 [M þ H]þ (DMF solution); IR (KBr): (cmꢀ1): 3399,
1615, 1552, 1494, 1310, 1104, 1068, 936, 841, 740, 613, 563, 492; UV–
m/z 292.1 [M þ H]þ; 1H NMR (DMSO-d6, 200 MHz, TMS)
d
: 8.637 (s,
Vis (DMF/H2O): lmax
L
(e
) ¼ 326 (41,600), 373 nm (32,100 Mꢀ1 cmꢀ1).
1 H, 11-CH]N), 8.343 (d, J ¼ 8.8 Hz, 1 H, 4-CH), 8.119 (d, J ¼ 8.8 Hz, 1
H, 3-CH), 7.936 (d, J ¼ 6.4 Hz, 2 H, 16,20-CH), 7.630–7.507 (m, 3 H,
17,18,19-CH), 7.467–7.387 (m, 2 H, 5,6-CH), 7.131 (d, 1 H, J ¼ 5.0 Hz,
7-CH); IR (KBr): (cmꢀ1): 3359, 3318, 1682, 1602, 1546, 1267; UV–Vis
m (DMF) ¼ 38.8 cm2
U .
ꢀ1 molꢀ1
5.4.2. Complex 2b
Yield: 87.1%; anal. calcd. for C34H30N8O16Dy2: C 36.04, H 2.65, N
9.89, Dy 28.71; found: C 36.15, H 2.66, N 9.86, Dy 28.82; ESI-MS m/z
1353.2 [M þ H]þ (DMF solution); IR (KBr): (cmꢀ1): 3406, 3216,1599,
1543, 1491,1309,1256, 1104,1058, 922, 839, 761, 636, 585, 489; UV–
(DMF/H2O): lmax
(
3
) ¼ 295 (35,500), 323 nm (21100 Mꢀ1 cmꢀ1).
5.3.2. 8-hydroxyquinoline-2-carboxaldehyde-(20-hydroxybenzoyl)
hydrazone (1b)
Vis (DMF/H2O): lmax
L
(e
) ¼ 327 (29,400), 377 nm (29,900 Mꢀ1 cmꢀ1).
Ligand 1b, a yellow precipitate, was obtained from equimolar
amounts of 8-hydroxyquinoline-2-carboxaldehyde and 2-hydrox-
ybenzoylhydrazine as the preparation of ligþand 1a. Yield: 81.0%;
m.p. ¼ 245–247 ꢂC; ESI-MS m/z 308.2 [M þ H] ; 1H NMR (DMSO-d6,
m (DMF) ¼ 36.0 cm2
U .
ꢀ1 molꢀ1
5.4.3. Complex 2c
Yield: 92.8%; anal. calcd. for C34H30N8O16Dy2: C 36.04, H 2.65, N
9.89, Dy 28.71; found: C 36.17, H 2.65, N 9.87, Dy 28.62; ESI-MS m/z
1353.2 [M þ H]þ (DMF solution); IR (KBr): (cmꢀ1): 3405, 3193,1598,
1548,1494,1336,1289,1104,1070, 926, 843, 741, 642, 564, 490; UV–
200 MHz, TMS)
d
: 8.621 (s, 1 H, 11-CH]N), 8.356 (d, J ¼ 8.6 Hz, 1 H,
4-CH), 8.113 (d, J ¼ 8.6 Hz, 1 H, 3-CH), 7.871 (d, 1 H, J ¼ 7.8 Hz, 20-
CH), 7.469–7.395 (m, 3 H, 5,6,18-CH), 7.133 (d, J ¼ 7.0 Hz, 1 H, 7-CH),
7.018–6.943 (m, 2 H, 17,19-CH); IR (KBr): (cmꢀ1): 3464, 3250, 1643,
Vis (DMF/H2O): lmax
L
(e
) ¼ 332 (42,700), 379 nm (37,300 Mꢀ1 cmꢀ1).
1607, 1532, 1288; UV–Vis (DMF/H2O): lmax
(
e) ¼ 294 (31600),
m (DMF) ¼ 38.9 cm2
U .
ꢀ1 molꢀ1
329 nm (23,600 Mꢀ1 cmꢀ1).
5.4.4. Complex 2d
5.3.3. 8-hydroxyquinoline-2-carboxaldehyde-(40-hydroxybenzoyl)
hydrazone (1c)
Yield: 85.8%; anal. calcd. for C32H28N10O14Dy2: C 34.85, H 2.54, N
12.70, Dy 29.49; found: C 35.01, H 2.55, N 12.64, Dy 29.42; ESI-MS
m/z 1323.2 [M þ H]þ (DMF solution); IR (KBr): (cmꢀ1): 3401, 1634,
1592, 1550, 1494, 1314, 1102, 1059, 935, 843, 740, 612, 566, 491; UV–
Ligand 1c, a pale yellow precipitate, was obtained from equi-
molar amounts of 8-hydroxyquinoline-2-carboxaldehyde and 4-
hydroxybenzoylhydrazine as the preparation of ligand 1a. Yield:
81.0%; m.p. ¼ 279–280 ꢂC; ESI-MS m/z 308.2 [M þ H]þ; 1H NMR
Vis (DMF/H2O): lmax
(
e
) ¼ 327 (44,900), 372 nm (36,600 Mꢀ1 cmꢀ1).
L
m (DMF) ¼ 34.9 cm2
U .
ꢀ1 molꢀ1
(DMSO-d6, 200 MHz, TMS) d: 8.594 (s, 1 H, 11-CH]N), 8.329 (d,
J ¼ 8.4 Hz, 1 H, 4-CH), 8.088 (d, J ¼ 8.4 Hz, 1 H, 3-CH), 7.834 (d, 2 H,
J ¼ 10.4 Hz, 16,20-CH), 7.493–7.379 (m, 2 H, 5,6-CH), 7.124 (d,
J ¼ 6.8 Hz, 1 H, 7-CH), 6.900 (d, J ¼ 10.4 Hz, 2 H, 17,19-CH); IR (KBr):
(cmꢀ1): 3320, 3139, 1660, 1632, 1581, 1277; UV–Vis (DMF/H2O):
5.5. Determination of crystal structures
X-ray diffraction data for a crystal were performed with
graphite-monochromated Mo K
APEX area-detector diffractometer and collected by the
a
radiation (0.71073 Å) on a Bruker
–2 scan
lmax
(3
) ¼ 300 (31,800), 326 nm (22,400 Mꢀ1 cmꢀ1).
u
q
technique at 273(2) K. The crystal structures were solved by direct
methods. All non-hydrogen atoms were refined anisotropically by
full-matrix least-squares methods on F2. A partial structure was
obtained by direct methods and the remaining non-hydrogen
atoms were located from difference maps. Hydrogen atoms were
5.3.4. 8-hydroxyquinoline-2-carboxaldehyde-(isonicotinyl)
hydrazone (1d)
Ligand 1d, a yellow precipitate, was also obtained from equi-
molar amounts of 8-hydroxyquinoline-2-carboxaldehyde and

Y.-c. Liu, Z.-y. Yang / European Journal of Medicinal Chemistry 44 (2009) 5080–5089
5089
[23] S.Y. Chiang, J. Welch, F.J. Rauscher, T.A. Beerman, Biochemistry 33 (1994) 7033.
[24] J.M. Woynarowski, M. Mchugh, R.D. Sigmund, T.A. Beerman, Mol. Pharmacol.
35 (1989) 177.
[25] A.Y. Chen, C. Yu, B. Gatto, L.F. Liu, Proc. Natl. Acad. Sci. U.S.A. 90 (1993) 8131.
[26] W.J. Geary, Coord. Chem. Rev. 7 (1971) 81.
located in geometrically defined positions and not refined. All
calculations were performed using the programs SHELXS-97 and
SHELXL-97 [62].
[27] M.M. Moawad, W.G. Hanna, J. Coord. Chem. 55 (2002) 439.
[28] T.M.A. Ismail, J. Coord. Chem. 58 (2005) 141.
Acknowledgement
[29] J.M. Ou-Yang, J. Inorg. Chem. 13 (1997) 315 (in Chinese).
[30] X.H. Lu, Q.Y. Lin, L.C. Kong, X.Q. He, Chem. Res. Appl. 18 (2006) 1380 (in
Chinese).
[31] D. Suh, J.B. Chaires, Bioorg. Med. Chem. 3 (1995) 723.
[32] R. Palchaudhuri, P.J. Hergenrother, Curr. Opin. Biotechnol. 18 (2007) 497.
[33] S. Satyanarayana, J.C. Dabrowiak, J.B. Chaires, Biochemistry 31 (1992) 9319.
[34] B.D. Wang, Z.Y. Yang, P. Crewdson, D.Q. Wang, J. Inorg. Biochem. 101
(2007) 1492.
The study was supported by the National Natural Science Foun-
dation of China (20475023) and Gansu NSF (3ZS 041-A25-016).
Appendix A. Supplementary data
CCDC 713869 (2a) and 713870 (2b) contain the supplementary
crystallographic data for this paper. These data can be obtained free
of charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif. Also can be found at, in the
online version, doi:10.1016/j.ejmech.2009.09.015.
[35] R.D. Snyder, Mutat. Res. 623 (2007) 72.
[36] J.K. Barton, A.T. Danishefsky, J.M. Goldberg, J. Am. Chem. Soc. 106 (1984) 2172.
[37] H.L. Lu, J.J. Liang, Z.Z. Zeng, P.X. Xi, X.H. Liu, F.J. Chen, Z.H. Xu, Transit. Met.
Chem. 32 (2007) 564.
[38] C. Behrens, N. Harrit, P.E. Nielsen, Bioconjugate Chem. 12 (2001) 1021.
[39] S. Frau, J. Bernadou, B. Meunier, Bioconjugate Chem. 8 (1997) 222.
[40] D.S. Sigman, A. Mazumder, D.M. Perrin, Chem. Rev. 93 (1993) 2295.
[41] Y. Wang, Z.Y. Yang, Q. Wang, Q.K. Cai, K.B. Yu, J. Org. Chem. 690 (2005) 4557.
[42] J.Z. Wu, B.H. Ye, L. Wang, L.N. Ji, J.Y. Zhou, R.H. Li, Z.Y. Zhou, J. Chem. Soc.,
Dalton Trans. 8 (1997) 1395.
[43] Y.C. Liu, Z.Y. Yang, J. Du, X.J. Yao, R.X. Lei, X.D. Zheng, J.N. Liu, H.S. Hu, H. Li,
Chem. Pharm. Bull. 56 (2008) 443.
[44] Y.C. Liu, Z.Y. Yang, J. Du, X.J. Yao, R.X. Lei, X.D. Zheng, J.N. Liu, H.S. Hu, H. Li,
Immunobiology 213 (2008) 651.
References
[1] L.H. Schmidt, Annu. Rev. Microbiol. 23 (1969) 427.
[2] A.A. El-Asmy, A.Z. El-Sonbati, A.A. Ba-Issa, M. Mounir, Transit. Met. Chem. 5
(1990) 222.
[3] D. Parker, R.S. Dickins, H. Puschmann, C. Crossland, J.A.K. Howard, Chem. Rev.
102 (2002) 1977.
[4] M. Albrecht, O. Osetska, R. Fro¨hlich, Dalton Trans. 23 (2005) 3757.
[5] J.Z. Li, G. Li, W.J. Yu, J. Chin. Rare Earth Soc. 18 (2001) 71 (in Chinese).
[6] Y.J. Ji, Z.H. Wang, J.L. Li, S.H. Peng, J. Health Toxicol. 8 (1994) 164 (in Chinese).
[7] S.X. Liu, L. Wang, Y.Y. Liu, E.B. Wang, J. Chin. Rare Earth Soc. 15 (1997) 59 (in
Chinese).
[8] R.B. Hunter, W. Walker, Nature 178 (1956) 47.
[9] H.X. Liu, X.T. Liu, J.F. Lu, R.C. Li, K. Wang, J. Beijing Med. Univ. 32 (2000) 203 (in
Chinese).
[10] D.M. Kramsch, A.J. Aspen, L.J. Rozler, Science 213 (1981) 1511.
[11] E.M. Hodnett, P.D. Mooney, J. Med. Chem. 13 (1970) 786.
[12] E.M. Hodnett, W.J. Dunn, J. Med. Chem. 15 (1972) 339.
[13] Y.B. Zeng, N. Yang, W.S. Liu, N. Tang, J. Inorg. Biochem. 97 (2003) 258.
[14] A.M. Pyle, T. Morii, J.K. Barton, J. Am. Chem. Soc. 112 (1990) 9432.
[15] J.K. Barton, J.M. Goldberg, C.V. Kumar, N.J. Turro, J. Am. Chem. Soc. 108 (1986)
2081.
[16] S. Mahadevan, M. Palaniandavar, Inorg. Chim. Acre 254 (1997) 291.
[17] S.J. Lippar, Acc. Chem. Res. 11 (1978) 211.
[18] S.M. Hech, Acc. Chem. Res. 19 (1986) 383.
[19] B.N. Ames, M.K. Shigenaga, T.M. Hagen, Proc. Natl. Acad. Sci. U.S.A. 90 (1993)
7915.
[20] A.A. Horton, S. Fairhurst, Crit. Rev. Toxicol. 18 (1987) 27.
[21] H.L. Wang, Z.Y. Yang, B.D. Wang, Transit. Met. Chem. 31 (2006) 470.
[22] S.F. Lo, V. Mulabagal, C.L. Chen, C.L. Kuo, H.S. Tsay, J. Agric. Food Chem. 52
(2004) 6916.
[45] L.A. Bagatolli, S.C. Kivatinitz, G.D. Fidelio, J. Pharm. Sci. 85 (1996) 1131.
[46] M.M. Yang, P. Yang, L.W. Zhang, Chin. Sci. Bull. 9 (1994) 31.
[47] A. Ayar, B. Mercimek, Process Biochem. 41 (2006) 1553.
[48] Z.L. Li, J.H. Chen, K.C. Zhang, M.L. Li, R.Q. Yu, Sci. China Ser B Chem. 11 (1991)
1193 (in Chinese).
[49] G. Scatchard, Ann. N.Y. Acad. Sci. 51 (1949) 660.
[50] L.M. Berezhkovskiy, I.V. Astafieva, C. Cardoso, Anal. Biochem. 308 (2002) 239.
[51] J.B. Chaires, N. Dattagupta, D.M. Crothers, Biochemistry 21 (1982) 3933.
[52] J.D. McGhee, P.H. von Hippel, J. Mol. Biol. 86 (1974) 469.
[53] M. Baldini, M. Belicchi-Ferrari, F. Bisceglie, G. Pelosi, S. Pinelli, P. Tarasconi,
Inorg. Chem. 42 (2003) 2049.
[54] Y. Wang, Z.Y. Yang, Z.N. Chen, Bioorg. Med. Chem. Lett. 18 (2008) 298.
[55] J.-I. Ueda, N. Saito, Y. Shimazu, T. Ozawa, Arch. Biochem. Biophys. 333
(1996) 377.
[56] C.C. Winterbourn, Biochem. J. 198 (1981) 125.
[57] R. Xing, H. Yu, S. Liu, W. Zhang, Q. Zhang, Z. Li, P. Li, Bioorg. Med. Chem. 13
(2005) 1387.
[58] F. Zsila, Z. Bika´di, M. Simonyi, Org. Biomol. Chem. 2 (2004) 2902.
[59] A.G. Krishna, D.V. Kumar, B.M. Khan, S.K. Rawal, K.N. Ganesh, Biochim. Bio-
phys. Acta 1381 (1998) 104.
[60] C.C. Winterbourn, Biochem. J. 182 (1979) 625.
[61] Z.Y. Guo, R.E. X, S. Liu, H.H. Yu, P.B. Wang, C.P. Li, P.C. Li, Bioorg. Med. Chem.
Lett. 15 (2005) 4600.
[62] G.M. Sheldrick, Acta Crystallogr. A 46 (1990) 467.