Synthesis, structural and spectroscopic characterization and biomimetic properties of new copper, manganese, zinc complexes: Identification of possible superoxide-dismutase mimics bearing hydroxyl radical generating/scavenging abilities

Article abstract of DOI:10.1016/j.jinorgbio.2010.03.013

A series of Cu(II), Zn(II) and Mn(II) coordination compounds has been synthesized by reaction of the corresponding metal salts and pyrazolyl-based ligands, i.e. the neutral 1-(2-(4-((2,2,2-tri(1H-pyrazol-1-yl)ethoxy)methyl)benzyloxy)-1,1-di(1H-pyrazol-1-yl)ethyl)-1H-pyrazole {p-C₆H₄[CH₂OCH₂C(pz)₃]₂, (L¹), and the anionic hydridotris(3-phenyl-5-methylpyrazolyl)borate (L²)^-, bis(pyrazolyl)acetate (L³) and bis(3,5-dimethylpyrazolyl)acetate (L⁴)^-: the species [L¹(CuCl₂)₂] (1), [L¹(Cu(OAc)₂)₂] (2), [L¹(Zn(OAc)₂)₂] (3), [(CuCl(L²)(Hpz^Ph,Me)] (4), [Mn(L³)₂]·2H₂O, (5), [ZnCl(L³)(imH)]·MeOH [CuCl(L⁴)(imH)]·2H₂O (7) have been obtained (Hpz^Ph,Me=3-phenyl-5-methylpyrazole, imH=imidazole). Complexes 1 and 4 have been structurally characterized, also using less conventional powder diffraction methods. The superoxide scavenging activity has been characterized by indirect assays including EPR analysis. All complexes exhibit superoxide scavenging activity with IC₅₀ in the μM range and they protect against the oxidative action of peroxynitrite in different ways. 1, 4 and 7 exert both an anti- and pro-oxidant effect depending on their concentration as evaluated by EPR and fluorescence methods. The pro-oxidative effects are absent in Zn(II) and Mn(II) complexes.

Full text of DOI:10.1016/j.jinorgbio.2010.03.013

Journal of Inorganic Biochemistry 104 (2010) 820–830
Contents lists available at ScienceDirect
Journal of Inorganic Biochemistry
journal homepage: www.elsevier.com/locate/jinorgbio
Synthesis, structural and spectroscopic characterization and biomimetic properties
of new copper, manganese, zinc complexes: Identification of possible
superoxide-dismutase mimics bearing hydroxyl radical
generating/scavenging abilities
Giulio Lupidi a, Fabio Marchetti b, Norberto Masciocchi c, Daniel L. Reger d, Sartaj Tabassum e, Paola Astolfi f,
Elisabetta Damiani g, Claudio Pettinari b,
⁎
a
Dipartimento di Biologia Molecolare Cellulare e Animale, Università degli Studi di Camerino, Via Madonna delle carceri, I-62032 Camerino, MC, Italy
Dipartimento di Scienze Chimiche, Università degli Studi di Camerino, Via S. Agostino 1, I-62032 Camerino, MC, Italy
Dipartimento di Scienze Chimiche e Ambientali, Università dell'Insubria, Via Valleggio 11- 22100 Como, Italy
Department of Chemistry and Biochemistry, University of South Carolina, Columbia, SC, 29208, USA
Department of Chemistry, Aligarh Muslim University, Aligarh, 202002, India
Dipartimento ISAC-Sezione Chimica, Università Politecnica delle Marche, I-60131 Ancona, Italy
Dipartimento di Biochimica, Biologia e Genetica, Università Politecnica delle Marche, I-60131 Ancona, Italy
b
c
d
e
f
g
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of Cu(II), Zn(II) and Mn(II) coordination compounds has been synthesized by reaction of the
corresponding metal salts and pyrazolyl-based ligands, i.e. the neutral 1-(2-(4-((2,2,2-tri(1H-pyrazol-1-yl)
ethoxy)methyl)benzyloxy)-1,1-di(1H-pyrazol-1-yl)ethyl)-1H-pyrazole {p-C6H4[CH2OCH2C(pz)3]2, (L1), and
the anionic hydridotris(3-phenyl-5-methylpyrazolyl)borate (L2)−, bis(pyrazolyl)acetate (L3) and bis(3,5-
Received 20 January 2010
Received in revised form 22 March 2010
Accepted 23 March 2010
Available online 7 April 2010
dimethylpyrazolyl)acetate (L4)−
:
the species [L1(CuCl2)2] (1), [L1(Cu(OAc)2)2] (2), [L1(Zn(OAc)2)2]
(3), [(CuCl(L2)(HpzPh,Me)] (4), [Mn(L3)2]·2H2O, (5), [ZnCl(L3)(imH)]·MeOH [CuCl(L4)(imH)]·2H2O (7) have
been obtained (HpzPh,Me=3-phenyl-5-methylpyrazole, imH=imidazole). Complexes 1 and 4 have been
structurally characterized, also using less conventional powder diffraction methods. The superoxide scavenging
activity has been characterized by indirect assays including EPR analysis. All complexes exhibit superoxide
scavenging activity with IC50 in the µM range and they protect against the oxidative action of peroxynitrite in
different ways. 1, 4 and 7 exert both an anti-and pro-oxidanteffectdepending on theirconcentrationas evaluated
by EPR and fluorescence methods. The pro-oxidative effects are absent in Zn(II) and Mn(II) complexes.
© 2010 Elsevier Inc. All rights reserved.
1. Introduction
augment endogenous cellular defence levels. In this context, the failure
of clinical attempts to use SOD, due to several reasons (instability,
The formation of reactive oxygen species (ROS) and the well known
deleterious consequences of their uncontrolled production, is the price
that respiring organisms have to pay during their life cycle. Hence, all
aerobic organisms have developed a host of defence mechanisms aimed
at minimizing, either directly or indirectly, the formation or propagation
of ROS. For example, one such line of defence is constituted by
antioxidant enzymes, such as the superoxide dismutases (SOD) which
catalytically scavenge the superoxide radical, an important agent of
oxygen toxicity [1,2]. However, under certain circumstances, the
imbalance between oxygen toxicity and antioxidant levels may call
forth the need for exogenously administered compounds, either
naturally occurring or completely synthetic molecules, which could
immunogenicity, bio-inaccessibility, high molecular weight) has lead to
the development of a number of low molecular weight SOD mimics
which could ideally overcome these limitations. These compounds all
bear a redox active metal centre, similar to the active site metals of the
native SODs, i.e. Cu, Fe, Zn or Mn complexes with ligands of different
chemical nature, capable of detoxifying superoxide radical [3–5].
Studies on the reactivity of low molecular weight complexes which
exhibit SOD like activity have attracted major attention for the
development of SOD mimics and for expanding the biomimetic
chemistry of transition metals complexes. The pharmaceutical use of
metal complexes has therefore excellent potential and a broad array of
medicinal applications have been investigated as summarized by
several recent reviews in this field [6–9]. Different problems should be
addressed during the use of SOD-mimics such as accumulation,
biodistribution and clearance of metal ions bound to the complex. For
this purpose it is necessary to consider the physiological responses of
⁎
Corresponding author. Fax: +39 0737637345.
E-mail address: claudio.pettinari@unicam.it (C. Pettinari).
0162-0134/$ – see front matter © 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.jinorgbio.2010.03.013

G. Lupidi et al. / Journal of Inorganic Biochemistry 104 (2010) 820–830
821
these candidate drugs with in vitro and in vivo studies before they enter
clinical trials. For the potential therapeutic use of metals in medicine,
several efforts have been made to obtain compounds with high catalytic
activity and different ligands have been used to increase their stability
under physiological conditions [6,8,10]. Important requirements for
SOD-like activity of the complexes appear to be a medium strength
donor power, flexibility of the ligands and coordination sites belong-
ing to N-heteroaromatic rings. Among the most successful metal-
complexes possessing SOD activity, dinuclear metal complexes [11,12]
have received much attention considering that in nature various
enzymes containing dinuclear metal active centres exist and are able
to bind oxygen and catalyze the transformation of reactive intermedi-
ates. As part of this ongoing search for new therapeutic antioxidant
agents, we have decided to start a systematic study on copper, zinc and
manganese complexes containing heterocyclic N-donor ligands, also
able to form dinuclear species, followed by investigations on their
superoxide scavenging activity [13–16].
Here we report on the synthesis of novel mono-, di and poly-
nuclear metal complexes based on different types of ligands
(Scheme 1), i.e. the neutral ligand 1-(2-(4-((2,2,2-tri(1H-pyrazol-1-
yl)ethoxy)methyl)benzyloxy)-1,1-di(1H-pyrazol-1-yl)ethyl)-1H-
pyrazole (L1), and the anionic ligands hydridotris(3-phenyl-5-
methylpyrazolyl)borate (L2)−, bis(pyrazolyl)acetate (L3)− and bis
(3,5-dimethylpyrazolyl)acetate (L4)− bearing either Cu(II) [17], Zn(II)
[18], Mn(II) [19], as active metal centres. Several zinc [20–23], copper
[24–29] and manganese derivatives [30–32] have already been
reported with ligands (L2)−, (L3)− and (L4)− [33,34]. The structural,
spectroscopic and biochemical properties of the obtained metal
complexes will be presented, including evaluation of SOD-like
activity, effects on H2O2 decomposition/peroxidase activity and on
peroxynitrite.
tested were dissolved in DMSO. Peroxynitrite was synthesized
according to the protocol reported by Uppu et al. [35] and stored at
−80 °C.
All the reactions and manipulations were performed in air. Solvent
evaporations were always carried out under vacuum using a rotary
evaporator. The samples for microanalysis were dried in vacuo to
constant weight (20 ° C, ca.0.1 Torr). Elemental analyses (C, H, N)
were performed in-house with a Fison Instrument 1108 CHNS-O
Elemental analyzer. IR spectra were recorded from 4000 to 200 cm−1
with a Perkin-Elmer Spectrum 100 FT-IR instrument. The intensity of
reported IR signals is defined as w=weak, m=medium and
s=strong. 1H and 13C{1H} NMR spectra were recorded on a 400
Mercury Plus instrument operating at room temperature (400 MHz
for 1H, 100 MHz for 13C). H and C chemical shifts (δ) are reported in
parts per million (ppm) from SiMe4 (1H and 13C calibration by internal
deuterium solvent lock). Peak multiplicities are abbreviated: singlet,
s; doublet, d; triplet, t; quartet, q; multiplet, m. Melting points are
uncorrected and were taken on an STMP3 Stuart scientific instrument
and on a capillary apparatus. The electrical conductivity measure-
ments (ΛM, reported as Ω−1 cm2 mol−1) of acetonitrile and water
solutions of the complexes were taken with a Crison CDTM 522
conductimeter at room temperature. Electrospray mass spectra were
obtained with a Series 1100 MSI detector HP spectrometer, using
MeOH as mobile phase. Solutions for electrospray ionization mass
spectrometry (ESI-MS) were prepared using reagent grade methanol
and/or acetonitrile and obtained data (masses and intensities) were
compared to those calculated by using the IsoPro isotopic abundance
simulator [36]. Peaks containing copper(II) and zinc ions are
identified as the centres of isotopic clusters. The magnetic suscept-
ibilities were measured at room temperature (22 °C) by the Gouy
method, with a Sherwood Scientific Magnetic Balance MSB-Auto,
using HgCo(NCS)4 as calibrant and corrected for diamagnetism with
the appropriate Pascal constants. The magnetic moments (in BM)
were calculated from the equation µeff =2.828 (χcmorr T)1/2. UV–visible
(UV/Vis) Spectra were recorded on a Shimadzu — 1700 UV–Vis
spectrometer and the data are reported as λmax/nm. Cyclic voltam-
metric studies were performed on a CH Instrument Electrochemical
Analyzer in a single compartmental cell with 0.4 M KNO3 as
supporting electrolyte. A three electrode configuration was used
comprising of a Pt wire as auxiliary electrode, platinum micro cylinder
as working electrode and Ag/AgCl as the reference electrode.
Electrochemical measurements were made under a dinitrogen
atmosphere. All electrochemical data were collected at 298 K and
are uncorrected for junctions potentials. The formal potentials, E° (or
voltammetric E1/2) were taken as the average of the anodic (Epa) and
cathodic peak potentials (Epc) obtained from the cyclic voltammetry.
2. Experimental section
2.1. Materials and methods
Hypoxanthine (HX), xanthine, xanthine oxidase (XO, from bovine
milk), superoxide dismutase (SOD, from bovine erythrocytes),
catalase (CAT, from bovine liver) cytocrome c, phenazine methosul-
fate (PMS), β-nicotinamide adenine dinucleotide (NADH), nitroblue-
di tetrazolium chloride (NBT) were obtained from Sigma Chemical Co.
(Milan, Italy) while all other chemicals and reagents including the
spin trap 5,5-dimethyl-1-pyrroline-N-oxide (DMPO) were purchased
from Aldrich Chemical Co. (Milan, ITALY). DMPO was used as
purchased since it gave no EPR signals when scanned at 100 mM
concentration. Xanthine was dissolved in 1 µM NaOH. All compounds
2.2. Synthesis and characterisation of the ligands
2.2.1. 1-(2-(4-((2,2,2-Tris(pyrazol-1-yl)ethoxy)methyl)benzyloxy)-1,
1-di(pyrazol-1-yl)ethyl)-pyrazole (L1)
The ligand L1 was prepared by a slight modification of a previously
reported synthesis [37]. α,α'-Dibromo-p-xylene, p-(BrCH2)2C6H4
(2.64 g, 10 mmol) and tris-2,2,2-(1-pyrazolyl)ethanol, HOCH2C(pz)3
(4.88 g, 20 mmol) were dissolved in dry tetrahydrofuran (THF)
(125 ml). This solution was added drop wise to a suspension of NaH
(1.5 g) in dry THF (300 ml) under an inert atmosphere. The mixture was
stirred under reflux for 48 h and then allowed to cool at room
temperature. To this solution, water (200 ml) was added drop wise to
consume the excess NaH and dissolve the resulting NaBr and NaOH. The
THF mixture was extracted with diethyl ether (4×100 ml) and the
combined organic extracts were washed with 100 ml saturated NaHCO3
solution with 100 ml NaCl solution and finally with 100 ml water. The
organic layer was dried over anhydrous Na2SO4, filtered and the solvent
removed under vacuum to afford the desired compound as a white–
yellow powder (8.4 mmol, 4.95 g, 84%). M.p. 150–155 °C. Anal. calcd
Scheme 1.

822
G. Lupidi et al. / Journal of Inorganic Biochemistry 104 (2010) 820–830
for C30H30N12O2. C, 60.96; H, 5.07; N, 28.45. Found: C, 61.07; H, 4.98,
N, 28.70. IR (nujol, mull, cm−1): 3148w, 3133w, 3104w, 2944w,
2891w, 1516 m. 1H NMR (CDCl3): δ, 7.65, 7.38 (d,m, 3H, 3H, J=1.6 Hz,
3,5-H pz), 7.18 (m,2H, C6H4), 6.32 (d of d, 3H, J HH=2.5, 4-H pz), 5.07
(s, 2H, OCH2Ph), 4.44 (s, 2H, OCH2 (pz)3).
H, 5.36; N, 14.75%. Λm (MeCN, 10−4 M, 293 K): 22.3Ω−1 cm2 mol−1
.
ESI MS (CH3CN) (+): 704 [(Cu(L2)(HpzPh,Me)]+. µeff =1.81 BM. IR
(nujol mull, cm−1): 3215 (NH), 2526m (BH), 1565w, 1540m, 1503m,
524m, 495m, 478w, 438w, 388w, 318w, 304w, 280br (Cu–Cl).
2.3.5. [(Mn(L3)2]·2H2O, 5
2.2.2. Potassium hydridotris(3-phenyl-5-methylpyrazolyl)borate: K[L2]
The potassium salt of the ligand hydridotris(3-phenyl-5-methyl-
pyrazolyl)borate was synthesized according to the literature [38].
To a stirred solution of HL3 (0.193 g, 1.0 mmol) in methanol, a
methanol solution of MnCl2·6H2O was added (0.234 g, 1.0 mmol).
The reaction mixture was stirred for 6 h. An amorphous colorless
powder was obtained, washed with methanol and n-hexane and
shown to be compound 5. The compound is soluble in DMSO. Re-
crystallised from MeOH. M.p. 230 °C. Elem. Anal. Calcd. for
C16H18MnN8O6, C, 40.60; H, 3.83; N, 23.67; O, 20.28. Found: C,
40.40; H, 3.76; N, 23.23; O, 19.80%. Λm (MeCN, 10−4 M, 293 K):
2.5Ω−1 cm2 mol−1. µeff =5.96 BM. IR (nujol, cm−1): 3360br, 3114w
(CH), 1657m, 1626m, 1528m, 1513m, 1455m (C=O), 417m, 229w.
2.2.3. Bis(pyrazol-1-yl)acetic acid (HL3) and bis(3,5-dimethylpyrazol-1-yl)
acetic acid (HL4)
The ligands HL3 (HL3 =bis(pyrazol-1-yl)acetic acid) and HL4
(HL4 =bis(3,5-dimethylpyrazol-1-yl)acetic acid) were prepared
according to literature methods. Their analytical and spectroscopic
data correspond to those reported in the literature [39,40].
2.3. Syntheses and characterization of the complexes
2.3.6. [(Zn(OAc)(L3)(Him)]·MeOH, 6
To a stirred solution of HL3 (1 mmol, 0.248 g) in methanol, Zn
(OAc)2·H2O (2.0 mmol)) and imidazole (2.0 mmol), were added. The
reaction mixture was stirred for 6 h. A light colorless amorphous
powder was obtained, which was washed with methanol and shown
to be compound 6. The compound is soluble in DMSO. M.p. 132–
135 °C. Elem. Anal. Calcd. For C18H26N6O5Zn: C, 45.82; H, 5.55; N,
17.81. Found: C, 46.00; H, 5.85; N, 17.62%. Λm (MeCN, 10−4 M, 293 K):
3.2Ω−1 cm2 mol−1. IR (nujol mull, cm−1): 3128w (CH), 1631s, 1513m,
1441m (C=O) 402w.
2.3.1. {(L1)[CuCl2]2}, 1
To a stirred solution of L1 (0.59 g, 1.0 mmol) in dichloromethane,
CuCl2·2H2O (0.342 g, 2.0 mmol), dissolved in methanol, was added. The
reaction mixture was heated at reflux for 6 h, then concentrated to half
of its volume by rotary evaporation, and kept for 3 days at 4 °C. A micro-
crystalline compound was obtained which was washed with methanol
and shown to be compound 1. The compound is soluble in DMSO. M.p
214 °C. Elem. Anal. Calcd. For C30H30Cl4Cu2N12O2:C, 41.92; H, 3.52; N,
19.55. Found: C, 41.76; H, 3.43; N, 19.23%. Λm (DMSO, 10−4 M, 293 K):
38.7Ω−1 cm2 mol−1. µeff =2.20 BM. IR (nujol, mull, cm−1): 3179w,
3142w, 3074w (C–H) 1586w, 1511w, 615 m, 603 m, 498w, 476w,
409 m, 401 m, 386w, 295 s (Cu–Cl).
2.3.7. [(CuCl(L4)(Him)]2·2H2O, 7
To a stirred solution of HL3 (1 mmol, 0.248 g) in methanol,
CuCl2·H2O (2.0 mmol) and imidazole (2.0 mmol), were added. The
reaction mixture was stirred for 6 h. A light blue amorphous powder
was obtained, which was washed with methanol and shown to be
compound 7. The compound is soluble in DMSO. M.p. 217–220 °C.
Elem. Anal. Calcd. For C30H42Cl2Cu2N12O6: C, 41.67; H, 4.90; N, 19.44.
Λm (MeCN, 10−4 M, 293 K): 3.5Ω−1 cm2 mol−1. Found: C, 42.00; H,
4.75; N, 19.71%. µeff =1.52 BM. IR (nujol mull, cm−1): 3200br, 3158w,
3128w (CH) 1662 s, 1640sh, 1557 m, 1534w, 1500w, 1446 (C=O),
2.3.2. {(L1)[Cu(OAc)2]2}, 2
To a stirred solution of L1 (0.59 g, 1.0 mmol) in dichloromethane,
Cu(OAc)2·2H2O (0.434 g, 2.0 mmol) dissolved in methanol was
added. The reaction mixture was heated at reflux for 6 h, then
evaporated and the residue washed with a dichloromethane/diethyl
ether mixture. A micro-crystalline compound was obtained which
was washed with methanol and shown to be compound 2. The
compound is soluble in DMSO and MeCN. M.p 160–165 °C. Elem. Anal.
Calcd. For C38H42Cu2N12O10: C, 47.85; H, 4.44; N, 17.62. Found: C, 47.45;
430w, 417w, 226 m. ESI MS (CH3CN) (+): 657 [Cu2Cl(L4)2]+
.
2.4. Crystal structures determination
H, 4.54; N, 17.93%. Λm (MeCN, 10−4 M, 293 K): 3.0 Ω−1 cm2 mol−1
.
µeff =1.87 BM. IR (nujol, mull, cm−1): 3131w, 2935w (C–H) 1562 s,
2.4.1. Single-crystal X-ray structure determination of
1517 m, 1420 s.
[(CuCl(L2)(HpzPh,Me)], 4
A crystal of [(CuCl(L2)(HpzPh,Me)] with approximate dimensions
0.12×0.20×0.25 mm was mounted on the tip of a glass fiber and put on
a goniometer head. Diffraction measurements were performed on
an Enraf-Nonius CAD4 diffractometer using graphite monochromated
Mo-Kα radiation. Data collection (ω-scans) and processing (cell
refinement, data reduction and empirical absorption correction) were
performed using the locally available programs. The structure was
solved by direct methods using SHELXS-97 [41] and refined by full-
matrix least-squares techniques on F2 with SHELXL-97. All hydrogen
atoms were located by difference maps and were refined isotropically as
riding on their pertinent C, or N atoms. Table 1 contains a summary of
crystal data, data collection parameters and structural analysis.
Crystallographic data for the structural analysis (including that reported
below from powder diffraction data) have been deposited with the
Cambridge Crystallographic Data Centre, CCDC No. 761615-761616.
Copies of this information may be obtained free of charge on application
to CCDC, 12 Union Road, Cambridge CB2 1EZ, UK (fax: 44 1223 336 033;
e-mail: deposit@ccdc.cam.ac.uk or www: http://www.ccdc.cam.ac.uk).
2.3.3. {(L1)[Zn(OAc)2]2}, 3
Compound 3 was synthesized following a procedure analogous to
that reported for 2. To a stirred solution of L1 (0.59 g, 1.0 mmol) in
dichloromethane, Zn(OAc)2·2H2O (0.44 g, 2.0 mmol) dissolved in
methanol (5 mL) was added. The solution was stirred overnight, then
evaporated to half of its volume and refrigerated at 4 °C until a colorless
precipitate formed, which was filtered, washed with diethyl ether and
shown to be compound 3. The compound in insoluble in all organic
solvents. M.p 204 °C. Elem. Anal. Calcd. For C38H42N12O10Zn2: C, 47.66;
H, 4.42; N, 17.55. Found: C, 48.11; H, 4.62; N, 18.11%. IR (nujol mull,
cm−1): 1664 s, 1583br, 1516 m, 1473w, 1446 m(C=O), 437 m, 426 m.
2.3.4. [(CuCl(L2)(HpzPh,Me)], 4
To a stirred solution of K[L2] (0.48 g, 1.0 mmol) in dichloro-
methane, CuCl2·2H2O (0.171 g, 1.0 mmol) was added. The reaction
mixture was stirred at room temperature for 6 h, then filtered off and
the solution evaporated. A pale-green residue formed which was
washed with diethyl ether and shown to be compound 4. The
compound is soluble in chlorinated solvents and DMSO. Re-crystal-
lised from dichloromethane/diethyl ether. M.p 170 °C. Elem. Anal.
Calcd. for C40H38BClCuN8: C, 64.87; H, 5.17; N, 15.13. Found: C, 64.35;
2.4.2. Powder X-ray structure determination of {(L1)[CuCl2]2}, 1
A powdered, microcrystalline sample of 1 was gently ground in an
agate mortar, then deposited in the hollow of an aluminium sample

G. Lupidi et al. / Journal of Inorganic Biochemistry 104 (2010) 820–830
823
Table 1
Summary of crystal data and refinement parameters for compounds 1 and 4.
2.5. In vitro evaluation of radical scavenging activity
2.5.1. Evaluation of superoxide radical scavenging activity
Compound
1
4
Superoxide dismutase (SOD) activity of the complexes was
determined with a nonenzymatic assay method. Superoxide radicals
were generated by the NADH/PMS system according to ref. [44]. The
mixture (1 ml) in 20 mM potassium phosphate buffer contained NBT
43 µM, NADH 166 µM, aliquots of compounds in DMSO (from 0 to
30 µM), and PMS 2.7 µM that was added to start the reaction. The
reaction was conducted at room temperature for 2 min. The reduction
of NBT to the blue chromogen formazan by superoxide radical was
monitored at 560 nm.
Formula
C30H30Cl4Cu2N12O2
859.56
293
Monoclinic
C2/c
Powder
a=13.5231(4)Å
b=11.6834(4)Å
c=22.8433(8)Å
α=90°
β=109.997(2) °
γ=90°
3391.6(2)
48.82
4
C40H38BClCuN8
740.58
293
Triclinic
P-1
Single Crystal
a=11.748(2)Å
b=12.326(2)Å
c=15.286(3)Å
α=102.07(2)°
β=97.55(1)°
γ=116.20(2)°
1870.0(6)
6.93
Molecular weight (g mol−1
Temperature (K)
Crystal system
Space group
)
Method
Cell parameters
V (Å3)
2.5.2. Hydroxyl radical assay
µM (cm−1
Z
)
The deoxyribose method for determining the scavenging effect of
the SOD mimics on hydroxyl radicals was performed according to the
method previously described in [45].
2
ρ
(g cm−3
)
1.683
1744
1.310
770
calc
F(000)
Radiation
2θ min–max
Independent refl. (or y data)/parameters
R(int)
Rp, Rwp, RBragg
R, wR2
Cu–Kα
7.0–105.0
4901/50
Mo–Kα
6.0–50.6
6805 / 468
0.070
n.a
2.5.3. EPR spin trapping of superoxide radical with DMPO
Samples were assayed by EPR spectrometry for superoxide radical
dismutation activity using XO (0.4 u/ml) mediated oxidation of HX
(1 mM) as a source of superoxide and employing DMPO (100 mM) as
spin trap. Experiments were carried out in PBS (100 mM phosphate
buffer, pH 7.4, 0.9% NaCl), 1 mM DTPA in a final volume of 150 µl. Test
compounds were prepared in DMSO and used at several concentra-
tions (max. conc of DMSO used was 10% v/v). As a control, the test
compound was replaced by DMSO (10% v/v). The reaction was started
by addition of XO, and after thorough mixing, the mixture was
transferred to a Pasteur pipette whose narrow end was sealed off
using a Bunsen burner flame which served as sample cell, and the EPR
spectra recorded 90 s after XO addition. EPR spectra were collected on
a Bruker EMX EPR spectrometer (Bruker, Karlsruhe, Germany)
equipped with an XL Microwave frequency counter with the following
settings: frequency 9.78 GHz, power 25 mW, modulation amplitude
1 G, gain 5×105, field width 100 G, time constant 1.28 ms and scan
time 42 s. For each experiment, 10 scans were recorded.
n.a.
0.026. 0.040, 0.024
n.a
0.0529, 0.0979
holder (equipped with a zero-background plate). Diffraction data
were collected with overnight scans (16 h long) in the 5–105° 2θ
range on a Bruker AXS D8 Advance diffractometer, equipped with a
linear position-sensitive Lynxeye detector, primary beam Soller slits,
and Ni-filtered Cu–Kα radiation (λ=1.5418 Å); sampling interval (in
continuous mode): Δ2θ=0,02°. divergence slit: 1.0°; goniometer
radius 300 mm; generator setting: 40 kV, 40 mA. Standard peak
search, followed by indexing with TOPAS [42] allowed the detection of
the approximate unit cell parameters later improved by LeBail
refinements. Indexing figure of merit M(22)=61.5. Space group
determination, performed using systematic extinction conditions, in
conventional mode as well using a structureless full pattern profile
match, indicated C2/c as probable space group (later confirmed by
successful structure solution and refinement). Structure solution was
performed (by the simulated annealing technique) as implemented in
TOPAS, using for “half” organic ligand (the crystallographic indepen-
dent portion) a rigid, idealized model [43], (flexible at the C-
pyrazole links and at the saturated O and C chain atoms) and one
independent metal and two chloride ions. The final refinements were
carried out by the Rietveld method, maintaining the rigid bodies
described above (with the organic ligands bisected by a twofold axis).
Peak shapes were defined by the Fundamental Parameters Approach
implemented in TOPAS, while crystal size effect was modelled by a
Lorentzian broadening. The background contribution to the total
scattering was modelled by Chebyshev's polynomials, also accounting
for a structured trace below Bragg peaks. One, refinable isotropic
thermal parameter was assigned to the metal atom, augmented by
2.0 Å2 for lighter atoms. No preferred orientation correction was found
to be necessary. The final Rietveld refinement plot is shown in Fig. 1.
Table 1 contains a summary of crystal data, data collection parameters
and structural analysis.
2.5.4. EPR spin trapping of hydroxyl radical with DMPO
Hydroxyl radical generating ability from test compounds was
examined by their addition to a reaction mixture containing 1 mM
H2O2 in the presence of 100 mM DMPO in PBS, 1 mM DTPA in a final
volume of 150 µl. The reaction was started by addition of H2O2 and
after thorough mixing, the mixture was transferred to a sample cell
and EPR spectra recorded 90 s after H2O2 addition as described above.
Hydroxyl radical scavenging ability of test compounds was
examined by their addition to the hydroxyl radical-generating Fenton
system consisting of 50 µM ferrous ammonium sulphate, 1 mM H2O2
in the presence of 100 mM DMPO in PBS, 1 mM DTPA in a final volume
of 150 µl. The reaction was started by addition of H2O2 and after
thorough mixing, the mixture was transferred to a sample cell and
EPR spectra recorded 90 s after H2O2 addition as described above.
2.5.5. Evaluation of peroxidase activity
Peroxidase activity was monitored using the DCFH fluorescence
method. This activity can be conveniently measured by monitoring
Fig. 1. Rietveld refinement plot for species 1, with difference plot and peak markers at the bottom. Horizontal axis, 2θ, °; vertical axis, counts.

824
G. Lupidi et al. / Journal of Inorganic Biochemistry 104 (2010) 820–830
oxidation of dichlorodihydrofluoresceine diacetate hydrolyzed freshly
for every experiment according to the published procedure [46].
Briefly, DCFH dissolved in ethanol was hydrolyzed by NaOH (0.01 M)
in the dark for 30 min at room temperature and stored in an ice bath
during the experiment. Peroxidase activity was measured as follows:
SOD-mimics were mixed with a solution of DCFH (2.5 µM) in
phosphate buffer (50 mM, pH 7.4, DTPA 0.1 mM). The reaction was
initiated by adding H2O2 and the increase in fluorescence emission
due to DCF formation was monitored at 520 nm for 30 min.
(excitation λ=495 nm).
remain coordinated to copper also in solution, as indicated by the
conductivity value found [47]. The IR spectra show all bands due to
the neutral bis(tripodal) ligands and to the counter-ion, and in the
case of 2 the existence of a bridging acetate [48]. The reaction between
L1 and Zn(OAc)2 yielded the 1:1 adduct 3 also when the reaction was
carried out in strong ligand excess (Scheme 2).
The synthesis of the polypyrazolylborate ligand TpPh,Me, here
described, is essentially the same as that used to prepare the parent
ligands Tp* by heating sodium or potassium tetrahydridoborate with
an excess of the HpzPh,Me [49]. Hydrogen gas is evolved during the
reaction and its progress was measured. The copper(II) chloro
complex 4 was obtained from the reaction of two equivalents of
K[TpPh,Me] with one equivalent of the copper salt in methanol and in
dichloromethane, respectively, at room temperature and in good yields
(Scheme 3). The presence of the neutral HpzPh,Me coordinated to copper,
is due to the hydrolysis of excess TpPh,Me, a phenomenon already
observed [50]. The Mn complex [Mn(L3)2]·2H2O 5 was obtained from
the reaction of two equivalents of the ligands HL3 in MeOH with one
equivalent of the metal salt (Scheme 3). Deprotonation of HL3 occurs
also in the absence of base. Whereas from the reaction of equimolar
quantity of HL4 with Zn and Cu salts in the presence of excess imidazole,
derivatives 6 and 7 formed (Scheme 3).
2.5.6. Measure of ONOO− scavenging activity
Peroxynitrite (ONOO−) scavenging ability of different complexes
was measured by monitoring the inhibition of oxidation of non-
fluorescent DCFH to the highly fluorescent DCF. DCFH was prepared as
previously described. To measure the scavenging activity of perox-
ynitrite, ONOO− at a fixed concentration (0.02 mM) was added to
DCFH (2.5 µM) in 2 ml of 50 mM phosphate buffer, pH 7.4, DTPA
0.1 mM, containing 25 µM of the different complexes, under contin-
uous stirring and the solution was incubated at 25 °C for 15 min. After
this time the fluorescence intensity of oxidized DCFH was measured at
excitation and emission wavelengths of 490 and 520 nm respectively.
3. Results
3.2. Spectroscopic characterization
3.1. Synthesis of ligands and complexes
Due to the limited solubility of compounds 1–3, polynuclear
structures (see also below Section 3.3) were hypothesized for all of
them, in which the L1 ligand bridges two metal centres. IR spectral
analysis seems in accordance with the proposed structures. In fact, in
the IR spectrum of 1 only a strong band due to Cu–Cl stretching mode
was observed at 295 cm−1, that is indicative of only one terminal
chloride. In the IR spectra of 2 and 3 a number of strong bands were
found in the range of 1400–1700 cm−1, assignable to ν(COO) of
acetates. Such an high number of bands cannot give a straightforward
indication of the bonding mode of acetates, however one could
tentatively hypothesize that they are bridged between two metal
The reaction between 1,4-C6H4[CH2OCH2C(pz)3]2 (L1) and CuX2·
nH2O (X=Cl, or OAc) in a mixture of dichloromethane and methanol
always yielded complexes of formula [(L1)(CuX2)], (1: X=Cl, 2:
X=OAc) (Scheme 2) independent of the stoichiometric ligand to
metal ratio employed, insoluble in MeCN and chlorinated solvents and
moderately soluble in DMSO, in which 2 is not-electrolyte, in contrast
to 1 which exhibits a conductivity value typical of 1:1 electrolyte [47],
suggesting only partial dissociation in DMSO upon breaking of the
polymeric structure. Whereas in 2, both acetate counterions likely
Scheme 2.

G. Lupidi et al. / Journal of Inorganic Biochemistry 104 (2010) 820–830
825
Scheme 3.
centres and also coordinated as terminal monodentates, as it is likely
should indicate partial quenching of magnetism, due to super-
exchange mechanisms through the bridging chloride ligands, whereas
the manganese derivative 5 shows a value of 5.96 B.M., in accordance
with a high spin d5 Mn(II) centre, thus indicating that (L3)− behaves
as a weak-field ligand.
in the case of compound 3, where a strong band was observed at
1664 cm−1
.
In the IR spectrum of compound 4 the typical ν(B–H) mode was
found at2526 cm−1, together withan intense bandat3215 cm−1 due to
the ν(N–H) of the 3-phenyl-5-methylpyrazole arising from decompo-
sition of the (L2)− ligand. Moreover a broad band at 280 cm−1 confirms
the presence of a Cu–Clbond. Finally, also the ESI-MS spectrumindicates
the presence of a Cu environment containing the tris(pyrazol-1-yl)
borate and the neutral pyrazole, as the peak at 704 can be easily
attributed to the cationic fragment [(Cu(L2)(HpzPh,Me)]+.
3.3. Structural characterization
[L1(CuCl2)2], 1, is a polymeric complex, built upon Cu(II) dimers of
ClCu(µ-Cl)2CuCl formulation, connected in one-dimensional chains
(shown in Fig. 2) by the large L1 ligand.
The IR spectra of compounds 5–7 containing the bis(pyrazol-1-yl)
acetate ligands, display the typical pattern due to monoanionic
tridentate ligands, in which the carboxylate arm is also coordinated to
the metal centre, the highest ν(COO) being always found in 5–7 at ca.
1630–1670 cm−1, whereas uncoordinated carboxylate groups are
generally observed at higher frequencies, above 1700 cm−1 [39,50].
The dinuclear nature of compound 7 was proposed mainly on the
basis of the ESI-MS spectrum that shows a peak at 657 assignable to
Both the dimer and the ligand are located, in the crystal, onto a
special position, an inversion centre for the former and a twofold axis
for the latter. Since it is well known that powder diffraction methods
on complex materials cannot attain the precision typical of conven-
tional single-crystal structure determinations, a rigid body description
(see Experimental section) was employed, in which only the flexible
torsional freedom is tolerated. Accordingly, our model shows
pentacoordinated Cu(II) ions in CuCl2N2 chromophores of square-
pyramidal shape, in which the apical Cu–Cl distance (2.82) Å) is
significantly longer than the equatorial ones (2.24–2.28 Å). Two
nitrogen atoms, in cis position and ca. 2.08 Å apart from the metal
centre complete the overall coordination, while the third pyrazolate
residue is, surprisingly, non-coordinating, with a Cu…N distance of
3.21 Å. The presence of a vacant site on the Cu(II) ion and the odd
conformation of what we (initially) thought being an ideal tripod-like
the cationic fragment [Cu2Cl(L4)2]+
.
The magnetic measurements on the paramagnetic compounds 1, 2,
4, 5 and 7 were performed to gain more information on the magnetic
status of the metal centre: apart from the copper derivatives 1, 2 and 4
that show dipole magnetic moments in the range of 1.87–2.20 B.M.,
typical of d9 Cu(II) ions with a unique unpaired electron, the copper
(II) derivative 7 displays a somewhat lower value of 1.52 B.M. which

826
G. Lupidi et al. / Journal of Inorganic Biochemistry 104 (2010) 820–830
Fig. 2. Schematic drawing of: a portion of the polymeric species 1, with partial labelling scheme. In the crystal, wavy chains run parallel to the [101] direction (the horizontal axis).
Vertical axis, b (the twofold axes bisecting the external 1,4-substituted aromatic rings of the L1 ligand). Relevant bond distances (Å) and angles (°): Cu–Cl1 2.278(9); Cu–Cl2 2.236(7),
Cu–Cl2′ 2.818(9), Cu–N21 2.069(6), Cu–N31 2.084(6), Cl1–Cu–N31 162.8(3), Cl2–Cu–N21 176.8(4). Copper: yellow; Chlorine: green; Nitrogen: blue; Oxygen: red.
ligand, prompted us to verify which structure had to be considered
correct, i.e. the open, or the closed one, or any intermediate
conformation. Several tests performed using restrained torsional
values (not shown here) clearly indicated that the open conformation
is the correct one, as also witnessed by the very low profile and Bragg
agreement factors.
Worthy of note, the overall crystal packing (not shown here) is
generated by juxtaposition of infinite, parallel and wavy chains
running along the [101] direction, with the L1 ligand moieties
alternating above, and below, the chain axis. This effect can be
graphically appreciated in Fig. 2, where fragmented lines connect the
longest Cu…Cl interactions.
defined by the Cu–N bond of the neutral ligand (2.034(2) Å), and of
one pyrazolate branch of the tripodal moiety (2.056(2) Å). In such a
crowded environment, phenyl rings of the tripod are (coherently, i.e.
propeller-like) twisted at an angle (in the 30–40° range) with respect
to the heteroaromatic rings. Similarly, also the torsional angle about
the C13–C14 bond, linking the two aromatic rings in the 3-phenyl-5-
methylpyrazole ligand, is significantly different from zero, approach-
ing 38°. Under these structural conditions, the Cu(II) ions are deeply
buried within the ligands sphere, but, thanks to a possible labilization
of the Cu–N bond of the neutral moiety, ligand dissociation and
substitution may occur, thus giving the whole complex a chemical
reactivity which, per se, is not evident from the (rigid) solid-state
structure.
At variance from 1, species 4, [(CuCl(L2)(HpzPh,Me)], contains
mononuclear Cu(II) complexes, the structure of which is schemati-
cally shown in Fig. 3; these molecules, in the crystal phase, weakly
interact with neighbouring (symmetry equivalent) complexes
through weak H…H (2.29–2.36 Å, and above) and H…Cl (2.90 Å)
contacts.
3.4. Evaluation of superoxide radical scavenging activity
Different methods (non enzymatic and enzymatic methods) can be
used to ascertain the Superoxide Radical Scavenging activity of various
compounds. The SOD-like activity for the reaction of different mimics
was determined studying the effect of the complexes on superoxide
generated by the NADH/PMS system as described in materials and
methods. This method is very sensitive for evaluation of superoxide
radical scavenging activity and was used to calculate the IC50 of the most
active compounds (Table 2). As reported, all complexes show a dif-
ferent rate of reaction with O2− radicals. Mimics [(Mn(L3)2]·2H2O (5),
[(CuCl(L4)(Him)]2·2H2O (7) and {(L1)[CuCl2]2} (1), show the greatest
activity withrespecttothe other compounds tested while less inhibition
was show by Zn complexes. Under the reaction conditions used, the O2−
radicals activity was almost completely inhibited by SOD at a
concentration of 0.10 µM (IC50=0.01 µM).
Each Cu(II) centre is linked to a tripod-like L2 ligand, to a
monodentate 3-phenyl-5-methylpyrazole moiety and to a terminally
bound chloride ion (Cu–Cl 2.2833(8) Å). The overall coordination
geometry is close to trigonal–bipyramidal, with the apical positions
The hydroxyl radical scavenging activity of the different complexes
was also evaluated using the deoxyribose assay. Deoxyribose is
degraded to malondialdehyde on exposure to hydroxyl radicals
generated by Fentonsystems. Under acidic conditionsand uponheating,
malondialdehyde reacts with thiobarbituric acid to form a pink
chromogen which may be easily detected spectrophotometrically [45].
Table 2
SOD-like activity of different complexes.
SOD-Mimics
IC50 µM
(1){(L1)[CuCl2]2}
1.82( 0.15)
Not tested
(2){(L1)[Cu(OAc)2]2}
Fig. 3. ORTEP drawing of the species 4, with partial labelling scheme. Thermal ellipsoids
drawn at the 50% probability level. Hydrogen atoms omitted for clarity. Relevant
bond distances (Å) and angles (°): Cu–Cl 2.2833(8), Cu N1 2.246(2), Cu N3 2.056(2),
Cu N5 2.035(2), Cu N8 2.035(2). N1 Cu Cl 116.14(6), N3 Cu Cl 93.37(6), N5 Cu Cl 149.02(6),
N8 Cu Cl 91.14(6), N3 Cu N1 89.43(8), N5 Cu N1 94.77(8), N8 Cu N1 88.78(8), N5 Cu N3
84.64(8), N8 Cu N3 175.49(8), N5 Cu N8 91.38(8).
(3){(L1)[Zn(OAc)2]2}
16.06( 0.15)
14.30 (( 0.40)
2.70( 0.30)
17.11( 0.50)
1.16( 0.10)
(4) [(CuCl(L2)(HpzPh,Me)]
(5) [(Mn(L3)2(H2O)2]
(6) [(Zn(OAc)(L3)(Him)]·MeOH
(7) [(CuCl(L4)(Him)]2·(H2O)2

G. Lupidi et al. / Journal of Inorganic Biochemistry 104 (2010) 820–830
827
The hydroxyl radical scavenging properties were tested for all SOD-
mimics using this method and no pro-oxidant effect was observed at the
concentration of complexes used. The assay was performed in the
absence of EDTA. As reported in Fig. 4 all compounds exhibited weak
comparable scavenging activity for hydroxyl radicals except for the
{(L1)[Zn(OAc)2]2} (3) that lacks this activity while for [(Zn(OAc)(L3)
(Him)]·MeOH (6) it is reduced with respect to the other SOD-mimics.
The superoxide scavenging activity of the complexes was also
examined by EPR spin-trapping, after having established the condi-
tions for detection of superoxide radical and interpretation of EPR
spectra. After addition of XO to the reaction mixture, the EPR signal of
DMPO–OOH adduct (both in the absence or presence of DMSO) was
recorded at room temperature after 90 s and scans were collected
thereafter for the subsequent 10min. The characteristic spectra
consisting of 12 lines with 2:4:4:2 pattern with hyperfine splitting
constants aN =13.89 G, aHβ =11.61 G e aγH =1.28 G are reported in
Fig. 1Sa,b (Supplementary material). The DMPO–OOH spectrum
decays within a few minutes and is replaced by the appearance of a
new signal typical of the DMPO–OH adduct (Fig. 1Sc, Supplementary
material, in the absence of DMSO) characterized by a 4 line spectrum
with a 1:2:2:1 pattern and hyperfine splitting constants aN =14.85 G,
aβH =14.85 G, or of the DMPO–CH3 adduct (Fig. 1Sd, Supplementary
material), in the presence of DMSO, characterized by a 6 line spectrum
of equal intensity with hyperfine splitting constants aN =16.1 G,
aβH =23.0 G. Both these signals arise partly from the well known
spontaneous decay of the DMPO–OOH adduct, and partly from the
spontaneous dismutation of superoxide radical since this is not
completely scavenged by DMPO. In the latter case, the reactions
leading to the two adducts are described in Eqs. (1)–(5):
{(L1)[CuCl2]2} (1), [(CuCl(L2)(HpzPh,Me)] (4), [(Mn(L3)2]·2H2O (5)
and [(CuCl(L4)(Him)]2·2H2O (7), were tested. The other metal
complexes could not be tested due to insolubility reasons at the
concentrations required in this experimental system. Different
concentrations for each compound were examined in order to find
the concentration at which the DMPO–OOH signal at 90 s (Fig. 5a) was
inhibited by roughly 50%, which corresponds to a SOD concentration
of 2.5 u/ml (Fig. 5b). For compounds [(Mn(L3)2]·2H2O (5) and [(CuCl
(L4)(Him)]2·2H2O (7), a 500 µM concentration lead to an appreciable
decrease in the DMPO–OOH signal (Fig. 5c,d), indicating superoxide
radical scavenging activity. For compound [(CuCl(L2)(HpzPh,Me)] (4),
only a slight decrease in this signal was observed (Fig. 5e) which
was unaffected by increasing its concentration. Instead complex
{(L1)[CuCl2]2} (1) lead to roughly a 50% decrease in the DMPO–OOH
signal at concentrations as low as 325 µM (Fig. 5f). However, at
concentrations 400 µM and above, the DMPO–OOH signal was
immediately replaced by the DMPO–CH3 signal. This can be explained
by admitting a Fenton-like activity which promotes hydroxyl radical
formation and hence detection of this adduct (Eqs. (3)–(5)). In fact,
Fig. 2Sa (Supplementary material) shows the DMPO–CH3 signal
obtained from 400 µM complex {(L1)[CuCl2]2} (1). This signal was also
observed with complex [(CuCl(L4)(Him)]2·2H2O (7) but at 750 µM
(Fig. 2Sb, Supplementary material). Complex [(Mn(L3)2]·2H2O (5)
showed no Fenton-like activity, even at concentrations as high as 1 mM
(Fig. 2Sc, Supplementary material). In this case no signal was observed
because at this concentration superoxide radical is efficiently scavenged
hence none is trapped by DMPO. For complex [(CuCl(L2)(HpzPh,Me)] (4)
(Fig. 2Sd, Supplementary material), no Fenton-like activity was
observed, even at 1 mM since the DMPO–OOH signal was the only
signal recorded.
O2•− + O2•− + 2Hþ→H2O2 + O2
O•2− + H2O2 + 2Hþ→O2 + H2O + OH•
DMPO + OH•→DMPO−OH
ð1Þ
ð2Þ
ð3Þ
ð4Þ
ð5Þ
To further confirm that complexes {(L1)[CuCl2]2} (1) and [(CuCl(L4)
(Him)]2·2H2O (7) were capable of generating hydroxyl radicals by
Fenton-type reactions, H2O2 (1 mM) was added to them in the presence
of DMPO and the DMPO–CH3 signal with traces of the DMPO–OH one
were immediately observed (Fig. 3Sa,b, Supplementary material).
Instead, no EPR signals were observed for complexes [(Mn(L3)2]·2H2O
(5) and [(CuCl(L2)(HpzPh,Me)] (4) (Fig. 3Sc,d, Supplementary material)
as expected.
OH• + CH3SðOÞCH3→CH3SðOÞOH + CH3•
DMPO + CH3• →DMPO−CH3:
Considering that [(Mn(L3)2]·2H2O (5) is a Mn complex and since
previous literature reports have demonstrated that Mn complexes are
capable of scavenging hydroxyl radicals besides possessing SOD-like
activity, the former activity was tested on this complex. Hydroxyl free
radicals were generated by the Fenton reaction (H2O2/Fe2+) and
indirectly demonstrated by the appearance of the DMPO–CH3 signal
with traces of the DMPO–OH one (Fig. 4Sa, Supplementary material).
As shown in Fig. 4Sb (Supplementary material), the presence of 1 mM
of [(Mn(L3)2]·2H2O (5) lead to a remarkable decrease in intensity of
this signal, a decrease which was concentration dependent (not
shown). This result thus suggests the hydroxyl scavenging ability of
this Mn complex which is in agreement with the results shown in
Fig. 4.
Addition of SOD (5 u/ml) to the reaction medium resulted in a loss
of signal intensity due to the rapid dismutation reaction rate
(2.3×109 M−1 s−1) [51] compared with the slow trapping rate
(10 M−1 s−1) [52].
Having established the conditions for superoxide generation
and interpretation of the EPR spectra, the following metal complexes,
Based on the results obtained by EPR-spin trapping, the potential
peroxidase activity of all the complexes using a fluorescent method
was investigated. As previously reported in [31], with this method it is
possible to measure the peroxidase activity of SOD: incubation of the
enzyme with a mixture containing DCFH and H2O2 causes a time
dependent increase in DCF fluorescence (Fig. 6). In accordance with
the EPR data, the complex [(CuCl(L4)(Him)]2·2H2O (7) shows the
highest peroxidase activity while only a slight increase in fluorescence
with respect to the control is shown for {(L1)[CuCl2]2} (1). For the
other SOD-like complexes, no peroxidase activity was observed.
An important mechanism by which SOD mimics can attenuate
inflammation is by reducing peroxynitrite formation by simply
removing superoxide before it can react with nitric oxide. These
compounds also have the possibility to react with peroxynitrite and
reduce the oxidative damage it causes as reported in Fig. 7. From the
Fig. 4. Hydroxyl radical scavenging activity effects of different SOD-mimics tested at
25 μM in DMSO.

828
G. Lupidi et al. / Journal of Inorganic Biochemistry 104 (2010) 820–830
Fig. 5. EPR spin trapping of superoxide radical with DMPO in the presence of SOD-like mimics. a) DMPO–OOH obtained from HX/XO+DMSO system recorded 90 s after addition of
XO; b) DMPO–OOH obtained from HX/XO+DMSO system in the presence of SOD 2.5 u/mL; c) DMPO–OOH obtained from a) in the presence of [(Mn(L3)2](H2O)2,(5) 500 μM;
d) DMPO–OOH obtained from a) in the presence of [(CuCl(L4)(Him)]2·(H2O)2 ,(7) 500 μM; and e) DMPO–OOH obtained from a) in the presence of [(CuCl(L2)(HpzPh,Me)], (4) 500 μM;
f) DMPO–OOH obtained from a) in the presence of {(L1)[CuCl2]2}, (1) 325 μM.
data obtained it is possible to note that the dinuclear metal complexes
show scavenging activity against this oxidant: a remarkable reactivity
was observed for [(Mn(L3)2]·2H2O (5) with respect to the Cu and Zn
derivatives (1), (7), (3) and (6) {(L1)[CuCl2]2} (1), [(CuCl(L4)
(Him)]2·2H2O (7), {(L1)[Zn(OAc)2]2} (3), [(Zn(OAc)(L3)(Him)]·
MeOH (6) (that show comparable activity) whereas this is totally
absent in the [(CuCl(L2)(HpzPh,Me)] (4) derivative. In particular, the
activity of the Mn complex was further investigated and as reported in
the inset of Fig. 7, its scavenging activity is concentration-dependent.
4. Discussion
Recent advances in medicinal inorganic chemistry demonstrate
significant perspectives for the utilization of metal complexes as drugs,
presenting a flourishing area of work for inorganic chemistry [7–10]. In
this study, different Cu, Zn and Mn dinuclear complexes with some
neutral and monoanionic pyrazole-based ligands were synthesized and
characterized and their radical scavenging properties evaluated. Some
of these compounds are not soluble in solvents that can be used for
biological purposes hence it was impossible to evaluate their potential
activities. All DMSO-soluble tested complexes, catalyze superoxide
dismutation with an IC50 in the micromolar range (Table 2). Even if in
strong polar solvents such as DMSO a certain degree of ionization of our
Fig. 6. Evaluation of peroxidase activity of different SOD-mimics. Time-dependent
fluorescence measurement of DCF formation (excitation λ=495 nm; emission
λ=520 nm) was performed on incubation mixtures containing DCFH (2.5 mM),
H2O2 (1 mM), phosphate buffer 0.1 M, pH 7.4 (containing DTPA (100 µM)) and (●)
DCFH alone; (○)[(CuCl(L4)(Him)]2·(H2O)2, (7); (Δ){(L1)[CuCl2]2}(1); (▲) [(CuCl(L2)
(HpzPh,Me)](4); (□)[(Zn(OAc)(L3)(Him)]·MeOH,(6); (■) [(Mn(L3)2](H2O)2,(5); (*){(L1)
[Zn(OAc)2]2} (3). The concentration for all compounds tested was 25 µM.

G. Lupidi et al. / Journal of Inorganic Biochemistry 104 (2010) 820–830
829
the case of Mn, and Zn complexes [(Mn(L3)2]·2H2O (5) and [(Zn
(OAc)(L3)(Him)]·MeOH (6). As an important non-enzymatic antiox-
idant, Zn complexes are attracting ample attention from researchers
[54] and it can be seen that the inhibitory effects of the zinc deriv-
atives {(L1)[Zn(OAc)2]2} (3) and [(Zn(OAc)(L3)(Him)]·MeOH (6)
against O−2 · and .OH radicals are related to concentration. Our results
are in line with those reported in different studies where it has been
demonstrated that different zinc salts [55] or zinc complexes
prepared with pharmacologically active ligands [56–58] possess
potent antioxidant activity, and can attenuate the biochemical
consequences of oxygen free radicals. In particular, Zn-complexes
seem to be active scavengers of OH·.
The pro-oxidant activity of the two Cu complexes is totally reduced
in the concentration range around the IC50 values (data not shown). In
view of the critical role of the superoxide radical in several disease
states, these new dinuclear SOD-like mimics also appear to react with
relevant biological oxidants, in particular with peroxynitrite. This is
important since the pro-inflammatory and cytotoxic effects of peroxy-
nitrite in tissues are numerous and the removal of peroxynitrite by
agents such as SOD-like mimics may be cytoprotective and anti-
inflammatory [59,60]. From the data reported in Fig. 7, the complexes
may be ranked according to their scavenging activity towards
peroxynitrite. The complex [(Mn(L3)2]·2H2O (5) shows the highest
reactivity which is twice that of the Cu and Zn derivatives. Therefore
from this preliminary screening on the scavenging properties of these
new dinuclear complexes, the only compound which appears to have
superior antioxidant properties compared to the others tested is the
complex [(Mn(L3)2]·2H2O (5) which besides scavenging superoxide
radical, does not show any prooxidant activity at high concentrations.
The dinuclear Mn compounds could be considered good candidates for
use as human pharmaceuticals as previously reported for other
mononuclear Mn complexes [6] for which biological and kinetic data
are available. Several biological studies have been published for Mn(III)
(salen) complexes, Mn(III)(porphyrinato) complexes, Mn(II)
(1,4,7,10,13-pentaazayclopentadecane) and Mn(II) bis(cyclohexylpyr-
idine)-substituted (M40403) derivatives (for review see Refs.
[6,10,12,61]), which have been shown to be promising active drugs in
clinical therapy for diseases mediated by superoxide radicals. The IC50
value exhibited by [(Mn(L3)2]·2H2O (5) is lower than those reported for
other Cu–SOD mimics, nevertheless, further investigations should be
done regarding with the stability of the metals in their complexes, their
toxicity and importantly, their biological effects at the cellular level.
These studies are in progress in our laboratory and currently we
continue to investigate new strategies to enhance the ROS-scavenging
systems of the cell, in an effort to attenuate the damage resulting from
oxidative stress. SOD is in fact one of the first line ROS-defence enzymes
in cells whose role is becoming increasingly important as more
pathological and disease states are discovered to be linked with
oxidative damage.
Fig. 7. Evaluation of peroxynitrite scavenging activity of different SOD-mimics.
Inhibition of peroxynitrite-mediated dichlorodihydrofluorescein oxidation by different
SOD mimics (25 µM) detected by fluorescence measurements. Data are expressed as
the mean SD for three replicate determinations. Experimental condition as reported
in methods. Inset: peroxynitrite scavenging activity as a function of different [(Mn(L3)2]
(H2O)2,(5) concentrations. The fluorescence measurement of DCF formation was
obtained after a 15-min incubation of mixtures (as described in Materials and methods)
at 25 °C, containing different concentrations of complex.
derivatives cannot be excluded, the remarkable SOD activity observed
for Cu-complexes ([(CuCl(L4)(Him)]2·2H2O (7) and {(L1)[CuCl2]2} (1))
is probably related to the resemblance of their coordination environ-
ments with that in the native SOD enzyme. Increasing the steric
hindrance in the Cu coordination environment as in derivative
[(CuCl(L2)(HpzPh,Me)] (4) does not increase the activity of the
mimics. The synthesis of dinuclear Zn mimics lead to two derivatives
with decreasing SOD-like activity that are however indicative for
constructing new promising Cu–Zn stable complexes. For these
derivatives it is important to determine whether a compound is in
fact acting as a true catalytic SOD mimic and to this end, the activity
of the complexes using EPR spectroscopy was evaluated which
confirms the data obtained with indirect methods. From the results
of the above EPR experiments, it may be concluded that with the
exception of [(CuCl(L2)(HpzPh,Me)] (4) which showed negligible
SOD-like activity, at least under the experimental conditions used in
this study, all the other complexes were capable of dismutating
superoxide radical, {(L1)[CuCl2]2} (1) being the most active, i.e. 50%
reduction in DMPO–OOH signal at concentrations b350 mM. How-
ever, besides this superoxide scavenging activity, both {(L1)[CuCl2]2}
(1) and especially [(CuCl(L4)(Him)]2·2H2O (7) also generate
hydroxyl radicals when used at high concentrations. This result is
not surprising since these two complexes both contain Cu as the
active metal for SOD mimic activity, which has the potential to
participate in Fenton chemistry. Therefore, the only complex which
appears to have superior antioxidant properties compared to the
others tested, is the Mn complex [(Mn(L3)2]·2H2O (5) which not
only scavenges superoxide radical, but also the toxic hydroxyl radical
(Fig. 4). An important consideration on the catalytic activity of the
different compounds as SOD-mimics is that they can act as oxido-
reductases in vitro and in vivo assays. As oxidoreductases, SOD-
mimics may react with superoxide in a reduction reaction but they
can also react with the product of dismutation, H2O2 hence it is
possible that an apparent catalytic redox cycle may possibly occur.
This behaviour is not new because even Cu–Zn–SOD enzymes show
this property, which is however absent in the mitochondria Mn–SOD
form of the enzyme [53]. The pro-oxidant properties of different
SOD-like mimics were evaluated using EPR and fluorescence
methods and as reported in the Results section, the dinuclear copper
complexes [(CuCl(L4)(Him)]2·2H2O (7) and {(L1)[CuCl2]2} (1) show
this activity under our experimental conditions, while it is absent in
Table of abbreviations
L1
1-(2-(4-((2,2,2-tris(pyrazol-1-yl)ethoxy)methyl)benzy-
loxy)-1,1-di(pyrazol-1-yl)ethyl)-pyrazole
potassium hydridotris(3-phenyl-5-methylpyrazolyl)borate
bis(pyrazol-1-yl)acetic acid
bis(3,5-dimethylpyrazol-1-yl)acetic acid
acetate
K[L2]
HL3
HL4
OAc
HpzPh,Me 3-phenyl-5-methylpyrazole
imH
THF
EPR
IC50
ROS
SOD
HX
imidazole
tetrahydrofuran
electron paramagnetic resonance
inhibiting concentration of 50% target substrate
reactive oxygen species
superoxide dismutases
hypoxanthine
XO
xanthine oxidase

830
G. Lupidi et al. / Journal of Inorganic Biochemistry 104 (2010) 820–830
[23] B.S. Hammes, M.T. Kieber-Emmons, J.A. Letizia, Z. Shirin, C.J. Carrano, L.N. Zakharov,
CAT
PMS
NBT
DMPO
DTPA
DCFH
catalase
A.L. Rheingold, Inorg. Chim. Acta 346 (2002) 227–238.
phenazine methosulfate
nitroblue-di tetrazolium chloride
5,5-dimethyl-1-pyrroline-N-oxide
Diethylene triamine pentaacetic acid
2,7-dichloro-fluorescin
[24] B. Kozlevčar, P. Gamez, R. de Gelder, W.L. Driessen, J. Reedijk, Eur. J. Inorg. Chem.
(2003) 47–50.
[25] G.A. Santillan, C.J. Carrano, Inorg. Chem. 46 (2007) 1751–1759.
[26] G.A. Santillan, C.J. Carrano, Dalton Trans (2008) 3995–4005.
[27] G.A. Santillan, C.J. Carrano, Inorg. Chem. 47 (2008) 930–939.
[28] B. Kozlevčar, A. Golobič, P. Gamez, I.A. Koval, W.L. Driessen, J. Reedijk, Inorg. Chim.
Acta 358 (2005) 1135–1140.
[29] M. Pellei, G. Gioia Lobbia, C. Santini, R. Spagna, M. Camalli, D. Fedeli, G. Falcioni,
Dalton Trans (2004) 2822–2828.
Acknowledgements
[30] L. Peters, E. Hübner, T. Haas, F.W. Heinemann, N. Burzlaff, J. Organomet.Chem 694
(2009) 2319–2327.
[31] E. Hübner, T. Haas, N. Burzlaff, Eur. J. Inorg. Chem. (2006) 4989–4997.
[32] N. Burzlaff, I. Hegelmann, B. Weibert, J. Organomet. Chem. 626 (2001) 16–23.
[33] N. Burzlaff, in: R. van Eldik (Ed.), Tripodal N,N,O-Ligands For Metalloenzyme
Models And Organometallics, Adv. Inorg. Chem., Vol. 60, Elsevier, London, 2008,
pp. 101–165.
Financial supports by the University of Camerino, Fondazione
CARIMA and Valle-esina Spa are gratefully acknowledged. Authors are
also grateful to the Italian Ministry of University and Research –
International University cooperation programs-2009 – for the financial
support.
[34] A. Otero, J. Fernández-Baeza, A. Lara-Sánchez, J. Tejeda, L. F: Sánchez-Barba, Eur. J.
Inorg. Chem (2008) 5309–5326.
[35] R.M. Uppu, G.L. Squadrito, R. Cueto, W.A. Pryor, Methods Enzymol. 269 (1996)
311–312.
[36] M.W. Senko, IsoPro Isotopic Abundance Simulator, v. 2.1, National High Magnetic
Field Laboratory, Los Alamos National Laboratory, Los Alamos, NM, 1994.
[37] D.L. Reger, T.D. Wright, R.F. Semeniuc, T.C. Grattan, M.D. Smith, Inorg. Chem. 40
(2001) 6212–6219.
[38] K.-W. Yang, Y.-Z. Wang, Z.-X. Huang, J. Sun, Polyhedron 16 (1997) 109.
[39] F. Marchetti, M. Pellei, C. Pettinari, R. Pettinari, E. Rivarola, C. Santini, B.W. Skelton,
A.H. White, J. Organomet. Chem. 690 (2005) 1878–1888.
[40] H. Kopf, C. Pietraszuk, E. Hübner, N. Burzlaff, Organometallics 25 (2006) 2533–2546.
[41] G.M. Sheldrick, SHELXL-97: Crystal Structure Refinement Program, University of
Göttingen, Germany, 1997.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.jinorgbio.2010.03.013.
References
[1] A.D. de Grey, Free Radic. Res. 40 (2006) 1244–1249.
[2] F. Johnson, C. Giulivi, Mol. Aspects. Med. 26 (2005) 340–352.
[3] J.R.J. Sorenson, Curr. Med. Chem. 9 (2002) 639–662.
[42] Version 3.0, Bruker AXS, 2005, Karlsruhe, Germany.
[43] D.L. Reger, R.F. Semeniuc, C. Pettinari, F. Luna-Giles, M.D. Smith, The cartesian
coordinates of the ligand L1 were derived from the experimentally determined
crystal structure of a ruthenium dimer, Crystal Growth Des. 6 (2006) 1068–1070.
[44] P. Valentão, E. Fernes, F. Carvalho, P.B. Andrade, R.M. Seabra, M.L. Bastos, J. Agric,
Food Chem. 49 (2001) 3476–3479.
[45] M. Paya, B. Halliwell, J.R.S. Hoult, Biochem. Pharmacol. 44 (1992) 205–214.
[46] H. Zhang, J.M. Gurney, D. Becker, B. Kalyanaraman, J. Biol. Chem. 277 (2002)
1013–1020.
[4] D. Salvemini, Z.-Q. Wang, J.L. Zweier, A. Samouilov, H. Macarthur, T.P. Misko, M.G.
Currie, S. Cuzzocrea, J.A. Sikorski, D.P. Riley, Science 286 (1999) 304–306.
[5] C.X. Zhang, S.J. Lippard, Curr. Op. Chem. Biol. 7 (2003) 481–489.
[6] D.P. Riley, Chem. Rev. 99 (1999) 2573–2588.
[7] H. Sakurai, Y. Kojlwe, Y. Yoshikawa, K. Kewebe, H. Yosul, Coord. Chem. Rev. 226
(2002) 167–178.
[8] B.J. Day, Antioxid. Redox Signal 10 (2008) 355–370.
[9] V. Afonso, R. Champy, D. Mitrovic, P. Collin, A. Lomri, Joint Bone Spine. 74 (2007)
324–329.
[47] W.J. Geary, Coord. Chem. Rev. 7 (1971) 81–122.
[48] K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordination
Compounds, 6th Ed., Part B, John Wiley & Sons, Hoboken, New Jersey, 2009,
pp. 62–67, and references reported therein.
[49] C. Pettinari, Scorpionates II: Chelating Borate Ligands, Imperial College Press,
London, 2008.
[50] C. Pettinari, F. Marchetti, R. Pettinari, A. Marinelli, A. Crispini, A. Bellusci, Inorg.
Chim. Acta 362 (2009) 4593–4598.
[51] E. Finkelstein, G.M. Rosen, E.J. Raukman, J. Am. Chem. Soc 102 (1980) 4994–4999.
[52] E.M. Fielden, P.B. Roberts, B.C. Bray, D.J. Lowe, G.N. Mautner, G. Rotilio, L.
Calabrese, Biochem. J. 139 (1974) 49–60.
[53] T. Offer, A. Russo, A. Samuni, FASEB J. 14 (2000) 1215–1223.
[54] A. Formigari, P. Irato, A. Santon, Comp. Biochem. Physiol., Part C 146 (2007) 443–459.
[55] D. Bagchi, M. Bagchi, S.J. Stohs, Gen. Pharmac. 28 (1997) 85–91.
[56] Q. Wang, Y. Wang, Z. Yang, Chem. Pharm. Bull. 56 (2008) 1018–1021.
[57] T.R. Li, Z.Y. Yang, B.D. Wang, Chem. Pharm. Bull. 55 (2007) 26–32.
[58] D. Qin, Z. Yang, G. Qi, Spectrochim. Acta A 74 (2009) 415–420.
[59] C. Muscoli, S. Cuzzocrea, D.P. Riley, J.L. Zweier, C. Thiemermann, Z.Q. Wang, D.
Salvemini, Br. J. Pharmacol. 140 (2003) 445.
[10] A.S. Potapov, E.A. Nudnova, G.A. Domina, L.N. Kirpotina, M.T. Quinn, A.I.
Khlebnikov, I.A. Schepetkin, Dalton Trans. 23 (2009) 4488–4498.
[11] M. Cejuedo, S. Alzuet, S. Ferrer, J. Borra, A. Castineiras, A. Mozzani, L. Casella, Inorg.
Chem. 43 (2004) 6805–6814.
[12] V.S. Sprakel, M.C. Feiters, W. Meyer-Klaucke, M. Klopstra, J. Brinksma, B.L. Feringa,
K.D. Karlin, R.J. Nolte, Dalton Trans (2005) 3522–3527.
[13] S. Tabassum, S. Parveen, F. Arjmand, Indian J. Chem. 43A (2004) 270–277.
[14] S. Tabassum, S. Parveen, F. Arjmand, Trans. Met. Chem. 30 (2005) 196–204.
[15] F. Arjmand, S. Parveen, S. Tabassum, Trans. Met. Chem. 30 (2005) 128–135.
[16] I. Schepetkin, A. Potapov, A. Klebnikov, E. Korotkova, A. Lukina, G. Malovichko, L.
Kirpotina, M.T. Quinn, J. Biol. Inorg. Chem. 11 (2006) 499–513.
[17] P. Chen, K. Fujisawa, E.I. Solomon, J. Am. Chem. Soc. 122 (2000) 10177–10193.
[18] G. Parkin, Chem. Rev. 104 (2004) 699–768.
[19] U.P. Singh, A.K. Sharma, P. Tyagi, S. Upreti, R.K. Singh, Polyhedron 25 (2006)
3628–3638.
[20] A.D. Schofield, M.L. Barros, M.G. Cushion, A.D. Schwarz, P. Mountford, Dalton Trans.
(2009) 85–96.
[21] A. Beck, B. Weibert, N. Burzlaff, Eur. J. Inorg. Chem. (2001) 521–527.
[22] I. Hegelmann, A. Beck, C. Eichhorn, B. Weibert, N. Burzlaff, Eur. J. Inorg. Chem.
(2003) 339–347.
[60] E. Esposito, S. Cuzzocrea, Front. Biosci. 14 (2009) 263.
[61] Y. Xu, B. Liu, J.L. Zweie, G. He, J. Pharmacol. Exp. Ther. 327 (2008) 402.