Dinitrosyl iron complexes [E5Fe(NO)2]- (E = S, Se): A precursor of Roussin's black salt [Fe4E3(NO)7]-

Article abstract of DOI:10.1016/j.ica.2006.02.035

[PPN][Se₅Fe(NO)₂] (1) and [K-18-crown-6-ether][S₅Fe(NO)₂] (2′) were synthesized and characterized by IR, UV-Vis, EPR spectroscopy, magnetic susceptibility, and X-ray structure. [PPN][Se₅Fe(NO)₂

Full text of DOI:10.1016/j.ica.2006.02.035

Inorganica Chimica Acta 359 (2006) 2525–2533
www.elsevier.com/locate/ica
Dinitrosyl iron complexes [E5Fe(NO)2]ꢀ (E = S, Se): A precursor of
Roussin’s black salt [Fe4E3(NO)7]ꢀ
a
a
a
b,
a
Tai-Nan Chen , Feng-Chun Lo , Ming-Li Tsai , Ko-Nien Shih *, Ming-Hsi Chiang ,
c
a,
*
Gene-Hsiang Lee , Wen-Feng Liaw
a
Department of Chemistry, National Tsing Hua University, Hsinchu 30043, Taiwan
Department of Radiological Technology, Yuanpei Institute of Science and Technology, Hsinchu, Taiwan
b
c
Instrumentation Center, National Taiwan University, Taipei 10764, Taiwan
Received 22 December 2005; received in revised form 17 February 2006; accepted 23 February 2006
Available online 3 March 2006
Abstract
[PPN][Se5Fe(NO)2] (1) and [K-18-crown-6-ether][S5Fe(NO)2] (20) were synthesized and characterized by IR, UV–Vis, EPR spectros-
copy, magnetic susceptibility, and X-ray structure. [PPN][Se5Fe(NO)2] easily undergoes ligand exchange with S8 and (RS)2 (R = C7H4SN
(5), o-C6H4NHCOCH3 (6), C4H3S (7)) to form [PPN][S5Fe(NO)2] and [PPN][(SR)2Fe(NO)2]. The reaction displays that [E5Fe(NO)2]ꢀ
(E = Se (3), S (4)) facilely converts to [Fe4E3(NO)7]ꢀ by adding acid HBF4 or oxidant [Cp2Fe][BF4] in THF, respectively. Obviously,
complexes 1 and 20 serve as the precursors of the Roussin’s black salts 3 and 4. The electronic structure of {Fe(NO)2}9 core of
[Se5Fe(NO)2]ꢀ is best described as a dynamic resonance hybrid of {Fe+1(ꢀNO)2}9 and {Feꢀ1(NO+)2}9 modulated by the coordinated
ligands. The findings, EPR signal of g = 2.064 for 1 at 298 K, implicate that the low-molecular-weight DNICs and protein-bound DNICs
may not exist with selenocysteine residues of proteins as ligands, since the existence of protein-bound DNICs and low-molecular-weight
DNICs in vitro has been characterized with a characteristic EPR signal at g = 2.03. In addition, complex 20 treated human erythroleu-
kemia K562 cancer cells exposed to UV-A light greatly decreased the percentage survival of the cell cultures.
ꢀ 2006 Elsevier B.V. All rights reserved.
Keywords: Dinitrosyl iron compounds; DNICs; Selenolate
1. Introduction
DNICs) and is probably released from cells in the form
of low-molecular-weight dinitrosyl iron complexes
(LMW-DNICs) [6–10]. Recent report showed that both
cysteine-bound DNICs (DNICs-CYS) and glutathione-
bound DNICs (DNICs-GSH) elicited a NO release associ-
ated relaxant effect in isolated arteries, and a faster rate of
NO release from DNICs-CYS than from DNICs-GSH was
observed [11]. The chemical versatility of sulfur in biology
is well established [12]. A genetic cordon for selenocysteine
incorporation into proteins has been established in pro-
karyotes in the past decade [13]. In fact, Bo¨ck et al. pro-
nounced selenocysteine as the 21st amino acid in
ribosome-mediated protein synthesis [14].
Extensive EPR studies have identified nitrosyl non-heme
iron complexes as products from the interaction of NO
with several iron–sulfur and other iron-containing proteins
[1–9]. Examples of nitric oxide coordination to iron and the
spectroscopic signals of dinitrosyl iron complexes (DNICs)
are of much interest, particularly in light of role(s) in sul-
fur-rich protein uptake and degradation [6,7]. DNICs have
been suggested as one of the two possible forms for storage
and transport of NO in biological system [8,9]. In vivo,
nitric oxide can be stabilized and stored in the form of dini-
trosyl iron complexes with proteins (protein-bound
In particular, nitric oxide derived from spontaneously
decomposing NO-donor compounds, e.g., Roussin’s salt,
has been known to sensitize hypoxic cell cultures to
*
Corresponding authors. Tel./fax: +88635721890.
E-mail address: wfliaw@mx.nthu.edu.tw (W.-F. Liaw).
0020-1693/$ - see front matter ꢀ 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.ica.2006.02.035

2526
T.-N. Chen et al. / Inorganica Chimica Acta 359 (2006) 2525–2533
c-radiation damage [15]. In model compounds, Glidewell
and co-workers reported that reactions of iron–sulfur pro-
teins model compounds [Fe2(l-S)2 (S,S-C6H4)2]2ꢀ/[Fe4S4-
(SR)4]2ꢀ and nitric oxide/nitrite, followed by acidic
work-up, yielded the anionic [Fe4S3(NO)7]ꢀ via the forma-
tion of the paramagnetic intermediate [(SH)2Fe(NO)2]ꢀ
[16,17]. Also, Roussin’s black salt [Fe4S3(NO)7]ꢀ was
reported to be isolated upon reaction of [Fe(CO)3(NO)]ꢀ
with elemental sulfur or polysulfide [18], and protonation
of [Fe2S2(NO)4]2ꢀ by HBF4/CF3COOH [19]. Interestingly,
the facile conversion of [Fe2S2(NO)4]ꢀ into [Fe4S3(NO)7]ꢀ
was also observed simply by dissolving [Fe2S2(NO)4]ꢀ salt
in CH2Cl2 [16,20].
powder, fluoroboric acid, ferrocenium tetrafluoroborate,
(Lancaster/Aldrich/Fluka/Acros), and 3-(4,5-dimethylthia-
zol-2-yl)-5-diphenyltetrazolium bromide (MTT) (Sigma)
were used as received. The compounds [PPN][Fe(CO)3-
(NO)], [FeII(S2CN(Et)2)2] (Fe(DTC)2), and [FeIII(S2-
CN(Et)2)3] (Fe(DTC)3), were synthesized and characterized
by published procedures [24,28]. Infrared spectra of the
mNO stretching frequencies were recorded on a Perkin–
Elmer model spectrum one spectrophotometer with sealed
solution cells (0.1 mm) and KBr windows. UV–Vis spectra
were recorded on a GBC Cintra 10e spectrophotometer.
Analyses of carbon, hydrogen, and nitrogen were obtained
with a CHN analyzer (Heraeus).
We have demonstrated that reversible transformation of
the {Fe(NO)2}9 [PPN][S5Fe(NO)2] (PPN = bis(triphenyl-
phosphoranylidene)ammonium) to the [S5Fe(l-S)2FeS5]2ꢀ
cluster by photolysis in the presence of the NO-acceptor
reagent [(C4H8O)Fe(S,S-C6H4)2]ꢀ, identified by UV–Vis,
EPR and IR [21,22], is consistent with reports of repair
of nitric oxide modified [2Fe–2S] ferredoxin by cysteine
desulfurase and L-cysteine in vitro [23]. A logical and signif-
icant extension of these efforts would be analogous studies
involving exploration of dinitrosyl iron derivatives in the
biologically relevant selenolate ligand field. In this study,
the analogue [PPN][Se5Fe(NO)2] (1) was isolated from
reaction of [PPN][Fe(CO)3(NO)] and Se powder in THF.
The facile conversion of complexes 1/[cation][S5Fe(NO)2]
(cation = N(PPh3)2 (2), K-18-crown-6-ether (20)) into
Roussin’s black salts [Fe4E3(NO)7]ꢀ (E = Se (3), S (4))
accompanied by the release of NO was observed upon reac-
tion of complexes 1/2 and HBF4/[Cp2Fe][BF4], individu-
ally. Transformation of complex 1 to dinitrosyl iron
complexes [PPN][(RS)2Fe(NO)2] (R = C7H4SN (5), o-C6-
H4NHCOCH3 (6), C4H3S (7)) was also investigated. In this
paper, we also demonstrate that NO is able to transfer
from [S5Fe(NO)2]ꢀ to the NO-trapping reagents ([FeII(S2-
CN(Et)2)2] or [FeIII(S2CN(Et)2)3]) to form [(NO)Fe(S2C-
N(Et)2)2] complex. Also described are effects on the
mortalities of human erythroleukemia K562 cells treated
with DNICs 20 alone and DNICs 20 combined with UV-
A light (wavelength = 400–315 nm).
2.1. Preparation of [PPN][Se5Fe(NO)2] (1)
The compounds [PPN][Fe(CO)3(NO)] (0.708 g, 1 mmol)
and Se powder (0.48 g, 6 mmol) were dissolved in 10 mL of
THF and refluxed under nitrogen at 70 ꢁC. The reaction
was monitored with FTIR. The spectrum (IR (THF):
1736 s, 1697 s (mNO) cmꢀ1) was assigned to the formation
of complex 1. The resulting mixture was filtered to separate
dark green solution (complex 1) and dark brown insoluble
solid. The dark green THF solution was then concentrated
and diethyl ether–hexane was added to precipitate the dark
green solid [PPN][Se5Fe(NO)2] (1) (0.423 g, yield 40%,
based upon total iron). Diffusion of diethyl ether into
THF solution of complex 1 at ꢀ15 ꢁC for 4 weeks led to
dark green crystals suitable for X-ray crystallography. IR
(THF): 1736 s, 1697 s (mNO) cmꢀ1. Absorption spectrum
(THF) [kmax/nm (e/Mꢀ1 cmꢀ1)]: 444 (3168), 588 (1378)
(Calc. for C36H30N3O2P2Se5Fe: C, 53.07; H, 3.71; N,
5.16. Found: C, 53.29; H, 3.90; N, 5.58%).
2.1.1. Preparation of [K-18-crown-6-ether][S5Fe(NO)2]
(20)
[K-18-crown-6-ether][Fe(CO)3NO] (0.484 g, 1 mmol)
and S8 (0.256 g, 1 mmol) were loaded into a 50-mL Schlenk
flask and dissolved in THF (10 mL). After the solution
mixture was stirred under nitrogen at ambient temperature
overnight, the solution IR spectrum (IR (THF): 1696 s,
1738 s (mNO) cmꢀ1) was assigned to the formation of [K-
18-crown-6-ether][S5Fe(NO)2]. The resulting mixture was
then filtered through Celite and diethyl ether was then
added to precipitate dark-brown solid [K-18-crown-6-
ether][S5Fe(NO)2] (20) (yield 0.20 g, 35%) [21]. Diffusion
of diethyl ether into THF solution of complex 20 at
ꢀ15 ꢁC for 4 weeks led to dark-brown crystals suitable
for single-crystal X-ray diffraction. IR (THF): 1738 s,
1696 s (mNO) cmꢀ1. Absorption spectrum (THF) [kmax/nm
(e/Mꢀ1 cmꢀ1)]: 364 (2887), 440 (2824), 568 (916).
2. Experimental
Manipulations, reactions, and transfers of samples were
conducted under nitrogen according to standard Schlenk
techniques or in a glovebox (argon gas). Solvents were dis-
tilled under nitrogen from appropriate drying agents
(diethyl ether from CaH2, acetonitrile from P2O5–CaH2,
methylene chloride from P2O5–CaH2, hexane and tetrahy-
drofuran (THF) from sodium-benzophenone) and stored
˚
in dried, N2-filled flasks over 4 A molecular sieves. Nitro-
gen was purged through these solvents before use. Solvent
was transferred to a reaction vessel via a stainless steel can-
nula under positive pressure of N2. The reagents bis(triphe-
nylphosphoranylidene)ammonium chloride ([PPN]Cl), iron
pentacarbonyl, sodium nitrite, elemental sulfur, selenium
2.2. Reactions of [PPN][E5Fe(NO)2] (E = Se (1); S (2))
and [Cp2Fe][BF4]/HBF4
A THF solution (2 mL) of [Cp2Fe][BF4] (0.0273 g,
0.1 mmol) (or fluoroboric acid (17.8 lL, 0.1 mmol)) was

T.-N. Chen et al. / Inorganica Chimica Acta 359 (2006) 2525–2533
2527
added dropwise by a cannula into the THF solution (3 mL)
of complex (0.1045 g, 0.1 mmol) (or complex
2.5. Reaction of [PPN][S5Fe(NO)2], Fe(DTC)3
(DTC = diethyldithiocarbamate) and S8
1
2
(0.1 mmol)) and stirred at room temperature overnight.
The reaction mixture was filtered to remove the insoluble
solid and then hexane was added to precipitate the known
dark green solid [PPN][Fe4E3(NO)7] (E = S (3), Se (4);
yield 65% for 3 and 21% for 4) [25–27]. Complex 3: IR
(THF): 1795 w, 1743 s, 1707 w (mNO) cmꢀ1. Complex 4:
IR (THF):1790 w, 1740 s, 1707 w (mNO) cmꢀ1. Recrystalli-
zation from saturated THF solution of complexes 3/4 with
hexane diffusion at ꢀ15 ꢁC gave dark green crystals suit-
able for X-ray crystallography.
A solution containing [PPN][S5Fe(NO)2] (0.1 mmol,
0.0815 g) and S8 (0.1 mmol, 0.0256 g) and Fe(DTC)3-
(0.1 mmol, 0.05 g) (or Fe(DTC)2 (0.2 mmol, 0.0704 g))
[28] in THF (10 mL) was stirred overnight under N2 (g)
at ambient temperature. The reaction solution was moni-
tored by IR. The spectrum (IR (THF): 1715 s (mNO) cmꢀ1
)
was assigned to the formation of the known [(NO)Fe-
(DTC)2] [28]. The reaction mixture was then filtered to sep-
arate the green solution [(NO)Fe(DTC)2] and the known
reddish-brown precipitate [PPN][S5Fe(l-S)2FeS5] (yield
62%, based on total iron) characterized by UV–Vis and sin-
gle-crystal X-ray diffraction [29,30].
2.3. Reaction of [PPN][Se5Fe(NO)2] and disulfides (RS)2
(R = 2-C7H4NS, C4H3S, C6H4-o-NHC(O)CH3)
A
THF solution of [PPN][Se5Fe(NO)2] (0.209 g,
2.6. Cell culture experiment
0.2 mmol) was transferred into a 50-mL Schlenk flask
loaded with disulfide species (RS)2 (R = 2-C7H4NS
(0.067 g, 0.2 mmol), C4H3S (0.046 g, 0.2 mmol), C6H4-o-
NHC(O)CH3 (0.067 g, 0.2 mmol)). The reaction mixture
was stirred for 10 min at room temperature, then filtered
to remove the insoluble solid, and hexane (15 mL) was
added to precipitate the known [PPN][(RS)2Fe(NO)2]
(R = C7H4NS (5) (79%), C6H4-o-NHC(O)CH3 (6) (80%),
C4H3S (7) (88%)), characterized by IR, UV–Vis, and sin-
gle-crystal X-ray diffraction [22]. Complex 5: IR (THF):
1766 s, 1716 s (mNO) cmꢀ1. Absorption spectrum (THF)
[kmax/nm (e/Mꢀ1 cmꢀ1)]: 465 (3500), 799 (712). Complex
Human erythroleukemia K562 cells were grown in sus-
pension in Dulbecco’s modified Eagle’s medium (DMEM)
supplemented with 10% fetal bovine serum (Hyclone), pen-
icillin (100 IU/ml), streptomycin (100 lg/ml), Fungizone
(50 lg/ml) and 2 mM L-glutamine. The cells were main-
tained at 37 ꢁC under 5% CO2 in a humidified incubator.
The cells were plated in 12-well plates at cell density of
2 · 105, 3 · 105, and 4 · 105 cells/well, respectively. Com-
plex [18-crown-6-ether-K][S5Fe(NO)2] (20) serves as a
NO-donor reagent. The cells were treated with different
concentrations of 20 (1, 10, and 100 lM) for 24 h and the
percentage of survival was measured by the colorimetric
MTT assay. The MTT solution (5 mg/ml in PBS) was
added to each well and then incubated for 3 h. The surviv-
ing cells convert MTT to MTT formazan that shows purple
color when formazan was dissolved in dimethyl sulfoxide
(DMSO), and the absorption intensity was measured at
570 nm by using a microplate reader for enzyme-linked
immunosorbent assays (ELx808). Complex 20 proved to
be not cytotoxic or cytostatic to be used in such experi-
ments. The cells were treated with compound 20 (10 lM)
for 2 h in DMEM and then exposed the cell to UVA light
(k = 430 nm) for 10 min. The cells were then cultured for
24 h after DNIC and UVA treatment. The collected cells
were centrifuged in 1000 rpm for 1 min to separate the
medium. Fresh DMEM was added and subsequently the
viability was determined by the colorimetric MTT assay.
6: IR (THF): 1752 s, 1705 s (mNO), 1690 s (mCO) cmꢀ1
.
Absorption spectrum (THF) [kmax/nm (e/Mꢀ1 cmꢀ1)]: 479
(2984), 774 (453). Complex 7: IR (THF): 1743 s, 1698 s
(mNO) cmꢀ1. Absorption spectrum (THF) [kmax/nm (e/
M
ꢀ1 cmꢀ1)]: 514 (2830), 798 (902) [22].
2.4. Photolysis of THF solution of complex 1 and
[PPN]2[Fe(S,S-C6H4)2]2
An 8 mL THF solution of compounds 1 (0.0525 g,
0.05 mmol)
and
[PPN]2[Fe(S,S-C6H4)2]2
(0.089 g,
0.05 mmol) was loaded into a reactor (20 mL). The
reaction mixture was then irradiated by UV lamp
(Imax = 366 nm) under N2 atmosphere at room tempera-
ture for 6 h after the reaction solution was stirred in the
dark for 2 h. The resulting mixture was filtered to separate
the red brown precipitate and the dark reddish brown
upper solution. The uncharacterized dark brown precipi-
2.7. EPR measurements
tate is insoluble in organic solvent. IR (mNO 1789 cmꢀ1
)
and UV–Vis spectra (497, 610, 1305 nm (THF)) of the fil-
trate (upper solution) indicated the formation of
[PPN][(NO)Fe(S,S-C6H4)2]. The THF solution of complex
[PPN][(NO)Fe(S,S-C6H4)2] was concentrated to 5 mL
under vacuum and hexane–diethyl ether was then added
to precipitate the dark reddish brown solid [PPN][(NO)-
Fe(S,S-C6H4)2] (0.080 g, 40%) identified by UV–Vis (497,
610, 1305 nm (THF)), and FTIR (mNO: 1789 s cmꢀ1
(THF)).
EPR measurements were performed at X-band using a
Bruker EMX spectrometer equipped with a Bruker
TE102 cavity and a Bruker VT2000 temperature control
unit (120–300 K). For liquid helium temperature measure-
ments, an Oxford ESR910 continuous flow cryostat (4–
200 K) was used. X-band EPR spectra of complex 1 frozen
in THF were obtained with a microwave power of 5 mW,
frequency at 9.4323 GHz, and modulation amplitude of
0.05 mT at 100 kHz (temperature = 77 K). X-band EPR

2528
T.-N. Chen et al. / Inorganica Chimica Acta 359 (2006) 2525–2533
spectra of complex 1 were measured at a frequency of
9.78410 GHz at room temperature.
-
-
[Fe(CO)3(NO)]
(a)
[Fe4E3(NO)7]
E= Se (3), S (4)
(c)
Se powder
2.8. Magnetic measurements
[Cp2Fe]+
(d) H+/[CP2Fe][BF4]
-
Se
-
The magnetization data were recorded on a SQUID
magnetometer (MPMS5 Quantum Design company) with
an external 0.5 T magnetic field for complex 1 in the tem-
perature range from 4 to 300 K. The experimental mag-
netic susceptibility data were corrected for diamagnetism
by the tabulated Pascal’s constants.
S
(b)
O
O
N
N
Se
N
N
O
O
Se
Se
S
Fe
Se
S
Fe
(2)
S
S8
(e)
S
(1)
(RS)2
-
[Fe(S,S-C6H4)2]2-
RS
RS
NO
Fe
(f)
NO
Se powder
2.9. Crystallography
S
N
S
(5)
O
N
-
S
Crystallographic data of complexes 1 and 20 are summa-
rized in the Supporting Information. The crystals of 1 and
20 chosen for X-ray diffraction studies measured
0.25 · 0.20 · 0.20 mm and 0.30 · 0.25 · 0.25 mm, respec-
tively. Each crystal was mounted on a glass fiber. Diffraction
measurements for complexes 1 and 20 were carried out at
150(1) K on a Bruker-Nonius Kappa CCD diffractometer
with graphite-monochromated Mo Ka radiation (k =
S
S
RS =
,
,
S
S
S
S
HN
(7)
Fe
O
(6)
Scheme 1.
˚
0.7107 A) and h between 1.45ꢁ and 27.50ꢁ for complex 1,
and between 1.43ꢁ and 27.50ꢁ for complex 20. Least-squares
refinement of the positional and anisotropic thermal param-
eters for the contribution of all non-hydrogen atoms and
fixed hydrogen atoms was based on F2. Absorption correc-
tion was made using SORTAV [31]. The SHELXTL structure
refinement program was employed [32]. In the case of
complexes 1 and 20, the [Se5Fe(NO)2] and [S5Fe(NO)2]
fragments were found at disordered positions ð½S5Fe-
ðNOÞ ꢁ:½Se0 Fe0ðN0O0Þ ꢁ ¼ 0:94:0:06 and ½S5FeðNOÞ ꢁ:½S0 -
Fe0ðN0O0Þ2ꢁ5¼ 0:5:0:5Þ, respectively, and were refined by
2
2
2
5
partial occupancies.
Fig. 1. ORTEP drawing and labeling scheme of the [(NO)2FeSe5]ꢀ anion
with thermal ellipsoids drawn at 50% probability.
3. Results and discussion
3.1. Synthesis and spectroscopic characterization
As presented in Scheme 1a, a straightforward synthetic
reaction of [PPN][Fe(CO)3(NO)] with six equivalents of
Se powder in THF at 70 ꢁC was conducted. The reaction
mixture led to the isolation of the THF-soluble dark brown
[PPN][Se5Fe(NO)2] (1) (Fig. 1). Complex 1 exhibits two IR
S = 1/2 rhombic EPR spectrum with principal g values at
g1 = 2.016, g2 = 2.034, and g3 = 2.146 (77 K) (Fig. 2b)
[6,9,21–23]. Replacing the sulfur ring with selenium ring
yielded the obvious changes of the EPR spectra indicating
that the unpaired electron is more localized in the
{Fe(NO)2}9 unit [21,22].
m
NO bands at 1736 s, 1697 s cmꢀ1 (THF). The mNO bands of
complex 1 are similar to the NO stretching frequencies of
[S5Fe(NO)2]ꢀ (mNO = 1739, 1695 cmꢀ1), [(PhS)2Fe(NO)2]ꢀ
(mNO = 1735, 1694 cmꢀ1) [21,22], and [(PhSe)2Fe(NO)2]ꢀ
(1741 s, 1697 s cmꢀ1 (CH2Cl2)) [33]. Thus, the electron-
donating ability of the [Se5]2ꢀ ligand to {Fe(NO)2} unit
3.2. Reactivity
The coordinated [Se5]2ꢀ ligand of complex 1 could be
replaced by the [S5]2ꢀ. Dark green solution of complex 1
with one equivalent of S8 in THF solution rapidly resulted
in color change to dark brown with little NO stretching
frequency shift from 1697, 1736 to 1695, 1739 cmꢀ1 for
[PPN]+ salt and 1696, 1738 cmꢀ1 for [K-18-crown-6-
ether]+ salt, which implicated the formation of [cat-
ion][S5Fe(NO)2] (cation = PPN (2), K-18-crown-6-ether
(20)) (Scheme 1b) [21,22]. This substitution reaction is
could be considered as the same as those of the [S5]2ꢀ
,
2[SePh]ꢀ and 2[SPh]ꢀ ligands, respectively [21,22,33,34],
where the iron–dinitrosyl unit has the formal electronic
assignment, {Fe(NO)2}9 [35]. In contrast to the EPR iso-
tropic signal at g = 2.03, the characteristic of dinitrosyl
iron complexes [(RS)2Fe(NO)2]ꢀ, complex 1 exhibits an
isotropic signal at g = 2.064 at 298 K (Fig. 2a), and a

T.-N. Chen et al. / Inorganica Chimica Acta 359 (2006) 2525–2533
2529
60000
40000
20000
0
40000
30000
20000
10000
0
-10000
*
*
-20000
-30000
-40000
-20000
-40000
2.20
2.15
2.10
2.05
2.00
1.95
2.20
2.15
2.10
2.05
2.00
1.95
Field (g)
Field (g)
a
b
Fig. 2. X-band EPR spectra of complex 1 (a) at 298 K (* indicates impurity), and (b) at 77 K.
Fig. 3. ORTEP drawing and labeling scheme of [K-18-crown-6-ether][(NO)2FeS5] with thermal ellipsoids drawn at 30% probability.
expected to be driven by the formation of six-membered-
ring FeS5 bonds that are sufficient to place the {Fe(NO)2}9
complex 2 in the more optimum electronic condition [22].
In contrast, complex 2 was stable in the presence of six
equivalents of Se in THF at 70 ꢁC. Also, a difference in
the photochemical behavior was observed between the
highly photosensitive (photoreactive) complex 2 and pho-
w cmꢀ1 (THF) (4)). The products are further confirmed
by X-ray diffraction analysis. Obviously, the mononuclear
dinitrosyl iron complexes 1 and 2 are the precursors of
complexes 3 and 4 in this reaction, providing a route for
reassembly in reactions involving a change of nuclearity
and release of NO.
In contrast to the inertness of reaction of complex 2 and
disulfides (RS)2 [22], complex 1 triggers the S–S bond acti-
vation of (RS)2 species to form the stable [(RS)2Fe(NO)2]ꢀ
(R = C7H4SN (5), o-C6H4NHCOCH3 (6), C4H3S (7)) as
indicated in the NO stretching frequency shift from 1736,
tostable complex 1 in the absence of NO-trapping agent
2ꢀ
½FeðS; S-C6H4Þ2ꢁ
.
2
As shown in Scheme 1c–d, complexes 1 and 2 reacted
with one equivalent of acid HBF4 (or oxidant [Cp2Fe]-
[BF4]) in THF at ambient temperature overnight yielded
the known dark green complexes [Fe4E3(NO)7]ꢀ (E = Se
(3), S (4)), respectively [9,15–17], as identified by IR mNO
(1790 w, 1740 s, 1707 w (THF) (3); 1795 w, 1743 s, 1707
1697 to 1766, 1716, 1752, 1705, and 1743, 1698 cmꢀ1
,
respectively (Scheme 1e) [22]. This result may implicate that
the {Fe(NO)2}9 core shows more affinity to the thiolate
ligands as compared to selenolate ligand.

2530
T.-N. Chen et al. / Inorganica Chimica Acta 359 (2006) 2525–2533
In our previous report [21,22], we have demonstrated
[FeIII(S,S-CN(Et)2)3] for NO radical [28]. In a similar fash-
ion, reaction of complex 1 and one equivalent of [Fe(S,S-
CN(Et)2)3] led to the isolation of [(NO)Fe(S,S-CN(Et)2)2]
and Se powder (no repaired product [Se5Fe(l-Se)2FeSe5]2ꢀ
was isolated) in THF overnight. Compared to complex 2,
the prolonged reaction time for the complete formation
of [(NO)Fe(S,S-CN(Et)2)2] from reaction of complex 1
and one equivalent of [Fe(S,S-CN(Et)2)3] also implicates
the less NO-release ability of complex 1.
the reversible transformation between [S5Fe(NO)2]ꢀ and
[S5Fe(l-S)2FeS5]2ꢀ. Reactions of [S5Fe(NO)2]ꢀ and NO
2ꢀ
trapping agent ½FeðS; S-C6H4Þ2ꢁ2
in DMSO yielded
[(NO)Fe(S,S-C6H4)2]ꢀ and [S5Fe(l-S)2FeS5]2ꢀ with kobs
=
1.5 (3) · 10ꢀ2 sꢀ1 (DMSO, 18 ꢁC). For comparisons of
NO-release activity of dinitrosyl iron–thiolate and iron–sel-
enolate compounds, the representative time courses (6 h) of
NO release trapped by one equivalent of ½FeðS; S-
2ꢀ
C6H4Þ2ꢁ2 in THF were monitored by the formation of
[(NO)Fe(S,S-C6H4)2]ꢀ with an intense absorption band
at 1298 nm. The stability of complexes 1/[(PhSe)2Fe-
(NO)2]ꢀ displaying a nearly identical UV–Vis absorption
spectrum suggests that the NO release of complexes 1/
3.3. Magnetic susceptibility study
Magnetic susceptibility data of a powdered sample of
complex 1 were collected in the temperature range of 4.00–
300 K in a 5 kG (0.5 T) applied field. The net molar magnetic
susceptibility (vM), as shown in Fig. 4, increases from
4.516 · 10ꢀ3 cm3 molꢀ1 at 300 K to 0.097 cm3 molꢀ1 at
4 K. Its corresponding vMT value and effective magnetic
moment (leff) decrease from 1.355 cm3 K molꢀ1 (3.293
BM) at 300 K to 0.390 cm3 K molꢀ1 (1.767 BM) at 4 K.
From a magnetochemical point of view, the {Fe+1(NO)2}9
unit is considered as a tri-spin system. The spin-only
(g = 2.00) vMT and leff values for three magnetically
independent centers are 2.625 cm3 K molꢀ1 and 4.583 BM,
respectively. The spin-only values for the S = 5/2 and
[(PhSe)2Fe(NO)2]ꢀ does not occur directly in the presence
2ꢀ
of NO-trapping agent ½FeðS; S-C6H4Þ2ꢁ
at ambient tem-
2
perature (Scheme 1f). This result demonstrates that
{Fe(NO)2}9[Se5Fe(NO)2]ꢀ and [(PhSe)2Fe(NO)2]ꢀ display
the poor NO-donor activity. Obviously, the selenium or
sulfur bound to the {Fe(NO)2}9 motif may serve to modu-
late the NO-release ability of DNICs. Such a regulatory
role of thiolate/selenolate ligands of the {Fe(NO)2}9
DNICs may be critical for {Fe(NO)2}9 DNICs to serve
as a NO transporter in biological systems. However, irradi-
ation of THF solution of complex 1 in the presence of NO-
2ꢀ
trapping agent ½FeðS; S-C6H4Þ2ꢁ2 for 6 h partially yielded
[(NO)Fe(S,S-C6H4)2]ꢀ (yield 40%) and Se powder. Presum-
ably, irradiation triggers NO release of complex 1. Photol-
ysis of THF solution of [(PhSe)2Fe(NO)2]ꢀ for 1 h yielded
insoluble solids (complete decomposition). The facile
release of NO under photolysis for Roussin’s red/black
salts has been studied by Ford et al [15].
0.20
0.15
0.10
0.05
0.00
Interestingly, reaction of complex 1 with one equivalent
2ꢀ
of ½FeðS; S-C6H4Þ2ꢁ2 in the presence of additional S8 pow-
der under photolysis resulted in the formation of
[(NO)Fe(S,S-C6H4)2]ꢀ, Se powder and the dianionic
[S5Fe(l-S)2FeS5]2ꢀ isolated in ꢂ70% yield [29,30]. Forma-
tion of the mixed-chalcogenide [Se5Fe(l-S)2FeSe5]2ꢀ and
[Se5Fe(l-Se)2FeSe5]2ꢀ from reaction of complex 1, S8 pow-
0
50
100
150
200
250
300
2ꢀ
Temperature, K
der and ½FeðS; S-C6H4Þ2ꢁ
under photolysis was not
2
observed. Obviously, ligand-exchange reaction preceded
NO release forming [S5Fe(NO)2]ꢀ, and subsequent NO
release accompanied by oxidative addition to yield
[S5Fe(l-S)2FeS5]2ꢀ via intermediate [S5FeI] [21].
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
3.5
3.0
2.5
In order to further demonstrate that complex 2 primar-
ily transports NO radical, the following reactions were con-
ducted. Reaction of complex 2 and two equivalents of
[Fe(S,S-CN(Et)2)2] in the presence of one equivalent of
S8 in THF at ambient temperature afforded the known
[(NO)Fe(S,S-CN(Et)2)2] and [S5Fe(l-S)2FeS5]2ꢀ immedi-
ately, as characterized by IR, UV–Vis and X-ray diffraction
[28–30]. Whereas only one equivalent of [Fe(S,S-CN-
(Et)2)3] was required to completely convert pne equivalent
of complex 2 into [S5Fe(l-S)2FeS5]2ꢀ along with the forma-
tion of [(NO)Fe(S,S-CN(Et)2)2] in the presence of one
equivalent of S8, consistent with the trapping specificity
of different redox forms, [FeII(S,S-CN(Et)2)2] versus
2.0
1.5
χ
1.0
0.5
0.0
0
50
100
150
200
250
300
Temperature, K
Fig. 4. Plots of molar magnetic susceptibility (top) and effective magnetic
moment as well as vMT (bottom) vs. temperature for complex 1. The solid
line is the best fit of the data to the appropriate theoretical expression.

T.-N. Chen et al. / Inorganica Chimica Acta 359 (2006) 2525–2533
2531
S = 1/2 systems are 4.375 cm3 K molꢀ1 (5.916 BM) and
0.375 cm3 K molꢀ1 (1.732 BM), respectively.
(mNO: 1706 m, 1743 s cmꢀ1) demonstrates the instability
of complex 20 dissolving in H2O–DMSO solution at room
temperature. In order to demonstrate complex 20 serving as
a NO-release reagent, the commercial NO sensor (WPI)
was used to identify the NO released when complex 20
was dissolved in an aerated pH 7.4 aqueous solution
(H2O–DMSO). The decay of the NO electrochemical sig-
nals can be attributed to NO autoxidation in this solution.
On the basis of NO-analyzer experiments, light accelerates
the release of NO when dissolving complex 20 in aqueous
solution (volume ratio of 1:500 for DMSO:PBS (pH 7.4)).
The Roussin’s black/red salts have been used to deliver
NO photochemically to greatly decrease the survival frac-
tions of the cell cultures when exposed to c-radiation
[15,36]. Treatment of human erythroleukemia K562 cells
with UV-A alone had little effect on the survival. This com-
plex 20 proved to be not cytotoxic or cytostatic to be used
in such experiments, with percentage survival of 99% for
treatment with 1 lM complex 20 solution or 1 lM
[S5Fe(l-S)2FeS5]2ꢀ solution. However, treatment with
100 lM complex 20 solution proved to be quite toxic; the
cells were noticeably stained with trypan blue (viable cells
exclude trypan blue, while dead cells stain blue due to
membrane disruption) and the percentage of survival was
less than 10%. Also investigated here are attempts to use
complex 20 as vehicle for delivering NO to cancer cell cul-
tures to induce cell death. Fig. 5 shows the survival rate of
human erythroleukemia K562 cell cultures when subjected
to complex 20 and complex 20 combined with UV-A light in
different cell numbers. Obviously, complex 20 treated can-
cer cells exposed to UV-A light greatly decreased the per-
centage survivals of the cell cultures giving the respective
percentage of survivals 39%, 26%, and 14%, respectively.
This result implicates that complex 20 may serve as a poten-
tially photochemical reagent in pharmacological delivery of
NO to various tumors.
The theoretical equations for data-fitting are derived
^ ^
^
from the Heisenberg Hamiltonian ðH ¼ ꢀ2JS1S2Þ and the
Van-Vleck equation for a single-J-value system in which J
is the exchange parameter between the Fe+ ion and the
ꢀNO radical under approximation of an ideal C2v symmetry.
Representation of the electronic state based on the Ene-
mark–Feltham model as {Fe(NO)2}9 for complex 1 is due
to the variable nature of NO coordinated to transition
metal as Mnꢀ1–NO+, Mn–NOꢀ and Mn+1–NOꢀ. If only
the lowest-energy electronic state is considered, both of
{Fe+(ꢀNO)2}9 and {Feꢀ1(+NO)2}9 in complex 1 are possi-
ble. Both states are paramagnetic. The former has its ST
either of 5/2 or of 1/2. The latter consists of a spin-only
value of 1/2. The experimental magnetic susceptibility
may be contributed from both electronic states if mixed
electronic states are present in complex 1 (equation is shown
below). Due to the undetermined exclusive electronic state
of complex 1, fitting of the magnetic data is based on
assumption of the presence of both electronic states
9
9
veMxp ¼ ð1 ꢀ pÞvMðfFe ð NOÞ g Þ þ pvMðfFe ðþNOÞ g Þ.
þ
ꢀ
ꢀ
2
2
A least-squares fit (R2 = 0.997) to the vM versus tempera-
ture (4–180 K) curve, shown as a solid line in Fig. 4, gave
2J = ꢀ 9 (1) cmꢀ1, g = 2.14, p = 0.812 (6), with TIP (temper-
ature-independent-paramagnetism) held constant at 200 ·
10ꢀ6 cm3 molꢀ1. The result from the fit indicates that domi-
nant fraction of {Feꢀ1(+NO)2}9 exists in the electronic struc-
ture of complex 1. In other words, [(NO)2FeSe5]ꢀ cannot bea
good reagent for the purpose of NO delivery. The retarded
NO-release ability of complex 1 was experimentally revealed
to compliment the magnetic fitting result. In the presence of
½FeðS; S-C6H4Þ2ꢁ22ꢀ, all derivatives of DNICs [(NO)2Fe-
(SR)2]ꢀ showed rapid reactions of NO removal [22]. How-
ever, most of complex 1 remained in the solution under the
same condition after 18 hours. Longer reaction duration
did not improve the yield of [(NO)Fe(S,S–C6H4)2]ꢀ.
The proposed electronic structure difference between
complexes 1 and 2 ({Fe+1(ꢀNO)2}9 versus {Feꢀ1(NO+)2}9
or combinations of {Fe+1(ꢀNO)2}9 and {Feꢀ1(NO+)2}9) is
also reflected in the NO ligand lability of complexes 1 and
2. The inertness of NO ligand of complex 1 was demon-
strated by exposing the THF solution of [Se5Fe(15NO)2]ꢀ
to 14NO atmosphere. Complex [Se5Fe(15NO)2]ꢀ was stable
under purge of 14NO gas in THF at room temperature for
half-an-hour. At the end of this time, the product was all
[Se5Fe(15NO)2]ꢀ. In contrast to [Se5Fe(15NO)2]ꢀ, complex
[S5Fe(15NO)2]ꢀ did show the reversibility of NO ligand labil-
ity to yield [S5Fe(14NO)2]ꢀ when a THF solution of
[S5Fe(15NO)2]ꢀ was exposed to 14NO atmosphere.
100
90
80
70
60
50
40
30
20
10
0
Control
DNIC
DNIC+UVA
3.4. Cell culture experiments
Fig. 5. Percentage survival of human erythroleukemia K562 cells in the
presence of 10 lM DNICs 20 with or without UV-A light for different cell
numbers 2 · 105 (s), 3 · 105 (j), 4 · 105 (d).
Complex 20 is slightly soluble in H2O but is soluble in
H2O–DMSO. The disappearance of IR mNO spectrum

2532
T.-N. Chen et al. / Inorganica Chimica Acta 359 (2006) 2525–2533
˚
3.5. X-ray crystal structure
2.412(1) A is shorter than those of complexes fac-
[FeII(CO)3(SePh)3]ꢀ (Fe–Seave = 2.459(2) A) [37], and
˚
[Se5Fe(l-Se)2FeSe5]2ꢀ (Fe–Seave = 2.424(2) A) [30], but
˚
Figs. 1 and 3 display the thermal ellipsoid plot of the anio-
nic complexes 1 and 20, respectively. Crystallographic data,
selected bond distances and bond angles are given in Tables
1 and 2, respectively. The nitrosyls are slightly bent (\Fe–
N(1)–O(1) = 170.1(6)ꢁ and \Fe–N(2)–O(2) = 167.4(7)ꢁ)
and flared toward each other (\O(1)–Fe–O(2) = 103.4ꢁ ver-
sus \N(1)–Fe–N(2) = 115.6ꢁ; attracto conformation). X-
ray structure determination of complex 1 shows that the
FeSe5 ring in the chair conformation is similar to its ana-
logue, [S5Fe(NO)2]ꢀ [22]. The Fe–Se average distance of
˚
comparable to the Fe–SePh distance (2.395(1) A (average))
in [(PhSe)2Fe(NO)2]ꢀ [33]. Alternation in Se—Se bond
˚
lengths, spanning the range from 2.320(1) to 2.326(1) A, is
observed in complex 1. The Fe–N(2)–O(2) bond is off linear-
ity by 13ꢁ and differs from the Fe–N(1)–O(1) bond, which is
coordinated in a nearly linear mode with Fe–N(1)–O(1)
bond angle of 170.1(6)ꢁ in complex 1.
4. Conclusion and comments
Complexes 1 and 20 were synthesized by the reaction of
[Fe(CO)3(NO)]ꢀ with Se8 and S8 in THF solution. Facile
conversion of complex 1 to 2 and 5–7 was observed by react-
ing complex 1 with one equivalent of S8 and (RS)2 in THF,
respectively. This result shows that the {Fe(NO)2}9 core
has more affinity to the thiolates as compared to selenolate
ligand. Compared to complex 2, magnetic susceptibility
Table 1
Crystallographic data for the complexes [PPN][(Se)5Fe(NO)2] (1) and [K-
18-crown-6-ether][S5Fe(NO)2] (20)
Complexes
1
20
Formula
Formula weight
Crystal size
Crystal system
Space group
Unit cell dimensions
C36H30FeN3O2P2Se5 C36H72Fe3K3N6O24S15
1049.22
0.35 · 0.35 · 0.32
monoclinic
P21/c
1738.75
0.30 · 0.25 · 0.25
triclinic
measurements (temperature-dependent,
leff = 1.77 BM
ꢀ
(4 K) and leff = 3.29 BM (300 K)), EPR spectroscopy
(g = 2.064) and NO-trapping experiments support the fact
that the dinitrosyl iron complex 1 might be assigned as a
P1
˚
a (A)
11.2637(6)
11.7431(6)
29.1076(15)
90.00
94.255(1)
90.00
3839.5(3)
4
1.815
9.5120(1)
14.2804(1)
27.2859(2)
90.2025(5)
93.5272(4)
90.6861(5)
3699.06(5)
2
1.561
1794
1.43–27.5
67617
˚
b (A)
resonance hybrid of {Fe+1(ꢀNO)2}9 (minor) and {Feꢀ1
-
˚
c (A)
(NO+)2}9 (major) [21,22]. On the basis of previous investiga-
tion and this study, the electronic structure of {Fe(NO)2}
core of {Fe(NO)2}9 DNICs [(L)2Fe(NO)2]ꢀ (L = Se5, S5,
thiolate) is best described as a dynamic resonance hybrid
of {Fe+1(ꢀNO)2}9 and {Feꢀ1(NO+)2}9, modulated by the
coordinated ligands. The variation of the p value from
a (ꢁ)
b (ꢁ)
c (ꢁ)
3
˚
Volume (A )
Z
Dcalc (g cmꢀ3
F(000)
)
2036
1.40–27.5
8823
the magnetic susceptibility fit ðveMxp ¼ ð1 ꢀ pÞvMðfFeþ1
-
h Range (ꢁ)
Total number of
reflections
Number of unique
9
9
ꢀ1
ꢀ
ð NOÞ2g Þ þ pvMðfFe ðþNOÞ2g ÞÞ is significantly con-
trolled by the thiolate/selenolate ligands bound to the
{Fe(NO)2} core, as shown in Scheme 2.
6206
467
16986 (0.0434)
data (Rint
)
This study also shows that thiolates or selenolates bound
to the {Fe(NO)2}9 core may serve to modulate the NO-
release ability of DNICs. As has been known, characteriza-
tion of both protein-bound and low-molecular weight
DNICs (LMW-DNICs) in vitro has been made possible
via their distinctive EPR signals at g = 2.03 [1–11]. In this
study, the findings, EPR signal of g = 2.064 for complex 1
and the poor NO-release activity of 1 and [(PhSe)2-
Fe(NO)2]ꢀ, imply that the LMW-DNICs and protein-bound
DNICs may notexist with selenocysteine residues ofproteins
Number of parameters
838
R1,a wR2 (%) [I > 2r(I)] 4.20, 9.73
5.30, 15.21
7.87, 16.52
1.503 and ꢀ0.597
b
R1,a wR2 (%) all datꢀa3
7.19, 11.51
0.082 and ꢀ0.602
b
˚
Dqmin and Dqmax(e A
)
P
P
a
R = |(Fo ꢀ Fc)|/ Fo.
P
P
2
2 1=2
b
RwF 2
¼
wðF 2o ꢀ F c2Þ = ½w=ðF 2oÞ ꢁ
.
Table 2
Selected bond distances and angles of complexes [PPN][(Se)5Fe(NO)2] (1)
and [K-18-crown-6-ether][S5Fe(NO)2] (20)
1
20
˚
Fe–N1 (A)
Fe–N2
1.680(4)
1.670(4)
2.400(8)
2.4103(8)
1.166(4)
1.174(5)
2.3238(6)
116.62(18)
108.80(0)
107.0
1.682(3)
1.675(4)
2.2859(11)
2.742(11)
1.176(4)
1.176(4)
2.0538(8)
121.26(17)
107.87(1)
171.8(8)
-
[(RE)2Fe(NO)2]
Fe–Se1/Fe–S1
Fe–Se5/Fe–S5
N1–O1
N
S
S
S
S
S
S5
S
Se5
Cl
RE =
S
N2–O2
NH
C=O
Se–Seavg/S–Savg
\N1–Fe–N2
N–Fe–Seavg/N–Fe–Savg
Fe–N–Oavg
Fe-1(NO+)2
domination
Fe+1(·NO)2
domination
Se1–Fe–Se5/S1–Fe–S5
112.23(3)
107.60(4)
See Figs. 1 and 2.
Scheme 2.

T.-N. Chen et al. / Inorganica Chimica Acta 359 (2006) 2525–2533
2533
[14] A. Bo¨ck, K. Forchhammer, J. Heider, W. Leinfelder, G. Sawers, B.
Veprek, F. Zinoni, Mol. Microbiol. 5 (1991) 515.
[15] J. Bourassa, W. DeGraff, S. Kudo, D.A. Wink, J.B. Mitchell, P.C.
Ford, J. Am. Chem. Soc. 119 (1997) 2853.
[16] A.R. Butler, C. Glidewell, M.-H. Li, Adv. Inorg. Chem. 32 (1988)
335.
[17] A.R. Butler, C. Glidewell, A.R. Hyde, J.C. Walton, Inorg. Chim.
Acta 106 (1985) L7.
as ligands. That one equivalent of [Fe(S,S-CN(Et)2)3] in the
presence of one equivalent of S8 was required to completely
convert one equivalent of complex 2 into [S5Fe(l-S)2FeS5]2ꢀ
along with the formation of [(NO)Fe(S,S-CN(Et)2)2] dem-
onstrates that complexes 2/20 act as a NO-delivering species.
This study also shows that complex 20 may serve as a poten-
tially NO-delivering reagent in cell killing in human erythro-
leukemia K562 cell cultures [38–40].
[18] A.R. Butler, C. Glidewell, A.R. Hyde, J.C. Walton, Polyhedron 4
(1985) 797.
[19] W. Beck, R. Grenz, F. Go¨tzfried, E. Vilsmaier, Chem. Ber. 114
(1981) 3184.
Acknowledgment
[20] A.R. Butler, I.L. Megson, Chem. Rev. 102 (2002) 1155.
[21] M.-L. Tsai, C.-C. Chen, I.-J. Hsu, S.-C. Ke, C.-H. Hsieh, K.-A.
Chiang, G.-H. Lee, Y. Wang, J.-M. Chen, J.-F. Lee, W.-F. Liaw,
Inorg. Chem. 43 (2004) 5159.
We thank the National Science Council (Taiwan) for
support of this work.
[22] F.-T. Tsai, S.-J. Chiou, M.-C. Tsai, M.-L. Tsai, H.-W. Huang, M.-
H. Chiang, W.-F. Liaw, Inorg. Chem. 44 (2005) 5872.
[23] H. Ding, B. Demple, Proc. Natl. Acad. Sci. USA 97 (2000) 5146.
[24] L. Hedberg, K. Hedberg, S.K. Satija, B.I. Swanson, Inorg. Chem. 24
(1985) 2766.
[25] T.W. Hayton, P. Legzdins, W.B. Sharp, Chem. Rev. 102 (2002) 935.
[26] C.T.-W. Chu, L.F. Dahl, Inorg. Chem. 16 (1977) 3245.
[27] M.J. Scott, R.H. Holm, Angew. Chem. Int. Ed. Engl. 32 (1993) 564.
[28] S. Fujii, K. Kobayashi, S. Tagawa, Y. Tetsuhiko, J. Chem. Soc.,
Dalton Trans. (2000) 3310.
Appendix A. Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.ica.2006.
02.035.
References
[1] M.C. Kennedy, W.E. Antholine, H. Beinert, J. Biol. Chem. 272
(1997) 20340.
[29] D. Coucouvanis, D. Swenson, P. Stremple, N.C. Baenziger, J. Am.
Chem. Soc. 101 (1979) 3392.
[2] R.H. Morse, S.I. Chan, J. Biol. Chem. 255 (1980) 7876.
[3] J.C. Drapier, C. Pellat, Y. Henry, J. Biol. Chem. 266 (1991) 10162.
[4] M. Lee, P. Arosio, A. Cozzi, N.D. Chasteen, Biochemistry 33 (1994)
3679.
[30] H. Strasdeit, B. Krebs, G. Henkel, Inorg. Chim. Acta 89 (1984) L11.
[31] R.H. Blessing, Acta Crystallogr., Sect. A 51 (1995) 33.
[32] G.M. Sheldrick, SHELXTL, Program for Crystal Structure Determina-
tion, Siemens Analytical X-ray Instruments Inc., Madison, WI, 1994.
[33] W.-F. Liaw, C.-Y. Chiang, G.-H. Lee, S.-M. Peng, C.-H. Lai, M.Y.
Darensbourg, Inorg. Chem. 39 (2000) 480.
´
[5] M. Lepoivre, J.-M. Flaman, P. Bobe, G. Lemaire, Y. Henry, J. Biol.
Chem. 269 (1994) 21891.
[6] M.W. Foster, J.A Cowan, J. Am. Chem. Soc. 121 (1999) 4093.
[7] V.M. Sellers, M.K. Johnson, H.A. Dailey, Biochemistry 35 (1996)
2699.
[34] H. Strasdeit, B. Krebs, G.Z. Henkel, Naturforsch. 41B (1986) 1357.
[35] J.H. Enemark, R.D. Feltham, Coord. Chem. Rev. 13 (1974) 339.
[36] J.B. Mitchell, D.A. Wink, W. DeGraff, J. Gamson, L.K. Keefer,
M.C. Krishna, Cancer Res. 53 (1993) 5845.
[37] W.-F. Liaw, C.-H. Lai, C.-K. Lee, G.-H. Lee, S.-M. Peng, J. Chem.
Soc., Dalton Trans. (1993) 2421.
[38] P.C. Ford, S. Wecksler, Coord. Chem. Rev. 249 (2005) 1382, and
references therein.
[39] P.C. Ford, J. Bourassa, B. Lee, I. Lorkovic, K. Miranda, L.
Laverman, Chem. Rev. 171 (1998) 185, and references therein.
[40] P.C. Ford, J. Bourassa, Coord. Chem. Rev. 200–202 (2000) 887, and
references therein.
[8] A.F. Vanin, Biochem (Moscow) 63 (1998) 782.
[9] P.C. Ford, J. Bourassa, K. Miranda, B. Lee, I. Lorkovic, S. Boggs,
S. Kudo, L. Laverman, Coord. Chem. Rev. 171 (1998) 185.
[10] C.E. Cooper, Biochim. Biophys. Acta 1411 (1999) 290.
[11] J.L. Alencar, K. Chalupsky, M. Sarr, V. Schini-Kerth, A.F. Vanin,
J.-C. Stoclet, B. Muller, Biochem. Pharmacol. 66 (2003) 2365.
[12] R.H. Holm, P. Kennepohl, E.I. Solomon, Chem. Rev. 96 (1996)
2239.
[13] T.C. Stadtman, Annu. Rev. Biochem. 65 (1996) 83.