Efficient increase of DNA cleavage activity of a diiron(III) complex by a conjugating acridine group

Article abstract of DOI:10.1002/ejic.200700383

A new diferric complex, Fe₂L_b, in which a DNA intercalator (acridine) is linked to a precursor diferric complex (Fe ₂L_a), has been designed and synthesised as a hydrolytic cleaving agent of DNA. Compared with Fe

Full text of DOI:10.1002/ejic.200700383

FULL PAPER
DOI: 10.1002/ejic.200700383
Efficient Increase of DNA Cleavage Activity of a Diiron(III) Complex by a
Conjugating Acridine Group
Xiao-Qiang Chen,[a] Xiao-Jun Peng,*[a] Jing-Yun Wang,[b] Yan Wang,[b] Song Wu,[b]
Li-Zhu Zhang,[a] Tong Wu,[a] and Yun-Kou Wu[a]
Keywords: DNA cleavage / Iron complex / Diiron(III) complex / Intercalator / Acridine
A new diferric complex, Fe2Lb, in which a DNA intercalator
(acridine) is linked to a precursor diferric complex (Fe2La),
has been designed and synthesised as a hydrolytic cleaving
agent of DNA. Compared with Fe2La (without the DNA inter-
calator) (La: 2,6-bis{[(2-hydroxybenzyl)(pyridin-2-yl)methyl-
amino]methyl}-4-methylphenol), Fe2Lb [Lb: 5-(acridin-9-yl)-
N-(3,5-bis{[(2-hydroxybenzyl)(pyridin-2-yl)methylamino]-
methyl}-4-hydroxybenzyl)pentanamide] leads to a 14-fold in-
crease in the cleavage efficiency of plasmid DNA due to the
binding interaction between DNA and the acridine moiety.
The interaction has been demonstrated by UV/Vis absorp-
tion, CD spectroscopy, viscidity experiments and thermal de-
naturation studies. The hydrolytic mechanism is supported
by evidence from T4 DNA ligase assay, reactive oxygen spe-
cies (ROS) quenching and BNPP [bis(4-nitrophenyl) phos-
phate, a DNA model] cleavage experiments. The pH depen-
dence of the BNPP cleavage by Fe2La in aqueous buffer me-
dia shows a bell-shaped pH–kobs profile with an optimum
point around a pH of 7.0 which is in good agreement with the
maximum point of the pH-dependent relative concentration
curve of active species from the pH titration experiments.
The determination of the initial rates at a pH of 7.36 as a
function of substrate concentration reveals saturation kinet-
ics with Michaelis–Menten-like behaviour and Fe2La shows
a rate acceleration increase of 4.7ϫ106 times in the hydroly-
sis of BNPP.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2007)
phosphodiester bonds of DNA are exceptionally resistant
to hydrolysis under physiological conditions and it is a great
challenge to find new catalysts to enhance the hydroly-
sis.[9–11] In the past decade, a variety of metal complexes
which can promote the hydrolysis of DNA and phosphodi-
esters have been reported, including complexes based on
CuII, ZnII, CoIII and lanthanide ions.[12–21] Furthermore,
many elegant dinuclear complexes have been developed as
structural or functional models of the enzymes.[22] Com-
pared with mononuclear complexes, dinuclear complexes
have the higher activity as a result of the cooperative inter-
action of metal ions in stabilising the transition state of
phosphodiester cleavage.
Among the physiologically relevant metal ions, iron has
been much less studied in the hydrolytic reaction[23–25] due
to the fact that Fe-dependent DNA cleavage is often in-
volved in oxygen free radical processes such as the Fenton
reaction. In fact, purple acid phosphatases (PAP), compris-
ing diiron active centers can catalyse the hydrolysis of cer-
tain phosphate esters, including nucleoside di- and triphos-
phates and aryl phosphates under weakly acidic condi-
tions.[26] On the other hand, many efforts have so far been
made to enhance the reactivity of metal complexes cleaving
DNA. One of these strategies is the conjunction of metal
complexes and intercalating groups to increase the DNA
affinity of cleavage agents.[27–29] Tecilla et al. synthesised a
series of ZnII complex-intercalator conjugates, the struc-
Introduction
Mimicking the activities of nucleases is currently an at-
tractive research area in molecular biology and therapy ow-
ing to its essential role in artificial restriction enzyme and
conformational probes of nucleic acids.[1–4] A number of
oxidative cleavage reagents have been used with great suc-
cess in DNA footprinting[5] for locating base mismatches
and loop regions,[6] for locating conformational variations
in DNA[7] and as chemotherapeutic agents.[8] However,
these oxidative cleavage agents require the addition of an
external agent (e.g., light or hydrogen peroxide) to initiate
cleavage. Furthermore, oxidative cleavage is mediated by re-
active oxygen species (ROS) which can cause other severe
cytotoxic effects. These drawbacks of the oxidative reagents
hinder their potential applications in vivo. For eliminating
the possibility of significant cytotoxic side effect of ROS,
pathways that result in DNA cleavage by hydrolysis mecha-
nisms have been considered to be preferable. However, the
[a] State Key Laboratory of Fine Chemicals, Dalian University of
Technology,
158 Zhongshan Road, Dalian 116012, P. R. China
Fax: +86-411-88993906
E-mail: pengxj@dlut.edu.cn
[b] Department of Bioscience and Biotechnology, Dalian Univer-
sity of Technology,
Dalian 116023, P. R. China
Supporting information for this article is available on the
WWW under http://www.eurjic.org or from the author.
5400
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 5400–5407

DNA Cleavage Activity of a Diiron(III) Complex
tures of which feature a cis,cis-triaminocyclohexane chelat- spacer is very important in enhancing the activity of the
ing subunit linked to an anthraquinone moiety through conjugate toward DNA cleavage. Consequently, 5-(acridin-
alkyl spacers with different lengths and reactivity studies 9-yl)pentanoic acid was chosen to link with the dinuclear
showed that the length of the spacer which tethers the inter- metal moiety through an imide bond.[28] The acid was syn-
calating unit to the catalytic group is a key element in the thesised according to the method provided by Nadukkudy
cleavage activity.[28] An earlier complex was reported by et al.[31] The compound 2-[3,5-bis(chloromethyl)-4-hydroxy-
Lawrence and co-workers, who appended acridine to a pen- benzyl]isoindole-1,3-dione, reported in our earlier publica-
tadentate N5 ligand and used it as a cofactor for FeII-pro- tions,[32] was incubated with 2-{[(pyridin-2-ylmethyl)-
moted DNA oxidative cleavage.[29] However, examples of amino]methyl}phenol in the presence of triethylamine to
dinuclear complexes appended to the intercalating groups yield the intermediate 2. After deprotection of 2 with hydra-
as hydrolytic agents are surprisingly rare.
zine hydrate in ethanol at room temperature, the amine ob-
In our work, the new diferric complex Fe2Lb is derived tained was then linked to the acridinium subunit by an
from Lb (Scheme 1), the structure of which features a binu- amide bond to afford the ligand Lb in a yield of 50%. The
clear FeIII complex subunit linked an acridine moiety diiron(III) complex [FeIII2(BBPMP)(µ-OAc)2] with acetate
through a long spacer. The intercalating role of acridine has bridging groups using La as a ligand has been previously
been confirmed by UV/Vis absorption, CD spectroscopy, reported by Neves and co-works.[30] In accordance with this
viscidity experiments and thermal denaturation studies. The report, we initially obtained two complexes with acetate
hydrolytic role of the diiron(III) subunit has also been con- bridging groups using La and Lb. Unfortunately, the solubil-
firmed from the efficient cleavage of the BNPP [bis(4-ni- ities in water were too low to carry out DNA experiments.
trophenyl) phosphate]. Furthermore, the conjugate can ef- Based on the modified synthesis, the complexes Fe2La and
fectively promote the hydrolysis of plasmid DNA.
Fe2Lb were obtained and both have high solubility in water.
Attempts to obtain the structures of Fe2La and Fe2Lb by
X-ray diffraction were not successful although Fe2La and
Fe2Lb were characterised by elemental analysis and spectro-
scopic methods (see Figures S1–S11 in the electronic sup-
porting information). The binuclear structures of the com-
plexes were supported by the calculated isotopic distribu-
tions which are in agreement with those measured using
high-resolution ESI-MS (Figures S2 and S6–S9, supporting
information). For Fe2La, the observed species at m/z(z = 1)
= 731.1150 corresponds to the singly-charged diiron species
[FeIII2(La–3H)(OCH3)2]+. The six positive charges due to the
two Fe3+ ions and the five negative charges due to three
phenolates (La–3 H) and the two methoxy ligands (from the
exchange of ligand during ESI-MS using CH3OH as flow-
ing phase) result in a single positively charged species. A
similar analysis can also be applied to Fe2Lb.
pBR322 DNA Cleavage by Binuclear FeIII Complexes
In the absence of any added reductant or oxidant, when
plasmid DNA pBR322 (0.02 µgµLϪ1, 32 µ bp) is incu-
bated at 37 °C for 1 h in 20 m HEPES at pH 7.0 with
Fe2La or Fe2Lb, activity can already be observed with less
than 50 µ of the catalysts and DNA becomes converted
from supercoiled (form I) to nicked (form II) to an extent
which depends on the concentration of the dinuclear com-
plexes (Figure 1). A comparison between the activity of
Fe2La and Fe2Lb provided clear evidence for the important
role of the acridine subunit in promoting the efficiency of
cleavage, while just 5 µ Fe2Lb is required to cleave 81% of
supercoiled DNA after 1 h at 37 °C, the concentration of
Fe2La necessary to provide the same effect is more than
Scheme 1. Molecular structures of ligands used.
Results and Discussion
Synthesis and Characterisation
La was prepared according to a procedure previously de- 20 µ. From Figures 2 and 3 it may also be seen that the
scribed by Neves et al.[30] Lb was synthesised according to relative amount of supercoiled plasmid DNA decreases
the reaction sequences depicted in Scheme S1. The previous upon increasing the incubation time. Through exponential
report by Tecilla and co-workers has shown that the flexible decay fitting, the rate constant of the supercoiled DNA
Eur. J. Inorg. Chem. 2007, 5400–5407
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
5401

X.-J. Peng et al.
FULL PAPER
cleavage for 20 µ Fe2Lb was determined to be pH of 8.0, the supercoiled form was completely degraded
1.0ϫ10–2 s–1 (Figure S12) which corresponds to a half-life immediately after mixing. However, the cleavage system was
of a minute. The rate constant represents a 14-fold increase very complicated because high activity depended on “oxi-
in the cleavage efficiency when compared with the Fe2La dative” conditions, i.e. the presence of H2O2 or O2 and a
species (7.1ϫ10–4 s–1, Figure S13). A similar rate accelera- reductant (dithiothreitol or ascorbate). No rate constant
tion (15-fold) was also reported by Tecilla et al. with a Zn- was reported which might result from the fact that the
triaminocyclohexane complex tethered to an anthraquinone cleavage occurred with the initial burst of activity and the
intercalator.[28] In order to verify the impact of the intercal- extent of cleavage did not increase after lengthening the re-
ation, we carried out the cleavage experiments in the pres- action time. The second diiron(III) complex Fe2(DTPB)
ence of the mixture of acridine (5 µ) and Fe2La (5 µ). As was investigated by Liu and co-workers.[23] In the presence
shown in Figure S14, the extent of DNA cleavage by the of such a complex at 37 °C, a pH of 7.0 and a concentration
mixture of acridine and Fe2La is similar to Fe2La and the of 100 µ, the supercoiled DNA was degraded with a rate
activity is much less than for Fe2Lb (covalently linked with constant of 2.0ϫ10–3 s–1 and the cleavage did not depend
acridine). Furthermore, the kinetic experiment using the on the presence of O2. Compared with Fe2(DTPB), the ac-
mixture show cleavage rate for form I is 6.9ϫ10–4 s–1 (Fig- tivity of Fe2La seems somewhat lower (7.1ϫ10–4 s–1) at
ure S15–S16) which is also similar to Fe2La (7.1ϫ10–4 s–1). 20 µ. However, for Fe2Lb with a rate constant of
In contrast, the rate constant of Fe2Lb is much larger 1.0ϫ10–2 s–1 at a concentration as low as 20 µ, the cleav-
(1.0ϫ10–2 s–1). All of this evidence indicates that conjuga- age efficiency is evidently better than for Fe2(DTPB).
tion between the acridinium group and the diiron(III) sub-
unit is necessary to get a fast cleavage rate.
Complex–DNA Interaction
The DNA binding ability of Fe2Lb was studied using UV
spectroscopy by following the absorbance intensity changes
of the acridine group at 253 nm. Upon addition of increas-
ing amounts of DNA to the complex, 48% hypochromism
was observed (Figure 4) which indicates the interaction be-
Figure 1. Agarose gel electrophoresis of pBR322 plasmid DNA
treated with Fe2La (lane 2–4) and Fe2Lb (lane 5–8) for 1 h in an
HEPES buffer (20 m, pH 7.0) at 37 °C. Lane 1: control DNA,
tween acridine group and DNA. From the plot of
[DNA]/(εa – εf) vs. [DNA], the intrinsic binding constant
of the complex with DNA was calculated to be
(5.0Ϯ0.1)ϫ105 –1. The moderate binding for this com-
plex is comparable to those observed for many other acri-
dine derivatives.[33]
lane 2: Fe2La 10 µ, lane 3: Fe2La 20 µ, lane 4: Fe2La 50 µ, lane
5: Fe2Lb 5 µ, lane 6: Fe2Lb 10 µ, lane 7: Fe2Lb 20 µ, lane 8:
Fe2Lb 50 µ.
Figure 2. Time-dependence of pBR322 plasmid DNA cleavage by
20 µ Fe2La in an HEPES buffer (20 m, pH 7.0) at 37 °C. Lane
1: control, lane 2: Fe2La 30 min, lane 3: 1 h, lane 4: 2 h, lane 5: 4 h,
lane 6: 16 h.
Figure 3. Time-dependence of pBR322 plasmid DNA cleavage by
20 µ Fe2Lb in an HEPES buffer (20 m, pH 7.0) at 37 °C. Lane
1: control, lane 2: Fe2Lb 5 min, lane 3: 10 min, lane 4: 15 min, lane
5: 20 min, lane 6: 30 min, lane 7: 50 min.
The DNA cleaving systems based on mono-ferric com-
plexes mostly involve the FeII ion and DNA cleavage by
these systems, such as Fe-bleomycin and the Fenton agent,
occur by an oxidative mechanism. The FeIII-based agents
so far reported are relatively few although FeIII is present
in the active site of some phosphatases. Two FeIII complexes
with high hydrolytic activity in DNA cleavage have been
reported and both of them are binuclear. The first complex
Fe2(HPTB), described by Schnaith, Que and co-workers in
1994[25] showed a surprising reactivity: in the presence of
Figure 4. Top: absorption spectra of Fe2Lb (0.67ϫ10–5 ) in the
presence of increasing amounts of CT DNA (6.27ϫ10–4 ) at
the complex at a concentration of 10 µ at 25 °C and a 25 °C. Bottom: plot of [DNA]/(εa – εf) vs. [DNA].
5402 Eur. J. Inorg. Chem. 2007, 5400–5407
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

DNA Cleavage Activity of a Diiron(III) Complex
As a means for further clarifying the binding of the binu- tion of DNA by the complexes, reactions were also per-
clear complex, viscosity measurements were carried out on formed under anaerobic conditions (Figure S19) and the
calf thymus DNA by varying the concentration of the com- results showed that Fe2La and Fe2Lb still effectively cleaved
plex. Acridine, a classical intercalator, causes a moderated pBR322 DNA with little inhibition. In contrast, the oxidat-
increase in the viscosity of the DNA solution due to the ive cleaving system Fe(EDTA)2–/DTT showed very strong
separation of base pairs at intercalation sites and hence an oxygen dependence. Another general way to discriminate
increase in overall DNA length (Figure 5). In contrast, the hydrolytic mechanism from the oxidative mechanism is
Fe2La causes a decrease in DNA solution viscosity under the examination of the DNA cleavage in the presence of
the same conditions which can be attributed to the binding reactive oxygen species (ROS) scavengers. Figure 7 indicates
between Fe2La and DNA through charge affinity leading to little effect on the cleavage efficiency of DNA with Fe2Lb
a distortion of the DNA structure. Compared with Fe2La, when DMSO or NaN3 are added and superoxide dismutase
Fe2Lb further decreases the viscosity of DNA and the result (SOD) can even promote the cleavage efficiency. The experi-
suggests that the cooperation between the diiron complex mental results suggest that hydroxyl radical, singlet oxygen
and the acridine subunit makes distortion of the DNA and superoxide are not involved in the diiron DNA cleavage
more intense. The thermal denaturation studies show that system. Additionally, the possibility of hydroxyl radical for-
the addition of Fe2Lb to the calf thymus DNA could raise mation in the presence of Fe2La can also be evaluated by
Tm from (78.3Ϯ0.2) °C to (80.0Ϯ0.5) °C but no change monitoring the changes in absorbance at 552 nm for the dye
was observed in the case of Fe2La (Figure S17). The CD rhodamine B. Degradation of the dye can provide a direct
spectra also show the difference in the interaction between measure of the concentration of the hydroxyl radical
Fe2La and Fe2Lb with CT DNA (Figure S18): Incubation formed.[36] In the presence of Fe2La, no obvious decrease in
of DNA and Fe2La caused the intensities of both the posi- absorbance of the dye was detected over the time of the
tive and negative ellipticity bands to decrease, while incu- experiment (Figure S20) even after H2O2 was added, indi-
bation of DNA and Fe2Lb caused the intensities of the cating that the hydroxyl radical is not formed under these
negative ellipticity band to increase and the positive ellip- conditions. An oxidative pathway can therefore be ruled out
ticity band blue-shift. This suggests the acridine moiety has in the cleavage of DNA by diiron(III) complexes and, evi-
a significant effect on the conformational changes of DNA. dently, the cleavage occurs through a hydrolytic mechanism.
Figure 6. Ligation of pBR322 DNA cleaved by Fe2La and Fe2Lb.
A, lane 1: supercoiled (upper) and nicked (lower) pBR322 DNA,
lane 2: linearised pBR322 DNA cleaved by Fe2La, lane 3: ligation
of linearised pBR322 DNA by T4 DNA ligase, lane 4: λ DNA-
EcoT14 I digest markers. B, lane 1: supercoiled (upper) and nicked
(lower) pBR322 DNA, lane 2: linearised pBR322 DNA cleaved by
Fe2Lb, lane 3: ligation of linearised pBR322 DNA by T4 DNA
Figure 5. Effect of increasing the amounts of the compounds on
the relative specific viscosity of calf thymus DNA. Viscosity is pre-
sented as [η/η0]1/3 in accord with the theory of Cohen and Eisenberg
ligase, lane 4: λ DNA-EcoT14 I digest markers.
(1969). ᭹ acridine + CT DNA, ᭡ Fe2La + CT DNA, ᭿ Fe2Lb
CT DNA. r = Ccompound/CDNA
+
.
Figure 7. Agarose gel electrophoresis of pBR322 plasmid DNA
treated with complex Fe2Lb in various inhibitors of ROS. Incu-
bation time: 1 h (37 °C). Lane 1: control, lane 2: Fe2Lb 50 µ, lane
3: Fe2Lb 50 µ + 1 m DMSO, lane 4: Fe2Lb 50 µ + 1 m NaN3,
lane 5: Fe2Lb 50 µ + 10 U of SOD.
Cleavage Mechanism Analysis
The earlier two diiron(III) complexes Fe2(DTPB) and
Fe2(HPTB), described respectively by Liu[23] and by Que,[25]
have shown the yielding linearised DNA can be quantita-
tively relegated by confirming the occurrence of a hydrolytic
mechanism. In our cases, the linearised DNA has also been
relegated by T4 ligase (Figure 6), although the relegation is
not complete. In some cases, the hydrolytic products either
Hydrolysis of BNPP Catalysed by Fe2La
Cleavage experiments of phosphodiesters were also car-
did not end at the required 5Ј-phosphate and 3Ј-OH (ri- ried out to further verify the hydrolytic activity of diiron
bose) termini or the complex sometimes bound to the ter- complexes. In these kinds of experiments, BNPP is often
mini of cleaved DNA. These reasons would result in the used as a DNA model compound in the assessment of pho-
relegation being incomplete or even failing completely.[34,35] sphodiesterase activity. For ensuring the solubility of the
To explore the effects of molecular oxygen on the degrada- binary of the catalyst-substrate compound, the mixed solu-
Eur. J. Inorg. Chem. 2007, 5400–5407
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
5403

X.-J. Peng et al.
FULL PAPER
tion including DMSO (10 vol.-%) was used. The reaction
was monitored by following the increase in the absorbance
at 400 nm which indicates the release of NP from BNPP.
The determination of the initial rates at a pH of 7.36 as a
function of substrate concentration revealed saturation ki-
netics with Michaelis–Menten-like behaviour (Figure 8). A
linearisation after Lineweaver–Burke gave the following
kinetics values: KM
=
(3.1Ϯ0.1)ϫ10–3 , Vmax
=
(3.1Ϯ0.6)ϫ10–9 molL–1 s–1, and the catalytic constant kcat
= Vmax/[Fe2La]0 = (5.2Ϯ0.2)ϫ10–5 s–1. The second-order
rate constant, kBNPP (= kcat/KM) for BNPP is
(1.7Ϯ0.1)ϫ10–2 –1 s–1 and the substrate binding constant

Kb = 1/KM = (3.2Ϯ0.1)ϫ102 –1. From the kinetic data
obtained under these conditions, a rate acceleration toward
BNPP of ca. (4.7 Ϯ0.2)ϫ106 times was found. The rate is
comparable to that observed for other binuclear transition
metal-based complexes.[17]
Figure 9. Dependence of the reaction rate on the pH under the
following conditions: 10 vol.-% DMSO solution, 50 °C, [Fe2La] =
6.0ϫ10–5 molL–1, [BNPP]
=
2ϫ10–4 molL–1, [buffer]
=
5.0ϫ10–2 molL–1 (buffer: MES or HEPES), I = 0.1 molL–1
(NaNO3).
Figure 10. Potentiometric titration of Fe2La (2.0ϫ10–4 ) with an
increasing amount of 0.01  NaOH in DMSO/H2O (1:9).
Figure 8. Top: saturation kinetic experiments for the cleavage of
BNPP by Fe2La at a pH of 7.36 and 50 °C. I = 0.10  (NaNO3),
[Fe2La]0 = 0.06 m, [HEPES buffer] = 50 m, 10 vol.-% DMSO
solution. Bottom: Lineweaver–Burke plot.
The distribution curves (Figure 11) of the deprotonated
products of Fe2La in a 1:9 DMSO/H2O mixture display
maximum monohydroxo species concentrations at a pH of
6.9 which is in good agreement with the optimum point for
the catalytic activity at a pH of 7.0. These results indicate
that the b species is the active species for the cleavage of
phosphodiesters. As proposed previously [equilibria (1) and
(2)], monohydroxo species can provide a labile Fe–OH2
bond for phosphodiester binding and another FeIII bound
hydroxide group for nucleophilic attack on the phosphorus
atom. The decrease in the reactivity at pH Ͼ 7.0 most prob-
ably arises from the presence of the fully deprotonated form
c, in which the tendency of the OH– ion from the Fe–OH
group to leave becomes reduced even though a more con-
centrated OH– solution can increase the nucleophilic reac-
tivity.
It was noted that the pH dependence of the catalytic ac-
tivity between pH values of 5.9 and 8.5 shows a bell-shaped
profile (Figure 9) with an optimum point at about pH 7.0.
On the other hand, potentiometric titration of Fe2La shows
two successive titratable protons in a 1:9 DMSO/H2O sol-
vent mixture (Figure 10). Sigmoidal fitting gives the pKa
values of 5.3 and 8.5, respectively. From the titration curve,
we propose a processes according to the equilibria repre-
sented by Equations (1) and (2).
5404
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 5400–5407

DNA Cleavage Activity of a Diiron(III) Complex
(CDCl3): δ = 8.60 (d, J = 4.4 Hz, 2 H, Py-H), 7.84 (m, 2 H, Py-
H), 7.65–7.70 (m, 4 H, PhCH2Ph-H), 7.26 (d, J = 6 Hz, 4 H, Py-
H), 7.23 (t, J = 4.4 Hz, 2 H, Py-H), 7.10 (t, J = 7.2 Hz, 2 H,H-Ph),
7.02 (d, J = 7.2 Hz, 2 H, Ph-H), 6.72–6.76 (t, J = 7.2 Hz, 4 H, Ph-
H), 4.71 (s, 2 H, PhϪCH2ϪPht), 3.83 (s, 4 H, NϪCH2ϪPy), 3.80
(s, 4 H, NϪCH2ϪPh), 3.77 (s, 4 H, NϪCH2ϪPh) ppm. ESI-MS:
m/z 706.3 [M + H]+, HRMS calcd. for C43H39N5O5 706.3029,
found 706.3022.
4-Aminomethyl-2,6-bis{[(2-hydroxybenzyl)pyridin-2-ylmethylamino]-
methyl}phenol (3): Compound 2 (0.52 g, 7.4 mmol) was suspended
in EtOH (10 mL) and hydrazine hydrate (0.22 g, 4.5 mmol) was
added. The solution was heated to reflux for 2 h and stirred at
ambient temperature overnight and the solvent was then evapo-
rated under reduced pressure. The resultant solid was treated with
2  HCl then extracted with CH2Cl2 and dried with Na2SO4. Evap-
oration of the solvent yielded 0.36 g (85%) of the product. 1H
NMR (CD3COCD3,): δ = 8.61–8.64 (m, 2 H, PyϪH), 7.79 (dt, J
= 7.7, 1.8 Hz, 2 H, PyϪH), 7.32–7.38 (m, 4 H, PyϪH), 7.07–7.14
(m, 6 H, PhϪH), 6.71–6.76 (m, 4 H, PhϪH), 4.30 (s, 2 H,
PhϪCH2 ϪN), 3.87 (s, 4 H, NϪCH2 ϪPy), 3.80 (s, 4 H,
NϪCH2ϪPh), 3.77 (s, 4 H, NϪCH2ϪPh) ppm. ESI-MS: m/z 576.3
[M + H]+. HR-MS calcd. for C35H37N5O3 576.2975, found
576.2980.
Figure 11. The distribution curves of the protonated products of
Fe2La in DMSO/H2O (1:9). (a: the binuclear FeIII complex with
two water molecules bound to the metal centre; b: the mono-
hydroxo species and c: the dihydroxo one).
Conclusions
In summary, we have designed and synthesised a new
diiron complex Fe2Lb with an acridine group as an intercal-
ator of DNA. The cleavage experiments indicate that the
intercalative function can lead to a 14-fold increase in the
cleavage efficiency. Furthermore, the hydrolytic mechanism
studies demonstrate the potential of binuclear ferric com-
plexes as catalyst models for artificial nucleases. In future
5-(Acridin-9-yl)pentanoic Acid (4):[31] Diphenylamine (1.69 g,
0.01 mol), hexanedioic acid (4.38 g, 0.03 mol) and anhydrous ZnCl2
(6.8 g, 0.05 mol) were mixed well and heated at 230 °C for 20 h.
H2SO4 (20 mL, 20 vol.-%) was then added to the reaction mixture
work, we plan to link diiron(III) complexes with recognis- which was subsequently heated to reflux for 4 h. The mixture was
cooled and neutralised using aqueous NH3 (25 vol.-%) solution. A
solid product precipitated. After purification by silica gel column
chromatography [eluted by a mixture (1:1) of EtOAc and petroleum
ether], the pure compound was obtained. 1H NMR (CD3COCD3,):
δ = 8.40 (d, J = 8.8 Hz, 2 H, Acr-H), 8.15 (d, J = 8.8 Hz, 2 H,
AcrϪH), 7.84 (t, J = 7.6 Hz, 2 H, AcrϪH), 7.64 (t, J = 7.4 Hz, 2
H, AcrϪH), 3.68 (t, J = 7.2 Hz, 2 H, CH2Acr), 2.3 (t, J = 6.4 Hz,
2 H, CH2COOH), 1.75 [m, 4 H, Ϫ(CH2)2Ϫ] ppm. ESI-MS m/z:
278.0 [M – H]–.
able functional groups in the hope of realising the sequence-
selective cleavage of DNA.
Experimental Section
General: When necessary, reactions and manipulations were carried
out under an atmosphere of nitrogen. All reagents and solvents
were analytical reagent grade and were used without further purifi-
cation unless otherwise noted. UV/Vis spectra were recorded with
a Lambda 35 UV/Vis spectrometer. ESI–MS spectra and high-reso-
lution mass spectra were recorded on an HP 1100 MSD and an
HPLC-Q-Tof MS (Micro) instrument, respectively. NMR spectra
were recorded with a 400-MHz Varian INOVA system. IR spectra
were obtained by using an FTIR-430 spectrometer. Elemental
analyses were performed on a Vario EL III instrument. The
potentiometric titration was performed on a PHS-3C pH meter.
Viscosity measurements were recorded with a NXE-1B viscometer.
Circular dichroism spectra of DNA were obtained with a JASCO J-
20 automatic recording spectropolarimeter. The ImageMaster VDS
system (Pharmacia Biotech) was used for electrophoresis experi-
ments.
5-(Acridin-9-yl)-N-(3,5-bis{[(2-hydroxybenzyl)(pyridin-2-yl)methyl-
amino]methyl}-4-hydroxybenzyl)pentanamide (Lb): 5-(Acridin-9-yl)-
pentanoic acid (0.126 g, 0.45 mmol) was heated to reflux in SOCl2
(1 mL) for 2 h and the excess SOCl2 was removed under reduced
pressure. The dry solid was redissolved in dry acetonitrile (2 mL)
and cooled in an ice bath. Compound 3 (0.22 g, 0.38 mmol) and
triethylamine (0.5 mL) were added to the acetonitrile solution
which was stirred under argon overnight as the temperature grad-
ually returned to room temperature. Removing the solvent and pu-
rification on a column of silica gel using MeCN/H2O (95:5) as the
eluent gave 0.16 g (50%) of the desired product. 1H NMR (CD3Cl):
δ = 8.56 (d, J = 4.8 Hz, 2 H, PyϪH), 8.35 (d, J = 8.8 Hz, 2 H,
AcrϪH), 8.2 (d, J = 8.8 Hz, 2 H, AcrϪH), 7.77 (t, J = 7.6 Hz, 2
H, AcrϪH), 7.60 (t, J = 7.8 Hz, 2 H, PyϪH), 7.53 (t, J = 7.6 Hz,
2 H, AcrϪH), 7.15–7.21 (m, 4 H, PyϪH), 7.10 (t, J = 7.6 Hz, 2 H,
PhϪH), 6.96 (s, 2 H, PhϪH), 6.92 (d, J = 8 Hz, 2 H, PhϪH), 6.79
(d, J = 6.4 Hz, 2 H, PhϪH), 6.93 (t, J = 7.6 Hz, 2 H, PhϪH), 6.36
(br., 1 H, ϪNHCOϪ), 4.27 (d, J = 5.2 Hz, 2 H, NHϪCH2ϪPh),
3.77 (s, 4 H, NϪCH2ϪPy), 3.71 (s, 4 H, NϪCH2ϪPh), 3.65 (s, 4
H, NϪCH2ϪPh), 3.61 (t, J = 7.2 Hz, 2 H, CH2ϪAcr), 2.34 (t, J =
7.2 Hz, 2 H, CH2ϪCONH), 1.95–1.99 (m, 2 H, ϪCH2CH2Ϫ),
Syntheses
Bis(4-nitrophenyl) phosphates (BNPP) were prepared and purified
following a modified literature method.[37]
2-(4-Hydroxy-3,5-bis{[(2-hydroxybenzyl)pyridin-2-ylmethylamino]-
methyl}benzyl)isoindole-1,3-dione (2) (Scheme S1): A solution of 2-
[3,5-bis(chloromethyl)-4-hydroxybenzyl]isoindole-1,3-dione (0.61 g,
3.0 mmol), 2-{[(pyridin-2-yl)methylamino]methyl}phenol (1.40 g,
6.5 mmol) and Et3N in CH2Cl2 was stirred at ambient temperature
for 2 d. The reaction mixture was then diluted with CH2Cl2 and
washed with brine. The organic phase was dried with Na2SO4 and
the solvent was removed by evaporation. Purification on silica gel
1.82–1.84 (m,
2 H, ϪCH2CH2Ϫ) ppm. ESI-MS m/z 837.5
[M + H]+, HRMS calcd. for C53H52N6O4 837.4128, found
837.4135.
Synthesis of the Binuclear Complexes Fe2LaCl3·3H2O (Fe2La) and
Fe2LbCl3·4H2O (Fe2Lb): La (0.05 g, 0.09 mmol) and 2 equiv.
1
using EtOAc as eluent gave the desired product (71%). H NMR
Eur. J. Inorg. Chem. 2007, 5400–5407
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
5405

X.-J. Peng et al.
FULL PAPER
FeCl3·6H2O (0.049 g, 0.18 mmol) were dissolved in H2O and
heated at 50 °C for 1 h to obtain a dark blue solution and the
solution was then concentrated under reduced pressure. The target
for the free iron complex and the extinction coefficient for the iron
complex in the fully bound form, respectively. The intrinsic binding
constant Kb was given by the ratio of the slope to intercept through
product Fe2La was the obtained as
a
solid residue.
the plot of CDNA/(εa – εf) vs. CDNA.
C35H39Cl3Fe2N4O6 (Fe2LaCl3·3H2O): calcd. C 50.66, H 4.74, N
Viscosity Measurements: Viscosity experiments were carried out on
a NXE-1B viscometer at 25 °C. The concentration of DNA was
280 µ in the nucleotide phosphate and the flow times were deter-
mined with a manually operated timer. Data were presented as
[η/η0]1/3 vs. [complex]/[DNA] (η is the viscosity of DNA solution
in the presence of the compound and η0 is the viscosity of DNA
alone).
6.75; found C 50.64, H 4.49, N 6.81. High-Resolution ESI-MS
(CH3OH as flowing phase) (Figure S1–S2) m/z: [FeIIIFeIII
-
,
(La–3H)(OCH3)(OCH3)]+ 731.1150; [FeIIIFeIII (La–3 H) (OCH3)]2+
350.1511. IR (KBr) (Figure S3): ν = 3427–3066 (br., s), O–H of
˜
water and C–H of aromatic ring; 2922 (m) and 2855 (m), C–H of
methane; 1609 (s) and 1475 (s), C=C of aromatic ring; 1264 (s), C–
O of phenol; 759 cm–1 (s), C–H of aromatic ring. UV/Vis (H2O):
λmax (εm, Lmol–1 cm–1) (Figure S4): 260 (shoulder, 20000) attributed
to the π–π* transition of the ligand; 531 nm(4400) attributed to the
iron(III)Ϫligand charge transfer. Similarly, spectroscopic data of
Fe2Lb were obtained. C53H57Cl3Fe2N6O8 (Fe2LbCl3·4H2O): calcd.
C 56.63, H 5.11, N 7.48; found C 56.92, H 4.45, N 7.62. High-
Resolution ESI-MS (CH3OH as flowing phase) (Figure S5–S9)
m/z: [FeIIIFeIII (Lb–3 H) (OCH3)H]3+ 325.6591; [FeIIIFeIII (Lb–3
H) (OH)]2+ 480.9854; [FeIIIFeIII (Lb–3 H) (OH)(HCl)]2+ 498.9745;
[FeIIIFeIII (Lb–3 H) (OCH3)(HCl)]2+ 505.9684; [FeIIIFeIII (Lb–3 H)
Ligation Experiment on the DNA Fragments: Enzymatic assays
were performed using T4 DNA ligase to determine whether the
cleaved products were consistent with hydrolysis of the phosphodi-
ester linkages in DNA. First, the linearised DNA was isolated from
the pBR 322 DNA cleavage products by Fe2La or Fe2Lb. Then, the
following reaction was carried out for relegation: 1.5 µL of 10ϫ
ligation buffer, 10 µL of linearised pBR 322 DNA, 1 µL of T4 li-
gase (2 units), and 2.5 µL of H2O were mixed and incubated at
16 °C for 24 h. The ligation products were monitored by the elec-
trophoresis and visualised by staining in an ethidium bromide solu-
tion.
(OAc)(HCl)]2+ 520.9583. IR (KBr) (Figure S10): ν = 3417–3000
˜
(br., s), O–H of water and C–H of aromatic ring; 2925 (m) and
2857 (m), C–H of methane; 1635 (s) and 1478 (s), C=C of aromatic
ring; 1265 (strong), C–O of phenol; 759 cm–1 (strong), C–H of aro-
matic ring. UV/Vis (H2O): λmax (εm, Lmol–1 cm–1) (Figure S11): 253
(56300) and 354 (13300) attributed to the π–π* transition of ligand;
531 nm (2980) attributed to the iron(III)Ϫligand charge transfer.
Anaerobic Reactions: Deoxygenated water was prepared by four
freeze-pump-thaw cycles. Before each cycles the water was equili-
brated with nitrogen to assist the deoxygenating process. The
deoxygenated water was stored under a nitrogen atmosphere prior
to use. All anaerobic stock solutions were prepared in Schlenk
tubes under nitrogen using the deoxygenated water. Reaction mix-
tures were prepared by the addition of the appropriate volumes of
the stock solutions to the tubes and were incubated in a nitrogen-
filled glovebag.
pBR322 DNA Cleavage Experiments by the Binuclear Metal Com-
plexes: DNA cleavage experiments were performed using pBR 322
(Takara) in a 20 m HEPES buffer, pH 7.0, with incubating DNA
(0.4 µg) at 37 °C in the presence and absence of the metal complex
for the specified time. The cleavage reactions were stopped by ad-
dition of 3 µL of a loading buffer (0.25% bromphenol blue, 25%
glycerol, 1 m EDTA, 2% SDS). The electrophoresis of DNA
cleavage products was performed on 1% agarose gel. The gels were
run at 130 V for 45 min in a 0.01  pH 7.0 TAE buffer. The re-
solved bands were visualised by ethidium bromide staining and
were then quantified. A correction factor of 1.22 was utilised to
account for the decreased ability of ethidium bromide to intercalate
into form I DNA compared with forms II and III.[38]
Thermal Denaturation of DNA: In 0.01  Na2HPO4 buffer
(pH 7.0), the absorption values of DNA as well the mixture of
DNA and dinuclear FeIII complexes at 260 nm were recorded by
increasing the temperature in 5 °C increments from 55 °C to 95 °C
in a thermostatic bath where the controls were the buffer and the
buffers in the absence and presence of the diiron complex, the con-
centration of which was the same as that in the sample.
CD Spectra Measurements: Circular dichroism spectra of DNA
were obtained on a JASCO J-20 automatic recording spectropolari-
meter operating at 25 °C. The region between 200 and 300 nm was
scanned for each sample.
Cleavage of pBR322 in the Presence of ROS Scavengers: Scavengers
of reactive oxygen intermediates (DMSO, NaN3 and SOD) were
added alternatively to the reaction mixtures. Cleavage was initiated
by the addition of Fe2Lb and stopped by 3 µL of a loading buffer
(0.25% bromphenol blue, 25% glycerol, 1 m EDTA, 2% SDS).
Further analysis was conducted using the standard procedures de-
scribed above.
Kinetic Measurements: The kinetic measurements were performed
at 50 °C in buffered solutions of a 1:9 mixture of DMSO/water. 2-
(Morpholino)ethanesulfonic acid (MES) and 2-[4-(2-hydroxyethyl)-
piperazin-1-yl]ethanesulfonic acid (HEPES) were used as buffers.
The ionic strength was fixed to 0.1  with sodium nitrate. In a
typical experiment, aqueous buffer solution (1.5 mL) was mixed
with complex Fe2La stock solution (in water) (0.9 mL) and DMSO
(0.3 mL) in a temperature-controlled spectrometric cell. After the
mixture was equilibrated for 15 min, BNPP stock solution (in
water) was added (0.3 mL) and data collection was started immedi-
ately. The cleavage of BNPP was measured by following the in-
crease in the absorption of 4-nitrophenolate (NP) at 400 nm. NP
concentrations were determined using the molar extinction coeffi-
cients for NP of 2500, 4500, 9000, 12800, 16800, 17900,
19900 –1 cm–1 at pH values of 5.95, 6.36, 6.96, 7.36, 7.85, 8.13 and
8.64, respectively. The activities of the complexes were determined
by the method of initial rates. At least three independent measure-
ments were made.
Absorption Titration of Fe2Lb Binding to DNA: The solutions of CT
DNA gave a UV absorbance ratio at 260 and 280 nm, A260/A280
Ͼ
1.8, indicating that the DNA was sufficiently free of protein.[39] The
concentrated stock solution of CT DNA (stored at 4 °C and not
used for more than 6 d) was prepared and the concentration of
DNA in the nucleotide phosphate was determined by UV ab-
sorbance at 260 nm after a 1:10 dilution. The molar absorption
coefficient of CT DNA was taken as 6600 –1 cm–1.[39] A solution
(3 mL) of 0.67ϫ10–5  Fe2Lb in water was incubated for 30 min at
25 °C. With addition of increasing amounts of 2 µL aliquots of
DNA (6.27ϫ10–4 ), the influence of volume was found to be neg-
ligible. The binding constant was determined using the following
equation: CDNA/(εa – εf) = CDNA/(εb – εf) + 1/Kb(εb – εf), where εa,
εf, and εb correspond to Aobsd/[complex], the extinction coefficient
Potentiometric Titration of Fe2La: To an aqueous solution of Fe2La
(2.0ϫ10–4 ) containing 10 vol.-% DMSO was added a solution of
5406
www.eurjic.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2007, 5400–5407

DNA Cleavage Activity of a Diiron(III) Complex
[14] R. Ren, P. Yang, W. J. Zheng, Z. C. Hua, Inorg. Chem. 2000,
39, 5454–5463.
[15] S. Matsuda, A. Ishikubo, A. Kuzuya, M. Yashiro, M. Komi-
yama, Angew. Chem. Int. Ed. 1998, 37, 3284–3286.
[16] O. Iranzo, A. Y. Kovalevsky, J. R. Morrow, J. P. Richard, J. Am.
Chem. Soc. 2003, 125, 1988–1993.
[17] J. Chen, X. Y. Wang, Y. G. Zhu, J. Lin, X. L. Yang, Y. Z. Li,
Y. Lu, Z. J. Guo, Inorg. Chem. 2005, 44, 3422–3430.
[18] N. H. Williams, J. Chin, Chem. Commun. 1996, 131–132.
[19] N. H. Williams, W. Cheung, J. Chin, J. Am. Chem. Soc. 1998,
120, 8079–8087.
0.01  NaOH dropwise. Each data point was obtained by adding
2–20 µL of NaOH solution at 25 °C.
Supporting Information (see also the footnote on the first page of
this article): Synthesis of Lb (Scheme S1). High-Resolution ESI-
MS spectrum of Fe2La (Figure S1, S2). IR spectra of Fe2La (Figure
S3). Absorption spectra of 60 µ Fe2La (Figure S4). High-Resolu-
tion ESI-MS spectrum of Fe2Lb (Figure S5–S9). IR spectra of
Fe2Lb (Figure S10). Absorption spectra of 60 µ Fe2Lb (Figure
S11). Time course of DNA cleavage by 20 µ Fe2Lb (Figure S12).
Time course of DNA cleavage by 20 µ Fe2La (Figure S13). Aga-
rose gel electrophoresis of pBR322 DNA treated with Fe2La, the
mixture of acridine and Fe2La, Fe2Lb (Figure S14). Time-depen-
dence of DNA cleavage by the mixture of 20 µ Fe2La and 20 µ
acridine (Figure S15). Time course of DNA cleavage by the mixture
of 20 µ Fe2La and 20 µ acridine (Figure S16). Thermal Denatur-
ation studies of DNA (Figure S17). The CD spectra of CT DNA
with complexes (Figure S18). Anaerobic and aerobic cleavage of
pBR322 DNA by diiron(III) complexes and Fe(EDTA)2–/DTT
(Figure S19). Quantification of hydroxyl radicals by following the
degradation of rhodamine B (Figure S20).
[20] P. E. Jurek, A. M. Jurek, A. E. Martell, Inorg. Chem. 2000, 39,
1016–1020.
[21] C. A. Chang, B. H. Wu, B. Y. Kuan, Inorg. Chem. 2005, 44,
6646–6654.
[22] N. H. Williams, B. Takasaki, M. Wall, J. Chin, Acc. Chem. Res.
1999, 32, 485–493.
[23]
[24]
[25]
[26]
C. L. Liu, S. W. Yu, D. F. Li, Z. R. Liao, X. H. Sun, H. B. Xu,
Inorg. Chem. 2002, 41, 913–922.
F. Verge, C. Lebrun, M. Fontecave, S. Menage, Inorg. Chem.
2003, 42, 499–507.
L. H. Schnaith, R. S. Hanson, L. Que Jr, Proc. Natl. Acad. Sci.
USA 1994, 91, 569–573.
B. Vincent, G. L. Olivier-Lilley, B. A. Averill, Chem. Rev. 1990,
90, 1447–1467.
Acknowledgments
[27] K. D. Copeland, M. P. Fitzsimons, R. P. Houser, J. K. Barton,
Biochemistry 2002, 41, 343–356.
[28] E. Boseggia, M. Gatos, L. Lucatello, F. Mancin, S. Moro, M.
Palumbo, C. Sissi, P. Tecilla, U. Tonellato, G. Zagotto, J. Am.
Chem. Soc. 2004, 126, 4543–4549.
This work was supported by the Ministry of Education of China
and National Natural Science Foundation of China (projects
20376010 and 20472012).
[29] G. Roelfes, M. E. Branum, L. Wang, L. Que Jr, B. L. Feringa,
J. Am. Chem. Soc. 2000, 122, 11517–11518.
[30] A. Neves, M. A. de Brito, I. Vencato, V. Drago, K. Griesar, W.
Haase, Inorg. Chem. 1996, 35, 2360–2368.
[31] V. E. Nadukkudy, S. Muthusami, R. Danaboyina, Synth. Com-
mun. 1999, 29, 4007–4014.
[32] F. Shi, H. Y. Li, X. J. Peng, R. Zhang, X. Q. Chen, J. L. Fan,
L. C. Sun, Chin. Chem. Lett. 2004, 15, 1407–1410.
[33] J. Joseph, E. Kuruvilla, A. T. Achuthan, D. Ramaiah, G. B.
Schuster, Bioconjugate Chem. 2004, 15, 1230–1235.
[34] M. Scarpellini, A. Neves, R. Hörner, A. J. Bortoluzzi, B. Szpo-
ganics, C. Zucco, R. A. N. Silva, V. Drago, A. S. Mangrich,
W. A. Ortiz, W. A. C. Passos, M. C. B. de Oliveira, H. Terenzi,
Inorg. Chem. 2003, 42, 8353–8365.
[1] K. E. Erkkila, D. T. Odom, J. K. Barton, Chem. Rev. 1999, 99,
2777–2796.
[2] K. Lang, J. Mosinger, D. M. Wagnerova, Coord. Chem. Rev.
2004, 248, 321–350.
[3] H. T. Chifotides, K. R. Dunbar, Acc. Chem. Res. 2005, 38, 146–
156.
[4] A. Sreedhara, J. A. Cowan, J. Biol. Inorg. Chem. 2001, 6, 337–
347.
[5] W. K. Pogozelski, T. J. McNeese, T. D. Tuilius, J. Am. Chem.
Soc. 1995, 117, 6428–6433.
[6] C. J. Burrows, S. E. Rokita, Acc. Chem. Res. 1994, 27, 295–301.
[7] M. D. Kuwabara, D. S. Sigman, Biochemistry 1987, 26, 7234–
7238.
[8] J. Stubbe, J. W. Kozarich, Chem. Rev. 1987, 87, 1107–1136.
[9] J. Rammo, R. Hettich, A. Roigk, H. J. Schneider, Chem. Com-
mun. 1996, 105–107.
[35] C. Sissi, F. Mancin, M. Gatos, M. Palumbo, P. Tecilla, U.
Tonellato, Inorg. Chem. 2005, 44, 2310–2317.
[36] A. Sreedhara, J. D. Freed, J. A. Cowan, J. Am. Chem. Soc.
2000, 122, 8814–8824.
[37] A. Williams, A. N. Richard, J. Chem. Soc. B 1971, 1973–1979.
[38] R. P. Hertzberg, P. B. Dervan, J. Am. Chem. Soc. 1982, 104,
313–315.
[10] M. Yashiro, A. Ishikuno, M. Komiyama, J. Chem. Soc., Chem.
Commun. 1995, 1793–1794.
[11] M. P. Fitzsimons, J. K. Barton, J. Am. Chem. Soc. 1997, 119,
3379–3380.
[12] M. W. Göbel, Angew. Chem. Int. Ed. Engl. 1994, 33, 1141–
1143.
[13] F. H. Fry, A. J. Fischmann, M. J. Belousoff, L. Spiccia, J.
Brugger, Inorg. Chem. 2005, 44, 941–950.
[39] M. E. Reichmann, S. A. Rice, C. A. Thomas, P. Doty, J. Am.
Chem. Soc. 1954, 76, 3047–3053.
Received: April 8, 2007
Published Online: October 16, 2007
Eur. J. Inorg. Chem. 2007, 5400–5407
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
5407