Synthesis, structural characterisation and solution chemistry of ruthenium(III) triazole-thiadiazine complexes

Article abstract of DOI:10.1039/b823271g

Two ruthenium(III) complexes structurally similar to the anticancer compound NAMI were prepared: Na[RuCl₄(DMSO)(L1)] (1) and Na[RuCl ₄(DMSO)(L2)] (2), where L1 and L2 are differently functionalised triazole-thiadiazine ligands. To fa

Full text of DOI:10.1039/b823271g

View Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Synthesis, structural characterisation and solution chemistry
of ruthenium(III) triazole-thiadiazine complexes†
a
a
a
b
b
Massimiliano Delferro, Luciano Marchi o` ,* Matteo Tegoni, Saverio Tardito, Renata Franchi-Gazzola and
Maurizio Lanfranchi
a
Received 24th December 2008, Accepted 25th February 2009
First published as an Advance Article on the web 20th March 2009
DOI: 10.1039/b823271g
Two ruthenium(III) complexes structurally similar to the anticancer compound NAMI were prepared:
Na[RuCl
4
(DMSO)(L1)] (1) and Na[RuCl (DMSO)(L2)] (2), where L1 and L2 are differently
4
+
functionalised triazole-thiadiazine ligands. To facilitate the crystallisation of the complex anions, Na
was substituted with the [bis(triphenylphosphoranylidene)ammonium] cation (PPN ), allowing the
+
X-ray characterisation of PPN[RuCl
4
(DMSO)(L1)]·2H
2
O (1a·2H
2
O) and PPN[RuCl
4
(DMSO)(L2)]·
3H
2
O (2a·3H
2
O), respectively. The two compounds undergo stepwise hydrolytic processes, as assessed
1
by means of UV-vis and H NMR spectroscopy. The first hydrolytic step consists of the replacement of
a chloride anion with a water molecule, with a half-life of 50 min (1) and 110 min (2), while the
subsequent hydrolytic steps are more complicated to describe since more than one product is generated
at the same time. The redox potential of the Ru(III)/Ru(II) couple (0.31 V for 1 and 0.28 V for 2)
suggests that these complexes can be reduced in the intracellular environment, in agreement with the
“
activation by reduction” mechanism proposed for NAMI and NAMI-A. 1 and 2 were tested on a
human cancer cell line derived from a fibrosarcoma (HT1080), and on non-cancerous primary human
fibroblasts (HF), where they showed a modest inhibitory effect.
Introduction
Ruthenium complexes have received considerable attention in
the past two decades, since they have proven to be a valid
alternative to cis-platin as anticancer compounds against spe-
1,2
cific tumours. The imidazole/sulfoxide complexes NAMI and
3
NAMI-A (Scheme 1) are antimetastatic agents whose activity is
4,5
not associated with a direct cytotoxicity in tumour cells. The
+
substitution of Na with the imidazolium ion confers to NAMI-A
a greater chemical stability with respect to NAMI. NAMI-A has
6
successfully completed phase I clinical trials. The other family of
Ru(III) complexes endowed with anticancer activity is represented
7,8
by ICR-like complexes, where two nitrogen donor ligands are
located in the trans positions (Scheme 1). In contrast to NAMI-
type complexes, these latter compounds are cytotoxic but do not
9
exhibit anti-metastatic activity. In particular, KP1019 is active
against colorectal cancer, where it induces apoptosis. KP1019
has recently completed phase I clinical trials for the treatment
10
of colorectal cancer.
Scheme 1
a
Dipartimento di Chimica Generale ed Inorganica, Chimica Analitica,
Chimica Fisica, Universit a` degli Studi di Parma, Viale G.P. Usberti 17/A,
I-43100, Parma, Italy. E-mail: marchio@unipr.it; Fax: +39 0521 905424;
Tel: +39 0521 905424
Dipartimento di Medicina Sperimentale, Sezione di Patologia Generale e
Clinica, Universit a` degli Studi di Parma, Via Volturno 39, I-43100, Parma,
Italy
Despite many efforts dedicated to the mechanism of action of
NAMI-A and KP1019, a clear picture of the machinery behind
the anticancer properties of these drugs is not yet complete.
There are nevertheless some relevant aspects that have been
established, where these are related to (i) the redox properties of
the Ru(III)/Ru(II) couple and (ii) the hydrolysis of the complexes.
Ru(III) species behave as pro-drugs that can be reduced in the
hypoxic environment found inside tumour tissue, thus generating
labile and more active Ru(II) species that can interact with
b
†
Electronic supplementary information (ESI) available: Experimental and
1
calculated FAR-IR spectrum of 1; ESI-MS spectra of 1 and 2; H NMR
spectra of 1 and 2 in D O (time range 0–180 min); solvolysis of 1 and 2 in
2
buffered water (pH 7.4); solvolysis of 1 in DMSO; cyclic voltammogram
of 2; antiproliferative activity of 1 and 2 on HF and HT1080 cells. CCDC
reference numbers 713340 and 713341. For ESI and crystallographic data
in CIF or other electronic format see DOI: 10.1039/b823271g
11,12
biomolecules such as DNA.
This is the so called “activation by
13
reduction” mechanism. In fact, both NAMI-A and KP1019 have
3
766 | Dalton Trans., 2009, 3766–3773
This journal is © The Royal Society of Chemistry 2009

View Online
redox potentials that allow for the reduction of the metal center
Results and discussion
3,14
in a biological environment.
In parallel with this, the other
Synthesis and characterisation of Na[trans-RuCl (DMSO)(L1)]
4
process that characterises these Ru(III) complexes is the stepwise
hydrolysis of chloride ions and, in the case of NAMI-A, also of
(
1) and Na[trans-RuCl
4
(DMSO)(L2)] (2)
15,16
DMSO.
This is an important aspect since it appears that the
The ligands 6-methyl-7H-[1,2,4]triazolo[3,4-b][1,3,4]thiadiazine
loss of chloro ligands is a requisite for the biological activity of
this type of complex. The hydrolysis is even more favoured when
Ru(III) has been converted to the more labile Ru(II).
(
(
L1) and 3,6-dimethyl-7H-[1,2,4]triazolo[3,4-b][1,3,4]thiadiazine
L2) can be prepared in high yield in ethanol by reacting
chloroacetone with 4-amino-1,4-dihydro-1,2,4-triazole-5-thione
or 4-amino-1,4-dihydro-3-methyl-1,2,4-triazole-5-thione, respec-
tively (Scheme 2). The parent ligands can adopt different coor-
dination modes with metal centers, the most common of which
is represented by N,S chelation with the thione and amino
functions.
The mechanism through which the Ru(III) complexes are
transported once dissolved in biological fluids has also been
extensively investigated, and it was shown that proteins such
as albumin and transferrin are responsible for the transport of
17–19
ruthenium species to their site of action.
26–30
The conversion of the thione group into a thioether
Furthermore, the electrochemical and hydrolytic properties of
Ru(III) complexes are strongly dependent on the type of donor
atoms surrounding the metal center, and this has stimulated the
synthesis of a number of Ru(III) complexes that are structurally
induces a dramatic decrease of the coordination capability of the
sulfur atom, and the nitrogen atoms of the triazole ring then
become the preferred donor sites.
13
31,25
The Ru(III) complexes 1 and 2 were prepared starting from
1
4,20–22
8,14,20,23,24
similar to NAMI-A
Herein, we report the synthesis and X-ray structures of two
novel NAMI-like Ru(III) complexes, Na[RuCl (DMSO)(L1/L2)],
where L1 and L2 are N-donor triazole-thiadiazine ligands
Schemes 2 and 3). The molecular structures and biological activity
of Cu(II) complexes with sulfurated-triazolo ligands was recently
or ICR.
the ruthenium precursor Na[RuCl
procedure reported in the literature for the synthesis of NAMI
4
(DMSO) ], in analogy to the
2
3
4
(
Scheme 3).
The solution characterisation of 1 and 2 was performed by
(
1
means of ESI-MS spectrometry and H NMR spectroscopy.
The ESI-MS spectra of 1 and 2 dissolved in acetonitrile show
25
reported, and it was concluded that the N,S coordination to
Cu(II) of the thione group of one of these ligands enhances the
cytotoxic activity of the resulting complex towards cancer cells. By
converting this thione function to thioether groups, as in the case of
L1 and L2 ligands, the cytotoxic potential severely diminishes but
the cell death process is nonetheless triggered. In the present work
we wished to evaluate the biological activity of Ru(III) complexes
with sulfurated-triazolo ligands that were able to yield complexes
structurally similar to NAMI and NAMI-A.
-
the peaks of the [RuCl
4
(DMSO)(L1/L2)] anionic complexes
-
and the corresponding molecular fragments: [RuCl
4
(DMSO)] ,
-
-
1
[
RuCl
4
(L1/L2)] , and [RuCl
4
] (see ESI†). The H NMR spectra
of 1 and 2 in D
2
O present several broad peaks, in accordance
with the presence of paramagnetic Ru(III) metal ions (Fig. 1). The
assignment of the signals of 1 and 2 was performed by comparison
1
with the H spectra of similar Ru(III) complexes and by means of
the inversion recovery technique, which provides the longitudinal
32
relaxation time T1. In fact, nuclei that are close to Ru(III) will
exhibit a shorter T1 if compared with nuclei that are at a certain
distance from the paramagnetic center, such as solvent molecules.
In the present case, the determination of T1 values is hindered
by the overlap of the broad signals, and only the peak at -1 ppm
of 2 could be properly analysed. The T1 value of this signal is
~
10 ms, considerably smaller than that of the acetone present as an
impurity (~200 ms). It is nonetheless evident from the experiment
that the remaining broad peaks exhibit comparable T1 values.
The broad signal at ~-15 ppm is consistent with S-coordinated
DMSO as reported for other NAMI-like complexes. A tentative
assignment of L1 and L2 signals was performed by considering
that their paramagnetic shift depends on their distance from
Scheme 2 Ligand synthesis.
3,22
the Ru(III) center: for 1, 0.9 ppm (–CH
2
–), 0.71 ppm (C–H),
–), -1 ppm
0
.62 ppm (–CH ); for 2, 1.1 ppm (CH ), 0.8 ppm (–CH
3
3
2
triazole
(
–CH
Any attempt to grow single crystals of 1 and 2 suitable for
X-ray structural determination were unsuccessful. Nevertheless,
3
).
Scheme 3 Synthesis of NAMI-like complexes 1 and 2.
+
+
when the Na cation was replaced with PPN [bis(triphenyl-
phosphoranylidene)ammonium], crystals of PPN[RuCl (DMSO)-
(L1)]·2H (1a·2H O) and PPN[RuCl (DMSO)(L2)]·3H
(2a·3H O) were obtained. The molecular structures of the complex
anion of 1a·2H O are depicted in Figs. 2 and
4
The hydrolysis and solvolysis of the complexes were investigated
by means of UV-vis spectroscopy in non-buffered, buffered
pH 7.4) aqueous solution and in DMSO. In addition, in order
to gain information on the electrochemical properties of the
complexes, cyclic voltammetry experiments were also performed.
The cytotoxic effect of both compounds was studied on normal
cells (human fibroblast, HF) and on tumour cells derived from a
fibrosarcoma (HT1080).
2
O
2
4
2
O
2
(
2
O and 2a·3H
2
3, respectively. Relevant geometric parameters are reported in
Table 1.
In both compounds, the metal adopts an octahedral geometry,
and the S-coordinated DMSO is in the trans position to L1
or L2, and four chloride ions occupy the equatorial positions.
This journal is © The Royal Society of Chemistry 2009
Dalton Trans., 2009, 3766–3773 | 3767

View Online
Fig. 2 Ortep drawing of 1a·2H
2
O. Thermal ellipsoids are drawn at the
0% probability level. The PPN cation and the crystallization water
+
3
molecules were omitted for clarity.
1
Fig. 1 H NMR spectra of 1 (above) and 2 (below) together with the
inversion recovery experiments. Asterisks define protons of the ligands L1
and L2.
Table 1 Selected bond lengths [ A˚ ] for 1a·2H
a·2H
O and 2a·3H
2
O
2
1
2
O
Ru-N(11)
Ru-S(12)
Ru-Cl(1)
P(1)-N(12)
2.079(8)
2.310(3)
2.363(3)
1.582(7)
Ru-Cl(2)
Ru-Cl(3)
Ru-Cl(4)
P(2)-N(12)
2.372(3)
2.323(3)
2.344(3)
1.602(7)
2
a·3H
2
O
Fig. 3 Ortep drawing of 2a·3H
2
O. Thermal ellipsoids are drawn at the
0% probability level. The PPN cation and the crystallization water
+
3
Ru-N(21)
Ru-S(12)
Ru-Cl(1)
P(1)-N(12)
2.134(2)
2.2981(9)
2.3734(9)
1.579(2)
Ru-Cl(2)
Ru-Cl(3)
Ru-Cl(4)
P(2)-N(12)
2.3477(9)
2.3677(9)
2.3700(9)
1.574(2)
molecules were omitted for clarity.
coordination of N(11) with respect to N(21), whereas for L2,
the steric hindrance would not constitute a discriminating factor
for the N(11) and N(21) coordination if we assume that the van
der Waals radii of a sulfur atom and a methyl group are nearly
equivalent. Additional proof that the steric hindrance may play a
role in the preferred coordination of the complex derives from the
observation that, in both compounds, the molecular plane of the
Interestingly, L1 and L2 are coordinated to the metal through two
different nitrogen atoms of the triazole ring, where this may be
attributed to the comparable donor ability of the two nitrogen
atoms. The discriminating factor in determining which nitrogen
atom participates in the metal coordination may be due to the
different steric hindrance that is present close to the N(11) atoms
of L1 and L2 and by the perturbation that this would cause on
triazole ligand is bisecting the Cl(1)–Ru–Cl(2) angle. In 1a·2H
2
O,
the O1w water molecule is involved in hydrogen bonds (HBs) with
˚
the RuCl
4
equatorial plane: a hydrogen atom in L1 favours the
O2w (d[O1w–O2w] = 2.83(1) A) and with the oxygen atom of a
3
768 | Dalton Trans., 2009, 3766–3773
This journal is © The Royal Society of Chemistry 2009

View Online
Table 2 UV-vis data (lmax and e) for the Ru(III) complexes
max (nm), e
of spectra can be restricted to the 0–170 min interval (Fig. 4,
upper diagram), and it is characterised by a decrease of the band
at 398 nm and a clear isosbestic point at 370 nm. After this time, the
band at 398 nm tends to decrease, but the isosbestic point at 370 nm
is lost and a new broad band appears at 545 nm in spectra up to
l
lmax (nm), e
Na[RuCl
Na[RuCl
NAMI
4
(L1)(DMSO)]
(L2)(DMSO)]
398, 3690
398, 3418
390, 3644
485, 590
485, 600
451, 488
4
a
5
0 h (Fig. 4, lower diagram). Although two apparent isosbestic
a
points appeared at 366 and 480 nm in the 540–3000 min time
range, a careful evaluation of the spectra through the Evolving
ref. 3.
39,40
Factor Analysis
revealed that they are not proper isosbestic
DMSO ligand of an adjacent complex molecule (d[O1w–O(12)’] =
points and that at least three different species should be present in
˚
solution. For all processes, however, the pCl→t Ru LMCT band
2
.881(1) A; ¢ = x, y, z - 1). In addition, the O2w molecule acts as a
2g
HB donor with respect to the sulfur atom of L1 (d[O2w–S(11)] =
centered at 398 nm tends to decrease, suggesting that the reactions
˚
occurring at the metal center most likely involve the replacement
of chloride ions with water molecules, as observed for other
3
.39(1) A). In 2a·3H
2
O, the O1w water molecule acts as a HB
˚
donor with respect to Cl(1) (d[O1w–Cl(1)] = 3.444(3) A) and O2w
20,22
1
˚
NAMI-A-type complexes.
Following the hydrolysis with H
(
d[O1w–O2w] = 2.859(5) A), whereas the O3w water molecule
acts as a HB donor with respect to O2w (d[O3w–O2w] = 2.873(5)
NMR experiments in the 0–3 h time range, the appearance of the
uncoordinated DMSO/ligand signals indicate that the hydrolysis
also affects these moieties, albeit to a lesser extent with respect to
the chloride hydrolysis. In fact, the signals of the free ligands do
not increase significantly with respect to those of the coordinated
ones (ESI†). While the process occurring in the 0–170 min time
range produce only a small change in the shape of the absorption
spectra, it is worth noting that the subsequent reactions produce a
marked spectral change, as evidenced by the appearance of the new
band at 540 nm. Moreover, the processes that occur in the 540–
˚
˚
A) and N(11) (d[O3w–N(11)] = 2.885(4) A).
Far-infrared spectra were recorded for 1 and 2 (see ESI†), and
the assignment of the Ru–Cl stretching frequencies was made on
7,33–37
the basis of FAR-IR spectra of similar compounds.
The L1
and L2 coordination at the metal atom causes a shift to higher
values of the Ru–S stretching vibration frequency with respect to
-
1
the Ru(III) precursor. The new peaks that appear around 450 cm
can be attributed to the Ru–N stretching.
3
000 min interval may involve a more extensive hydrolysis with
Solution stability assays (Hydrolysis and solvolysis).
substitution of the triazole, chloride or DMSO ligands, giving rise
to a more complicated spectral set. The spectral behaviour of 2 is
analogous to that of 1, and the spectral changes over a 50 h time
course are reported in the ESI.†
The spectra collected for 1 and 2 in the 0–170 min time interval
were used to calculate the kinetic rate constant for the first process
by means of a least squares regression on all of the 270–700 wave-
lengths. The spectra could be fitted with a first order kinetic law
Compounds 1 and 2 are soluble in water, acetonitrile and DMSO,
and the UV-vis spectra recorded in water or acetonitrile are
characterised by relatively intense bands at ~400 nm, and by
a second and weaker band at ~485 nm (Table 2). The former
band can be ascribed to pCl→t2gRu LMCT (Ru-Cl) transitions
38
involving the four coplanar chlorides. These findings are in line
3,23
with the data reported for NAMI or NAMI-like complexes.
In order to investigate the hydrolysis and solvolysis processes, 1
and 2 were dissolved in water or DMSO and their UV-vis spectra
were recorded over a period of 50 h (water, Fig. 4), 24 h (DMSO,
ESI†), or 3 h (buffered water pH = 7.4, ESI†). In water, different
families of spectra can be identified for both complexes, indicating
at least two/three different consecutive reactions. The first family
(
v = k[complex]) and, for both complexes, the absorbance residuals
after the fitting process were less than 1% over the entire spectral
range. A plot of the experimental and calculated absorbance at
two representative wavelengths (342 and 398 nm) is reported in
Fig. 5.
4
Fig. 4 Upper diagram: UV-vis spectra of 1 Na[RuCl (DMSO)(L1)] in
water at 7.5 min intervals (0–170 min after dissolution). Lower diagram:
UV-vis spectra of 1 in water at 180 min intervals (540–3000 min after
dissolution).
Fig. 5 Experimental (dots) and calculated (lines) absorbances for the
hydrolytic reaction of Na[RuCl
4
(DMSO)(L1)] (1) in water (0–170 min
after dissolution) at 342 and 398 nm.
This journal is © The Royal Society of Chemistry 2009
Dalton Trans., 2009, 3766–3773 | 3769

View Online
Attempts to fit the spectra in the range 170–3000 min with
different kinetic models gave inconsistent results, suggesting that
several processes may occur at the same time. Although the
spectral change for the two complexes is similar, the reaction rate
for the first process (0–170 min) for complex 1 is ca. twice that
-
4
-1
of complex 2. In fact, k is 2.33(7) ¥ 10
s
for 1, and 1.05(7) ¥
-
4
-1
10
s
for 2, corresponding to a half-life of 50 and 110 min,
respectively. These rate constants are within the same order of
magnitude as expected by assuming the same hydrolytic process
for both complexes, and their difference can be ascribed to the
different coordination behaviour of L1 vs. L2 to the metal center.
These rate constants are one order of magnitude higher than
41
those determined for NAMI-A in acidic buffers. On the other
hand, when the hydrolysis is measured in buffered water solution
at physiological pH (7.4), the first hydrolysis process is much
faster, having an half-life of approximately 9 min (ESI†). This
is in agreement with the values determined for the NAMI-A-like
-
4
Fig. 6 Cyclic voltammograms of a 5 ¥ 10 M solution of 1 dissolved in
water. Supporting electrolyte: KNO 0.1 M. Scan rate: 10 mV/s. First scan
black line), second scan (gray line).
3
(
22
20
complexes with substituted triazolo-pyrimidine and thiazole
1
7
ligands in buffered solutions (pH 7.4). The hydrolysis profile of 1
and 2 at pH 7.4 is characterized in the first step by the decrease
of the bands centered at 395, and 465 nm, and by the increase of
the band centered at 350 nm. In a second stage, this latter band
decreases, and additional bands centered at 330, 440 and 575 nm
prevail.
containing nitrogen donor ligands. The redox behaviour of
2 is analogous to that of 1, with the redox potential for
the [RuCl (DMSO)(L2)] /[RuCl (DMSO)(L2)] couple being
0.28 V. The cyclic voltammogram of 2 recorded at 25 mV/s is
-
2-
4
4
reported in the ESI.†
The solvolysis in DMSO was investigated due to the fact that
in some cases, this solvent is employed for cytotoxic studies
at a concentration of 9ꢀ in water. The spectrophotometric
profiles of 1 and 2 as a function of time are reported in the
ESI.† After 2 h, the band at 410 nm shifts to 421 nm and the
Antiproliferative studies
Complexes 1 and 2 were tested on the human cancer cell line
derived from a fibrosarcoma (HT1080), and on non-cancerous
primary human fibroblasts (HF) in order to evaluate their effect on
cell proliferation. The compounds tested show a modest inhibitory
effect on the viability of both cell types (ESI†). At the highest
concentration tested (100 mM), all complexes cause a ª20% and
spectrum becomes superimposable with that of the precursor
-
[
trans-RuCl
4
(DMSO)
2
] . This may imply that L1 and L2 in 1 and 2,
respectively, are replaced by DMSO molecules within a few hours.
1
0% decrease in the viability in HT1080 and HF, respectively,
This is probably a consequence of the trans-effect exerted by the
DMSO molecules of the complexes.
42
when compared to untreated cells. This effect is possibly due to a
decrease in the proliferation rate of both normal and cancer cells,
since no signs of cytotoxicity were evident in treated cell cultures.
It is known that NAMI-A exerts its antitumour effects by
reducing the metastatic potential of cancer cells. Its mechanism
of action is related to the ability to decrease the motility of cells, im-
pairing their invasiveness. For this reason, its antimetastatic activ-
ity is not generally associated with an antiproliferative/cytotoxic
effect. Thus far, we cannotexclude the effectiveness ofthesenew Ru
complexes as antimetastatic agents, and more focused biological
studies are needed to prove this hypothesis.
Electrochemistry
45
4
3
The redox behaviour of NAMI-A was extensively investigated,
◦
and the formal redox potential E ’ measured in different aqueous
solutions varied in the range of +0.268/+0.254 V. These values
have important consequences for the biological properties of the
compound. In fact, they imply that the Ru(III) → Ru(II) reduction
is possible in the intracellular environment of proliferating cells.
This is due to the high concentration of glutathione in the cytosol
4
4
and to the hypoxic environment present in tumour tissues.
Following the reduction, NAMI-A first undergoes hydrolysis of
the imidazole ligand and, in a second stage, a chloride ion is
substituted by a water molecule.
Conclusions
43
Ru(III) complexes are actively investigated as valid alternatives to
cis-platin as anticancer compounds. The most promising ruthe-
nium compounds are represented by NAMI-A and KP1019, which
1
The redox behaviour of 1 and 2 is very similar to that of
NAMI-A. In Fig. 6, the cyclic voltammogram of 1 in non-
buffered water is shown. In the first scan, a reduction peak
is observed at 0.27 V (peak A), which is associated with the
oxidation peak at 0.36 V (peak B). After the reduction, some
of the Ru(II) complex undergoes hydrolysis and gives rise to an
additional oxidation peak in the reverse scan at 0.51 V (peak
C), which is then associated with a new reduction peak at 0.42 V
6,10
have recently completed phase I clinical trials. The mechanism
of action of NAMI and NAMI-A has been partly interpreted, and
it involves Ru(III) → Ru(II) reduction in the intracellular environ-
ment followed by partial hydrolysis of the complex. Prompted by
these results, a number of Ru(III) compounds have been prepared
by modifying the nitrogen donor ligand, with the aim of improving
the biological properties and the therapeutic potential of the
resulting complexes. In the present work, two Ru(III) complexes,
which are structurally similar to NAMI, were prepared and
-
2-
(
peak D). The [RuCl
4
(DMSO)(L1)] /[RuCl (DMSO)(L1)] redox
4
couple that defines the A–B system exhibits a redox potential
of 0.31 V vs NHE, in line with NAMI-A-type complexes
3
770 | Dalton Trans., 2009, 3766–3773
This journal is © The Royal Society of Chemistry 2009

View Online
structurally characterised, namely Na[trans-RuCl
4
(DMSO)(L1)]
38.95; S, 20.80; H, 3.92; N, 36.34; found: C, 38.62; S, 20.72; H,
1
(
1) and Na[trans-RuCl (DMSO)(L2)] (2), where L1 and L2 are
4
3.89; N, 36.29. H NMR (DMSO-d
6
): d = 8.97 (s, 1H, CH), 3.87
-
1
triazole-thiadiazine ligands. The two complexes behave similarly
to NAMI-A once dissolved in water, as they are subjected
first to the hydrolysis of a chloride ion, and in later stages to
a more extensive aquation. Furthermore, the redox potential
of the Ru(III)/Ru(II) couple (0.31 and 0.28 V for 1 and 2,
respectively) would imply the reduction of the two complexes in
(s, 2 H, CH
2
), 2.26 (s, 3 H, CH
3
) ppm. IR (KBr, cm ): 3112 m,
3096 m, 2956 m, 2902 w, 1629 m, 1488 s, 1426 s, 1281 s, 1160 s,
1021 m, 948 s, 795 s, 641 s.
2H-3,6-dimethyl-[1,2,4]-triazole-[3,4-b]-[1,3,4]-thiadiazine (L2).
To a suspension of 4-amino-1,4-dihydro-3-methyl-1,2,4-triazole-
5-thione (2.095 g, 16 mmol) in absolute ethanol (30 ml) was added
chloroacetone (95%) (1.35 ml, 16 mmol) with stirring. The yellow
clear solution was refluxed for 4 h. The mixture was cooled to room
temperature and neutralised with sodium bicarbonate and vacuum
dried. The solid was then treated with absolute ethanol and
13
the hypoxic environment of tumour tissues. In order to evaluate
the antitumour properties of the two complexes, preliminary
antiproliferative studies were performed with normal cells (HF)
and tumour cells derived from a fibrosarcoma (HT1080). Both
complexes showed only modest antiproliferative activity against
HT1080 cells (20%), and more studies are necessary in order to
investigate their antimetastatic potential.
ethyl acetate (1:1), and the inorganic salts (NaHCO , NaCl) were
3
removed by filtration. The clear brown solution was concentrated,
and upon addition of diethyl ether, a brown precipitate formed
(
Yield: 89%). C
6
H
8
N
4
S (%): calcd C, 42.84; S, 19.06; H, 4.79;
Experimental
1
N, 33.31; found: C, 42.70; S, 18.99; H, 4.72; N, 33.26. H NMR
(
3
1
triazole
DMSO-d
6
): d = 3.80 (s, 2H, CH
2
), 2.37 (s, 3H, CH
3
), 2.28 (s,
General procedures
thiadiazine
-1
H, CH
3
) ppm. IR (KBr, cm ): 2991 w, 2962 m, 2900 m,
All solvents and reagents are commercially available and were used
as received, except Na[RuCl
540 s, 1471 s, 1441 s, 1394 s, 1296 m, 1017 m, 808 m.
34
4
(DMSO)
2
], 4-amino-1,4-dihydro-
29
1
,2,4-triazole-5-thione and 4-amino-1,4-dihydro-3-methyl-1,2,4-
Na[trans-RuCl
Na[trans-RuCl (DMSO)
(10 mL) was added to a solution of L1 (0.098 g, 0.64 mmol)
in CH CN (5 mL). The mixture was stirred overnight. A
brown precipitate formed, which was filtered, washed with an
ethanol/diethyl ether mixture (1:1), dried in a vacuum and
collected (Yield: 78%). C
16.87; H, 2.42; N, 11.24; S, 12.87; found: C, 16.75; H, 2.38;
N, 11.20; S, 12.81. H NMR: (D
4
(DMSO)(L1)] (1). An orange solution of
4
6
triazole-5-thione, which were prepared as described elsewhere.
Elemental analyses were carried out by Carlo Erba mod. EA1108
microanalyser. Infrared spectra were recorded on a Nicolet 5PC
4
2
] (0.180 g, 0.43 mmol) in CH CN
3
3
-
1
FT-IR spectrometer in KBr pellets (4000–400 cm ) and in CsI
-
1
pellets (600–200 cm ). Electrospray ionization mass spectrometry
ESI-MS) was carried out in acetonitrile with a Micromass
(
7
H
12Cl
4
N
4
ONaRuS (498.20): calcd. C,
2
-
4
ZMD instrument (conc. = 2 ¥ 10 M for all complexes); the
given m/z values, originating from the most intense isotope,
were obtained by the mass linearization procedure. Expected and
experimental isotope distributions were compared in order to
assign molecular fragment peaks. H NMR spectra were recorded
on a Bruker Avance 300 spectrometer, and chemical shifts are
reported in ppm referenced to residual solvent protons (DMSO-
1
2
O): d = 0.89 (broad), 0.71
-1
(broad), 0.62 (broad), -14.52 (broad, DMSO) ppm. IR (KBr cm ):
3114 m, 3005 w, 2919 w, 2896 w, 1635 w, 1505 s, 1459 s, 1419
m, 1384 w, 1312 w, 1096 s, 1021 s, 971 w, 784 m. Far-IR
1
-
1
(CsI cm ): 458, 436, 364, 343. ESI-MS (CH
3
CN, negative ion
-
-
mode): m/z = 475.7 [RuCl
4
(DMSO)(L1)] , 395.8 [RuCl
4
(L1)] ,
-
-
d
6
and D
2
O). The paramagnetic Ru(III) complexes 1 and 2 were
O using the inversion recovery pulse technique:
321.8 [RuCl
4
(DMSO)] , 243.6 [RuCl
4
] .
characterised in D
2
PPN[trans-RuCl
.010 mmol) was dissolved in water (2 mL). To the orange solution,
4
(DMSO)(L1)]·2H
2
O (1a·2H
2
O). 1 (0.050 g,
p–t–p/2-FID. 10 spectra were recorded with t that varied in
the range of 0.00001–10 sec. Cyclic voltammetry was performed
using an Autolab PGSTAT 20 potentiostat (Ecochemie, The
Nethrlands) with GPES 4.9 software in a one-compartment three-
electrode cell using a 3.0-mm diameter platinum disk working
electrode, a platinum rod counter electrode and Ag/AgCl/KCl
0
+
-
bis(triphenylphosphoranylidene)ammonium chloride (PPN Cl )
(
0.057 g, 0.01 mmol) dissolved in 1 ml of water/ethanol (1:0.2 v:v)
was added with stirring. Light-orange crystals of PPN[trans-
RuCl O (1a·2H O) suitable for X-ray analysis
(DMSO)(L1)]·2H
formed immediately (Yield: 73%).
4
2
2
◦
◦
(
3M) as a reference electrode (E = +211.5 mV at 20 C). Cyclic
voltammograms were collected in nonbuffered water solution
using KNO as the supporting electrolyte. The voltammograms
were referenced to NHE.
Na[trans-RuCl
Na[trans-RuCl (DMSO)
10 mL) was added to a solution of L2 (0.107 g, 0.64 mmol)
4
(DMSO)(L2)] (2). An orange solution of
3
4
2
] (0.180 g, 0.43 mmol) in CH CN
3
(
2
H-3-methyl-[1,2,4]-triazole-[3,4-b]-[1,3,4]-thiadiazine
To a suspension of 4-amino-1,4-dihydro-1,2,4-triazole-5-thione
2.734 g, 23.5 mmol) in absolute ethanol (30 mL) was added
(L1).
in CH CN (5 mL). The mixture was stirred overnight. A
brown-orange precipitate formed, which was filtered, washed
with ethanol/diethyl ether, and dried in vacuo. (Yield: 68%).
3
(
chloroacetone (95%) (2 mL, 23.5 mmol) with stirring. The brown
clear solution was refluxed for 4 h. The mixture was cooled to
room temperature and neutralised with sodium bicarbonate and
vacuum dried. The solid was then treated with absolute ethanol
C
8
H
14Cl
4
N
4
NaORuS
S, 12.52. Found: C, 18.62; H, 2,71; N, 10.90; S, 12,45. H NMR
(D
2
(512.23): calcd. C, 18.75; H, 2.75; N, 10.93;
1
2
O): d = 1.10 (broad), 0.71 (broad), -1.0 (broad), -14.5 (broad,
-1
DMSO) ppm. IR (KBr, cm ): 3018 w, 2964 w, 2925 w, 1558 m,
1471 m, 1428 m, 1308 m, 1292 m, 1095 s, 1025 s, 685 w. Far-IR
(CsI, cm ): 448, 431, 356, 332. ESI-MS (CH
and ethyl acetate (1:1), and the inorganic salts (NaHCO , NaCl)
3
-
1
were removed by filtration. The volatiles were removed in vacuo.
The brown sticky product was triturated with hexane, which gave
3
CN, negative ion
-
-
mode): m/z = 491.9 [RuCl
4
(DMSO)(L2)] , 404.9 [RuCl
4
(L2)] ,
-
-
a brown precipitate (Yield: 91%). C
5
H
6
N
4
S (154.19): calcd. C,
321.8 [RuCl
4
(DMSO)] , 243.6 [RuCl
4
] .
This journal is © The Royal Society of Chemistry 2009
Dalton Trans., 2009, 3766–3773 | 3771

View Online
Table 3 Crystallographic data of 1a·2H
2
O and 2a·3H
2
O
UV-visible and solution studies
1
a·2H
2
O
2a·3H
2
O
UV-Vis spectra were recorded in the 270–700 nm range on Perkin-
Elmer lambda 25 and Varian Cary 50 spectrophotometers using
matched, teflon-capped quartz cuvettes (d = 1 cm). Solvolysis of 1
and 2 in water was studied by dissolving the complexes in doubly
Formula
Fw
C
43
H
46Cl
4
N
5
O
3
P
2
RuS
2
C
44
H
50Cl
4
N
5
O
4
P
2
RuS
2
1049.78
Monoclinic
Cc
10.277(4)
31.887(9)
14.738(7)
90
104.333(9)
90
4679(3)
4
1.490
Mo Ka (0.71073 A˚ )
293(2)
1081.82
Monoclinic
P2 /c
cryst. syst.
Space group
a,A˚
b,A˚
c,A˚
a
b
-
4
1
distilled water (CRu = 2.5 ¥ 10 M) and collecting the spectra over
time using dedicated Varian software (every 2.5 for 0–540 min, and
every 30 min for 540–3000 min). Solvolysis of 1 and 2 in buffered
water solution (HEPES 50 mM, NaCl 0.9%, pH = 7.4) was
performed in the 0–300 min time range collecting the spectra every
30 sec. Solvolysis of 1 and 2 in DMSO was studied by dissolving the
11.063(2)
14.691(3)
30.977(6)
90
99.83(2)
90
4961(2)
4
1.449
Mo Ka (0.71073 A˚ )
293(2)
0.726
53967/11291
0.0730
g
V,A˚
3
-4
complexes at CRu = 2 ¥ 10 M and collecting the spectra 1, 3, 6, 8,
Z
D/Mgm-3
24, 32, 48, 72 and 96 h after dissolution. Calculations of the kinetic
rate constants for the solvolysis processes in water were performed
l
40
T/K
using the software SPECFIT/32, using the whole spectral range
270–700 nm).
-1
m, mm
0.765
24350/8320
0.1118
545
0.688
(
Rflc collected/unique
R
int
No. of params. refined
GOF
581
1.036
Cell lines and culture conditions
R
1
(obs.), wR
2
0.0459, 0.0560
0.0431, 0.0582
0.730 and -0.464
Human fibroblasts (HF) were derived from a skin biopsy from
an healthy donor and were stored in liquid nitrogen at a stock
of 10 cells/ml. HT1080, a cell line originating from a human
fibrosarcoma, were provided from the “Instituto Zooprofilattico di
Brescia”. HT1080 cells and HF were routinely grown in Dulbecco’s
Modified Eagle’s Medium (DMEM) supplemented with 100 U/ml
Lrgst diff. peak/hole A˚ 0.480 and -0.375
-3
ꢀꢀF
ꢀ/ꢀ|F
= [ꢀw(F
ꢀwF
1
2
6
2
2
2
4
R
1
=
o
| - |F
c
o
|; wR
2
o
- F
c
) /
o
]
PPN[trans-RuCl
RuCl (DMSO)(L2)] (0.05 g, 0.0097 mmol) was dissolved
in water (2 mL). To the orange solution, bis(triphenyl-
4
(DMSO)(L2)]·3H
2
O (2a·3H
2
O). Na[trans-
penicillin, 100 mg/ml streptomycin and 10% fetal bovine serum
4
◦
(
FBS). Cells were cultured at 37 C in a 5% CO
2
atmosphere at
+
-
pH 7.4.
phosphoranylidene)ammonium chloride (PPN Cl ) (0.056 g,
.0097 mmol), dissolved in 1 ml water/ethanol (1:0.2 v:v),
was added with stirring. Yellow crystals of PPN[trans-
RuCl O) suitable for X-ray analysis
0
Viability assay
4
(DMSO)(L2)]·3H
2
O (2a·3H
2
The cells used for this study were in the exponential growth phase.
The cells were seeded in 96-well plates at a density of 10000 and
were immediately obtained (Yield: 70%).
4
000 cells/well for HF, and 4000 for HT1080, in 100 ml of medium
Crystallographic structure determination
and incubated for 24 h. The growth medium was then substituted
with a fresh one containing the substances to be tested at the
experimental concentrations. 1 and 2 were dissolved in 0.9% NaCl
physiological solution. 6 wells were treated with vehicle (control)
and 6 wells were treated with cis-platin (25 mM), which is used as
a standard cytotoxic drug. Cell viability was tested after 48 h with
A
summary of data collection and structure refinement
for PPN[RuCl O (1a·2H O) and PPN[RuCl
(DMSO)(L1)]·2H
O) are reported in Table 3. Single
4
2
2
4
-
(
DMSO)(L2)]·3H
2
O (2a·3H
2
crystal data were collected with a Bruker AXS Smart 1000
˚
area detector diffractometer (Mo Ka; a = 0.71073 A). Cell
52
the Resazurin assay as previously described: growth medium
parameters were refined from the observed setting angles and
detector positions of selected strong reflections. Intensities were
integrated from several series of exposure frames covering a
was replaced with a DMEM solution of resazurin. After 2 h, the
samples’ fluorescence was measured at 572 nm with a microplate
2
reader (Wallac 1420 Victor Multilab Counter).
47
hemisphere of reciprocal space. No crystal decay was observed.
48
Absorption correction using the program SADABS was applied,
which resulted in minimum and maximum transmission factors of
Acknowledgements
0
.974–1.000 (1a·2H
2
O) and 0.755–1.000 (2a·3H
2
O). The structures
49
This work was supported by Universit a` degli Studi di Parma, Italy.
We thank CIM (Centro Interdipartimentale Misure “Giuseppe
Casnati”) of the University of Parma for the use of the NMR
facilities. We are also grateful to CIRCMSB (Consorzio Interuni-
versitario di Ricerca in Chimica dei Metalli nei Sistemi Biologici).
were solved by direct methods (SIR97) and refined with full-
matrix least squares (SHELXL-97) using the Wingx software
50
51
package. Non-hydrogen atoms were refined anisotropically. The
hydrogen atoms of the two water molecules of 1a·2H O could
not be located from the difference Fourier map, whereas the
O were
2
hydrogen atoms of the three water molecules of 2a·3H
2
found and refined. The remaining hydrogen atoms were placed at
their calculated positions. The maximum and minimum peaks on
the final difference Fourier maps corresponded to 0.480/-0.375
Notes and references
1
2
3
M. J. Clarke, F. C. Zhu and D. R. Frasca, Chem. Rev., 1999, 99, 2511.
M. J. Clarke, Coord. Chem. Rev., 2002, 232, 69.
E. Alessio, G. Balducci, A. Lutman, G. Mestroni, M. Calligaris and
W. M. Attia, Inorg. Chim. Acta, 1993, 203, 205.
-
3
˚
O) e A . Graphical material
(
1a·2H
2
O) and 0.730/-0.464 (2a·3H
2
5
1
was prepared with the ORTEP3 for Windows program.
3
772 | Dalton Trans., 2009, 3766–3773
This journal is © The Royal Society of Chemistry 2009

View Online
4
5
6
7
8
9
G. Sava, S. Pacor, G. Mestroni and E. Alessio, Clin. Exp. Metastasis,
992, 10, 273.
G. Sava, I. Capozzi, K. Clerici, G. Gagliardi, E. Alessio and G.
Mestroni, Clin. Exp. Metastasis, 1998, 16, 371.
27 R. W. Clark, P. J. Squattrito, A. K. Sen and S. N. Dubey, Inorg. Chim.
Acta, 1999, 293, 61.
28 R. M. McCarrick, M. J. Eltzroth and P. J. Squattrito, Inorg. Chim. Acta,
2000, 311, 95.
29 T. W. Kajdan, P. J. Squattrito and S. N. Dubey, Inorg. Chim. Acta, 2000,
300, 1082.
30 P. Aslanidis, P. J. Cox, S. Divanidis and P. Karagiannidis, Inorg. Chim.
Acta, 2004, 357, 1063.
31 M. B. Cingi, M. Lanfranchi, M. A. Pellinghelli and M. Tegoni,
Eur. J. Inorg. Chem., 2000, 4, 703.
32 H. Gunther, NMR Spectroscopy, John Wiley & Sons Ltd, 2001.
33 K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordi-
nation Compounds, Part B, 5th ed., John Wiley & Sons, New York,
1997.
34 E. Alessio, G. Balducci, M. Calligaris, G. Costa, W. M. Attia and G.
Mestroni, Inorg. Chem., 1991, 30, 609.
35 A. Garza-Ortiz, P. U. Maheswari, M. Siegler, A. L. Spek and J. Reedijk,
Inorg. Chem., 2008, 47, 6964.
1
J. M. Rademaker-Lakhai, D. van den Bongard, D. Pluim, J. H. Beijnen
and J. H. M. Schellens, Clin. Cancer Res., 2004, 10, 3717.
B. K. Keppler, W. Rupp, U. M. Juhl, H. Endres, R. Niebl and W. Balzer,
Inorg.Chem., 1987, 26, 4366.
K. G. Lipponer, E. Vogel and B. K. Keppler, Met.-Based Drugs, 1996,
3
, 243.
C. G. Hartinger, S. Zorbas-Seifried, M. A. Jakupec, B. Kynast, H.
Zorbas and B. K. Keppler, J. Inorg. Biochem., 2006, 100, 891.
1
0 C. G. Hartinger, M. A. Jakupec, S. Zorbas-Seifried, M. Groessl,
A. Egger, W. Berger, H. Zorbas, P. J. Dyson and B. K. Keppler,
Chem.Biodivers., 2008, 5, 2140.
1 J. Malina, O. Novakova, B. K. Keppler, E. Alessio and V. Brabec, J. Biol.
Inorg. Chem., 2001, 6, 435.
2 D. Pluim, R. C. A. M. van Waardenburg, J. H. Beijnen and J. H. M.
Schellens, Cancer Chemother. Pharmacol., 2004, 54, 71.
3 E. Reisner, V. B. Arion, B. K. Keppler and A. J. L. Pombeiro, Inorg.
Chim. Acta, 2008, 361, 1569.
4 E. Reisner, V. B. Arion, M. Fatima, C. G. da Silva, R. Lichtenecker,
A. Eichinger, B. K. Keppler, V. Y. Kukushkin and A. J. L. Pombeiro,
Inorg. Chem., 2004, 43, 7083.
5 G. Sava, A. Bergamo, S. Zorzet, B. Gava, C. Casarsa, M. Cocchietto,
A. Furlani, V. Scarcia, B. Serli, E. Iengo, E. Alessio and G. Mestroni,
Eur. J. Cancer, 2002, 38, 427.
6 A. Kung, T. Pieper, R. Wissiack, E. Rosenberg and B. K. Keppler,
J. Biol. Inorg. Chem., 2001, 6, 292.
1
1
1
1
36 B. Serli, E. Zangrando, T. Gianfeffara, L. Yellowlees and E. Alessio,
Coord. Chem. Rev., 2003, 245, 73.
37 M. Zuber and Z. Ciunik, Polyhedron, 2007, 26, 1643.
38 C. M. Duff and G. A. Heath, J. Chem. Soc. Dalton Trans., 1991, 9,
2401.
39 H. Gampp, M. Maeder, C. J. Meyer and A. D. Zuberbuhler, Talanta,
1
1986, 33, 943.
40 R. A. Binstead, A. D. Zuberb u¨ hler, and B. Jung, SPECFIT/32
ver. 3.0.35, 2004, Spectrum Software Associates, http://www.bio-
logic.info/rapid-kinetics/software.html.
1
1
41 M. Bouma, B. Nuijen, M. T. Jansen, G. Sava, A. Flaibani, A. Bult and
J. H. Beijnen, Int. J. Pharm., 2002, 248, 239.
42 S. A. Cotton, Chemistry of Precious Metals, Blackie Academic &
Professional, London, UK, 1997.
7 M. Groessl, E. Reisner, C. G. Hartinger, R. Eichinger, O. Semenova,
A. R. Timerbaev, M. A. Jakupec, V. B. Arion and B. K. Keppler, J. Med.
Chem., 2007, 50, 2185.
1
1
2
2
2
8 M. Brindell, I. Stawoska, J. Supel, A. Skoczowski, G. Stochel and R.
43 M. Ravera, S. Baracco, C. Cassino, P. Zanello and D. Osella, J. Chem.
van Eldik, J. Biol. Inorg. Chem., 2008, 13, 909.
Soc. Dalton Trans., 2004, 15, 2347.
9 L. Messori, P. Orioli, D. Vullo, E. Alessio and E. Iengo, Eur. J. Biochem.,
44 F. Q. Schafer and G. R. Buettner, Free Rad. Biol. Med., 2001, 30,
1191.
2
000, 267, 1206.
0 P. Mura, M. Camalli, L. Messori, F. Piccioli, P. Zanello and M. Corsini,
Inorg. Chem., 2004, 43, 3863.
1 I. Turel, M. Pecanac, A. Golobic, E. Alessio, B. Serli, A. Bergamo and
G. Sava, J. Inorg. Biochem., 2004, 98, 393.
45 F. Frausin, V. Scarcia, M. Cocchietto, A. Furlani, B. Serli, E. Alessio
and G. Sava, J. Pharmacol. Exp. Ther., 2005, 313, 227.
46 M. B. Cingi, F. Bigoli, M. Lanfranchi, M. A. Pellinghelli, A. Vera and
E. Buluggiu, J. Chem. Soc. Dalton Trans., 1992, 21, 3145.
47 SMART (control) and SAINT (integration) software for CCD systems,
Bruker AXS, Madison, WI, USA, 1994.
2 A. H. Velders, A. Bergamo, E. Alessio, E. Zangrando, J. G. Haasnoot,
C. Casarsa, M. Cocchietto, S. Zorzet and G. Sava, J. Med. Chem., 2004,
4
7, 1110.
48 Area-Detector Absorption Correction, Siemens Industrial Automation,
2
2
2
3 P. Mura, F. Piccioli, C. Gabbiani, M. Camalli and L. Messori, Inorg.
Chem., 2005, 44, 4897.
Inc., Madison, WI, 1996.
49 A. Altomare, M. C. Burla, M. Camalli, G. L. Cascarano, C. Giacov-
azzo, A. Guagliardi, A. G. G. Moliterni, G. Polidori and R. Spagna,
J. Appl. Cryst., 1999, 32, 115.
50 G. M. Sheldrick, SHELX97. Programs for Crystal Structure Analysis
(Release 97–2). University of G o¨ ttingen, Germany, 1997.
51 L. J. Farrugia, J. Appl. Cryst, 1999, 32, 837.
52 S. Tardito, O. Bussolati, F. Gaccioli, R. Gatti, S. Guizzardi, J. Uggeri,
L. Marchi o` , M. Lanfranchi and R. Franchi-Gazzola, Histochem. Cell
Biol., 2006, 126, 473.
4 V. B. Arion, E. Reisner, M. Fremuth, M. A. Jakupec, B. K. Keppler,
V. Y. Kukushkin and A. J. L. Pombeiro, Inorg. Chem., 2003, 42, 6024.
5 S. Tardito, O. Bussolati, M. Maffini, M. Tegoni, M. Giannetto, V.
Dall’Asta, R. Franchi-Gazzola, M. Lanfranchi, M. A. Pellinghelli, C.
Mucchino, G. Mori and L. Marchi o` , J. Med. Chem., 2007, 50, 1916.
6 F. Dallavalle, F. Gaccioli, R. Franchi-Gazzola, M. Lanfranchi, L.
Marchi o` , M. A. Pellinghelli and M. Tegoni, J. Inorg. Biochem., 2002,
2
9
2, 95.
This journal is © The Royal Society of Chemistry 2009
Dalton Trans., 2009, 3766–3773 | 3773