Cu(II) complexes with hydrazone-functionalized phenanthrolines as self-activating metallonucleases

Article abstract of DOI:10.1016/j.ica.2017.11.015

Cu(II) phenanthroline complexes are able to cleave DNA and show cytotoxic activity against cancer cells. Also, the biological activity of hydrazone-based compounds has been reported in the literature. This motivated us to combine these two systems. Hydraz

Full text of DOI:10.1016/j.ica.2017.11.015

Inorganica Chimica Acta xxx (2017) xxx–xxx
Contents lists available at ScienceDirect
Inorganica Chimica Acta
journal homepage: www.elsevier.com/locate/ica
Research paper
Cu(II) complexes with hydrazone-functionalized phenanthrolines as
self-activating metallonucleases
⇑
Julian Heinrich, Jessica Stubbe, Nora Kulak
Institut für Chemie und Biochemie, Freie Universität Berlin, Fabeckstraße 34/36, 14195 Berlin, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 24 July 2017
Received in revised form 5 November 2017
Accepted 8 November 2017
Available online xxxx
Cu(II) phenanthroline complexes are able to cleave DNA and show cytotoxic activity against cancer cells.
Also, the biological activity of hydrazone-based compounds has been reported in the literature. This
motivated us to combine these two systems. Hydrazone-functionalized phenanthroline ligands with
acetyl and benzoyl substituents were synthesized and characterized. The DNA cleavage activity of the
corresponding Cu(II) complexes 3 and 4 was compared to the one of the well-known [Cu(phen)2]2+ com-
plex (1) and [Cu(phenD)2]2+ complex (2, phenD = 1,10-phenanthroline-5,6-dione, the precursor of hydra-
zone-functionalized phenanthrolines). In the presence of ascorbate, but also in its absence, oxidative
cleavage of DNA was proven by quenching therein involved reactive oxygen species (ROS), specifically
hydrogen peroxide and superoxide anion radicals. Only complexes 2–4 were active in the absence of
ascorbate, but not [Cu(phen)2]2+ (1). The surprising occurrence of ROS in the absence of any activating
agent indicates that 2–4 represent self-activating artificial metallonucleases. Thereby, self-activation
was most prominent for the novel hydrazone-based complexes 3 and 4. Cyclic voltammetry and circular
dichroism spectroscopy were applied to get an insight in redox behavior and DNA binding characteristics
of the Cu(II) complexes and thus to explain their different nuclease activity.
Keywords:
Copper(II) complexes
Self-activating metallonucleases
Phenanthroline
Hydrazone
DNA cleavage
DNA binding
Ó 2017 Elsevier B.V. All rights reserved.
1. Introduction
versatile biological activity [9]. Combining both systems to form
a Cu(II) complex with hydrazone-functionalized phenanthrolines
In the last few decades 1,10-phenanthroline (phen) has played
an important role in coordination chemistry due to its broad field
of applications. This ligand has received enormous interest due
to its rich functionalization chemistry, its chelating properties
towards several metal ions and biological activity of the corre-
sponding metal complexes [1]. The derivatization of phenanthro-
line in positions 2 and 9 as well as 5 to hydrazones is well
investigated. Thereby hydrazone-functionalized phenanthrolines
can coordinate several metal ions, e.g. Re(I), Pb(IV), Sn(IV), Gd
(III), La(III) and Cu(II) at either the phenanthroline moiety or both
the phenanthroline and hydrazone moieties [2–5]. To the best of
our knowledge, coordination of phenanthrolines with hydrazone
functionalization in position 5 to Cu(II) ions over the phenanthro-
line scaffold has not been described before.
could lead to novel artificial nucleases with enhanced biological
activity. Such an approach is of general interest in biological/
medicinal inorganic chemistry, but it also bears potential for the
development of well-tolerated anticancer drugs. The most success-
ful metal-based anticancer drug, cisplatin, suffers from causing
severe side effects. Whereas a compound comprising copper as
an endogenous metal, instead of platinum, could guarantee lower
systemic toxicity [10].
In general there are two common pathways, in which artificial
Cu(II) metallonucleases degrade DNA: oxidative and hydrolytic
cleavage. The latter cleavage mechanism involves the phosphate
backbone of DNA, and is enzymatically reversible [11]. The oxida-
tive cleavage mechanism implies a double-strand break subse-
quent to the generation of reactive oxygen species (ROS) [12,13].
The latter cleaving mechanism is irreversible, which is important
for the potential application of such metallonucleases as
chemotherapeutic agents. In general, the oxidative DNA cleavage
is common for complexes, which can be activated by a reductant.
They can then initiate the generation of ROS in an aerobic environ-
ment while being re-oxidized themselves [14,15]. Besides the ini-
tiation of redox processes, another possibility for the activation
The Cu(II) complex of the unsubstituted phenanthroline, [Cu
(phen)2]2+, is still up to date one of the most efficient artificial met-
allonucleases since its discovery in the ‘70 s [6–8]. On the other
hand, hydrazone-based compounds have been exploited for their
⇑
Corresponding author.
E-mail address: nora.kulak@fu-berlin.de (N. Kulak).
https://doi.org/10.1016/j.ica.2017.11.015
0020-1693/Ó 2017 Elsevier B.V. All rights reserved.
Please cite this article in press as: J. Heinrich et al., Inorg. Chim. Acta (2017), https://doi.org/10.1016/j.ica.2017.11.015

2
J. Heinrich et al. / Inorganica Chimica Acta xxx (2017) xxx–xxx
of metallonucleases is using light as a trigger. This has been
broadly covered in the literature [16]. Some of the rarest DNA
cleaving agents are the self-activating metallonucleases. In this
case, the ligand should be redox active replacing the external
reducing agent. On the one hand for Cu(II) complexes, the ligand
could be oxidized – as seen with prodigiosins or hydrazones – to
create a reduced Cu(I) species, which in turn generates ROS in a
cascade manner as described above [17–19]. On the other hand,
the reduced form of the ligand can be stabilized – as seen with
hydroxy-salens – to gain a rare Cu(III) species, which can bind
molecular oxygen for the generation of superoxide anion radicals
without the need for any reducing agent [20].
The respective hydrazide was dissolved in ethanol, phenD and a
catalytic amount of p-toluenesulfonyl chloride was added drop-
wise. The mixture was stirred under reflux for 6 h. After storing
overnight at À24 °C the formed precipitate was filtered off. The
crude product was recrystallized in ethanol to obtain the respec-
tive hydrazone-functionalized phenanthroline.
2.2.2.1. N’-(6-oxo-1,10-phenanthroline-5(6H)-ylidene)acetohydrazide
(phenAH). Acetohydrazide (0.194 g, 2.515 mmol), ethanol (100
mL), phenD (0.505 g, 2.403 mmol) and p-toluenesulfonyl chloride
(0.015 g, 0.077 mmol) were used for the synthesis of the ligand
phenAH, a yellow solid. Yield: 51%. Anal. Elemental analysis: Calc.:
Only a few examples of self-activating metallonucleases based
on Cu(II) complexes with either phenanthroline or hydrazone
ligands, where DNA is cleaved by generation of ROS in the absence
of a reducing agent, are described in literature [20–22]. Thus, the
aim of this work was the synthesis of novel Cu(II) complexes with
hydrazone-functionalized phenanthrolines and the evaluation of
their biological activity regarding the DNA cleavage properties in
a self-activating manner.
C
14H10N4O2 (%): C, 63.15; H, 3.79; N, 21.04. Found: C, 63.15; H,
3.95; N, 21.03. 1H NMR (400 MHz, DMSO d6): d = 2.49 (s, 3Hacetyl);
7.73 (m, 2Hphen); 8.59 (m, 2Hphen); 8.87–9.11 (dd, 2Hphen, J = 4.40
Hz, J = 90.06 Hz); 13.96 (s, 1HH) ppm. HR ESI-MS: m/z [M + H]+
Calc.: 267.0877, Found: 267.0892; [M + Na]+ Calc.: 289.0696,
Found: 289.0741.
2.2.2.2.
N’-(6-oxo-1,10-phenanthroline-5(6H)-ylidene)benzohy-
drazide (phenBH). Benzhydrazide (0.324 g, 2.380 mmol), ethanol
(100 mL), phenD (0.502 g, 2.390 mmol) and p-toluenesulfonyl
chloride (0.013 g, 0.070 mmol) were utilized for the synthesis of
the ligand phenBH, a yellow-orange solid. Yield: 78% (literature:
83% [2]). Anal. Elemental analysis: Calc.: C19H12N4O2 (%): C,
69.51; H, 3.68; N, 17.06. Found: C, 69.59; H, 3.76; N, 17.12. 1H
2. Experimental part
2.1. Materials and methods
All chemicals and solvents were purchased from Acros Organics,
Alfa Aesar, Fisher Scientific, Carl Roth or Sigma Aldrich and were
used without further purification. 1H NMR spectra of the ligands
were recorded in CDCl3 or DMSO d6 on a Jeol ECX 400 spectrometer
at room temperature. Chemical shifts are given relative to TMS
with positive d values indicating a low field shift. The characteriza-
tion of the ligands and their Cu(II) complexes was performed via
elemental analysis on a vario EL CHNS, UV/Vis spectroscopy on
an Agilent Cary 100 instrument and electrospray ionization mass
spectrometry (ESI-MS) on an Agilent 6210 ESI-TOF using methanol
and ethanol, respectively, for the ligands and acetonitrile for the
complexes (flow rate 10 mL/min).
NMR (400 MHz, DMSO d6): d = 7.70 (m, 5Hbenzoyl); 8.01 (d, 2Hphen
,
J = 7.87 Hz); 8.64 (dd, 2Hphen, J = 8.03 Hz, J = 16.99 Hz); 8.90–9.11
(dd, 2Hphen, J = 2.77 Hz, J = 86.39 Hz) ppm. HR ESI-MS: m/z [M +
H]+ Calc.: 329.1033, Found: 329.1060; [M + Na]+ Calc.: 351.0852,
Found: 351.0893.
2.3. Synthesis of the Cu(II) complexes
The Cu(II) complexes were prepared according to the following
general procedure based on literature reports for the synthesis of
the compounds [Cu(phen)2](NO3)2 (1(NO3)2) and [Cu(phenD)2]
(NO3)2 (2(NO3)2) [24,25].
Cu(NO3)2Á3H2O was dissolved in a small amount of ethanol. The
ligand was suspended in ethanol and the suspension was heated
up to 60 °C before the Cu(NO3)2Á3H2O solution was slowly added.
A 1:2 ratio of Cu(II):ligand was ensured, when both components
were combined. The occurring blue (1), green (2), red-brownish
(3) or green-brownish (4) solution was heated to reflux for 1 h.
Upon storage of the solution at À24 °C overnight the turquoise 1
(NO3)2, mint green 2(NO3)2, brown 3(NO3)2 and lime green 4
(NO3)2 precipitates were filtered off and dried in vacuo.
2.2. Synthesis of the ligands
2.2.1. 1,10-Phenanthroline-5,6-dione (phenD)
The ligand phenD was prepared by oxidation of commercially
available phen according to the literature [23].
A mixture of potassium bromide (29.750 g, 0.250 mol) and phen
(5.000 g, 0.028 mol) in 96% sulfuric acid (75 mL) and 65% nitric acid
(37.5 mL) was refluxed and stirred for 2 h. The reaction mixture
was diluted with water (1000 mL) and neutralized with sodium
bicarbonate. The mixture was then extracted three times with
dichloromethane (3 Â 300 mL). The organic phase was dried over
sodium sulfate for 30 min and the solvent was evaporated under
reduced pressure. The obtained crude product was recrystallized
in methanol to obtain phenD as a yellow powder. Yield: 30%. Anal.
Elemental analysis: Calc.: C12H6N2O2∙0.2 CH3OH (%): C, 67.65; H,
3.16; N, 12.93. Found: C, 67.28; H, 2.95; N, 13.23. 1H NMR (400
MHz, CDCl3): d = 7.59 (dd, 2Hphen, J = 4.68 Hz, J = 7.87 Hz); 8.50
(dd, 2Hphen, J = 1.85 Hz, J = 7.87 Hz); 9.12 (dd, 2Hphen, J = 1.84 Hz,
J = 4.70 Hz) ppm. HR ESI-MS: m/z [M + H]+ Calc.: 211.0502, Found:
211.0533; [M + Na]+ Calc.: 233.0321, Found: 233.0365.
2.3.1. [Cu(phen)2](NO3)2 (1(NO3)2)
Cu(NO3)2Á3H2O (0.135 g, 0.560 mmol), phen (0.200 g, 0.110
mmol) and 30 mL ethanol were used for the synthesis of 1(NO3)2,
a turquoise solid. Yield: 73%. Anal. Elemental analysis: Calc.:
C
24H16CuN6O6 (%): C, 52.61; H, 2.94; N, 15.34. Found: C, 52.63; H,
3.06; N, 15.41. HR ESI-MS: m/z [1]2+ Calc.: 211.5330, Found:
211.5376; [1 + NO3]+ Calc.: 485.0544, Found: 485.0612; [1 + Cl]+
Calc.: 458.0354, Found: 458.0424.
2.3.2. [Cu(phenD)2](NO3)2 (2(NO3)2)
Cu(NO3)2Á3H2O (0.059 g, 0.244 mmol), phenD (0.103 g, 0.488
mmol) and 15 mL ethanol were used for the synthesis of 2(NO3)2,
2.2.2. Hydrazone-functionalized phenanthrolines
a mint green solid. Yield: 71%. Anal. Elemental analysis: Calc.: C24-
The ligands N’-(6-oxo-1,10-phenanthroline-5(6H)-ylidene)ace-
tohydrazide (phenAH) and -benzohydrazide (phenBH) were pre-
pared according to the following general procedure by
condensing phenD with the appropriate hydrazide [2].
H12CuN6O10 (%): C, 47.42; H, 1.99; N, 13.82. Found: C, 47.58; H,
2.04; N, 13.91. HR ESI-MS: m/z [2]2+ Calc.: 241.5072, Found:
241.5118; [2 + NO3]+ Calc.: 545.0027, Found: 545.0122; [2 + Cl]+
Calc.: 517.9838, Found: 517.9951.
Please cite this article in press as: J. Heinrich et al., Inorg. Chim. Acta (2017), https://doi.org/10.1016/j.ica.2017.11.015

J. Heinrich et al. / Inorganica Chimica Acta xxx (2017) xxx–xxx
3
2.3.3. [Cu(phenAH)2](NO3)2 (3(NO3)2)
procedure as described above was applied in the presence of ascor-
bate (1 mM) and one of the following ROS scavengers: DMSO (400
mM), NaN3 (10 mM), pyruvate (2.5 mM) and superoxide dismutase
(SOD, 625 U/mL). In the case of pyruvate as a ROS scavenger a
MOPS buffer concentration of 100 mM was needed to keep the
pH value constant at 7.4. Combining the ROS scavenger pyruvate
(hydrogen peroxide scavenger) with SOD ensured immediate
degradation of enzymatically produced hydrogen peroxide. Addi-
tionally, one sample included all scavengers described above.
Furthermore, to investigate the self-activating DNA cleavage of
complex 3 (37.5 mM CuCl2/78.5 mM ligand) the same experiment
described above was carried out with all ROS scavengers in the
absence of ascorbate as reducing agent.
Cu(NO3)2Á3H2O (0.068 g, 0.282 mmol), phenAH (0.150 g, 0.563
mmol) and 20 mL ethanol were used for the synthesis of 3(NO3)2,
a brown solid. Yield: 15%. Anal. Elemental analysis: Calc.: C28H20
-
CuN10O10∙2.5 H2O∙0.5 C2H5OH (%): C, 44.19; H, 3.58; N, 17.77.
Found: C, 43.69; H, 3.07; N, 17.29. Note: Although the agreement
of calculated and found values is slightly outside the commonly
accepted range of 0.4%, we observed the same DNA cleavage activ-
ity for this complex as for the in situ prepared compound (vide
infra). HR ESI-MS: m/z [3]2+ Calc.: 297.5446, Found: 297.5493;
[3 + NO3]+ Calc.: 657.0776, Found: 657.0872; [3 + Cl]+ Calc.:
630.0587, Found: 630.0672.
2.3.4. [Cu(phenBH)2](NO3)2 (4(NO3)2)
Cu(NO3)2Á3H2O (0.055 g, 0.228 mmol), phenBH (0.150 g, 0.457
mmol) and 20 mL ethanol were used for the synthesis of 4(NO3)2,
a lime green solid. Yield: 50%. Anal. Elemental analysis: Calc.:
2.5.3. Bis-nitrophenyl phosphate (BNPP) assay
For the investigation of hydrolytic DNA cleavage activity a mix-
ture of BNPP (50 mM) buffered in MOPS buffer (50 mM, pH 7.4) was
incubated exemplarily with complex 3 (50 mM CuCl2/105 mM
ligand or 100 mM CuCl2/210 mM ligand) for 2 h at 37 °C. As a posi-
tive control the enzymatic BNPP cleavage by phosphodiesterase I
(0.05 U/mL) was carried out. The volume of all samples was
adjusted to 1 mL, all samples contained 12.5% DMSO. Following
the incubation, the degree of BNPP cleavage was monitored
UV/Vis spectroscopically in the range of 200–550 nm for each sam-
ple with the respective baseline (the baseline included all com-
pounds except BNPP).
C
38H24CuN10O10 (%): C, 54.06; H, 2.87; N, 16.59. Found: C, 54.04;
H, 3.17; N, 16.57. HR ESI-MS: m/z [4]2+ Calc.: 359.5603, Found:
359.5660; [4 + NO3]+ Calc.: 781.1089, Found: 781.1212; [4 + Cl]+
Calc.: 754.0900, Found: 754.1039.
2.4. Cyclic voltammetry
Cyclic voltammograms were recorded with a PAR VersaStat 4
potentiostat (Ametek) by working in anhydrous and degassed
dimethylformamide with 0.1 M NBu4PF6 (dried, >99.0%, electro-
chemical grade, Fluka) as supporting electrolyte. Concentrations
of the Cu(II) complexes (1À4) were about 10À4 M. A three-elec-
trode setup was used with a glassy carbon working electrode, a
coiled platinum wire as counter electrode, and a coiled silver wire
as a pseudoreference electrode. The ferrocene/ferrocenium redox
couple was used as internal reference.
2.6. Circular dichroism (CD) spectroscopy
CD spectra of calf thymus (CT) DNA (100 mM) in MOPS buffer
(50 mM, pH 7.4) were recorded on a Jasco J-810 spectrometer in
a range of 220–320 nm with a measuring velocity of 100 nm/min
and a data point interval of 0.1 nm. Cu(II) complexes 1–4 were
added stepwise (5 mM CuCl2/10.5 mM ligand to 15 mM CuCl2/31.5
mM ligand) to investigate their mode of interaction towards DNA.
The volume of all samples was adjusted to 1 mL, all samples
contained 10% DMSO.
2.5. DNA cleavage studies
Plasmid DNA pBR322 was purchased from Carl Roth. All DNA
cleavage experiments were performed at least in triplicate in order
to ensure reproducibility.
Metal complex solutions (1À4) were prepared in situ as follows:
500 mL of a 0.84 mM ligand solution in DMSO and 500 mL of an
aqueous 0.40 mM CuCl2 solution were combined resulting in 1
mL of a 0.40 mM complex solution in water:DMSO 1:1 (Cu:ligand
ratio 1:2.1).
3. Results and discussion
3.1. Synthesis of the Cu(II) complexes
Proceeding from the commercially available 1,10-phenanthro-
line (phen) the oxidized compound 1,10-phenanthroline-5,6-dione
(phenD) was synthesized according to the literature [23]. N’-(6-
Oxo-1,10-phenanthroline-5(6H)-ylidene)benzohydrazide (phenBH)
was prepared by condensing phenD with benzhydrazide [2]. This
synthetic route for hydrazone-functionalization in position 5 of
phenanthrolines was adjusted to prepare the compound N’-(6-oxo-
1,10-phenanthroline-5(6H)-ylidene)acetohydrazide (phenAH), which
has not been described in the literature before. All ligands (phenD,
phenBH, phenAH) were characterized via elemental analysis, 1H
NMR spectroscopy and high-resolution mass spectrometry
(Section 2.2.).
The synthesis of the novel Cu(II) complexes of phenBH and phe-
nAH was based on a procedure for the synthesis of Cu(II) com-
plexes of phen and phenD [24,25]. All Cu(II) complexes [Cu
(phen)2](NO3)2 (1(NO3)2), [Cu(phenD)2](NO3)2 (2(NO3)2), [Cu
(phenAH)2](NO3)2 (3(NO3)2) and [Cu(phenBH)2](NO3)2 (4(NO3)2)
were characterized via elemental analysis and high resolution
mass spectrometry (Section 2.3.) as well as UV/Vis spectroscopy.
UV/Vis spectra of the isolated complexes (NOÀ3 counter ions) as
well as the in situ prepared complexes (ClÀ counterions) 1–4
revealed that in aqueous solution presumably the same species
were present during DNA cleavage for these complexes prepared
2.5.1. General procedure
For the investigation of oxidative DNA cleavage activity of the
complexes a mixture of plasmid DNA pBR322 (0.025 mg/lL) buf-
fered in MOPS (50 mM, pH 7.4) was incubated for 2 h at 37 °C with
the respective compound (25 mM CuCl2/52.5 mM ligand from the
above described solutions) in the presence of ascorbate (1 mM)
as reducing agent or in its absence. For the latter case double con-
centrations of the complexes were used (50 mM CuCl2/105 mM
ligand). Thereby all samples contained 12.5% DMSO. Gel elec-
trophoresis was carried out for 2 h at 40 V using a 1% agarose gel
containing ethidium bromide (0.2 mg/mL) in 0.5X TBE buffer. The
bands of supercoiled (S), open-circular/nicked (N), and linear (L)
DNA were visualized by fluorescence imaging of intercalating
ethidium bromide on a Bio-Rad GelDoc EZ Imager. Data analysis
was performed with Bio-Rad’s Image Lab Software (Version 3.0).
Due to the decreased affinity of ethidium bromide to supercoiled
DNA a correction factor of 1.22 was used [26].
2.5.2. Reactive oxygen species (ROS)
For the determination of ROS exemplarily generated by com-
plex 3 (12.5 mM CuCl2/26.25 mM ligand) the general DNA cleaving
Please cite this article in press as: J. Heinrich et al., Inorg. Chim. Acta (2017), https://doi.org/10.1016/j.ica.2017.11.015

4
J. Heinrich et al. / Inorganica Chimica Acta xxx (2017) xxx–xxx
by different means (cf. S11). Whereas 1 and 2 show characteristic
d-d transition bands at ꢀ700 nm (see inset in S11) [25], 3 and 4
exhibit broad, intense MLCT bands (400–600 nm) instead. The lat-
ter indicates formation of Cu(I) species in aqueous solution [25]
(vide infra).
The composition and structures of the complexes are shown in
Chart 1. To the best of our knowledge, 3 and 4 represent the first Cu
(II) complexes of phenanthroline with hydrazone-functionalization
in position 5.
3.2. DNA cleavage studies
3.2.1. Nuclease activity
In order to investigate the oxidative DNA cleavage activity of
[Cu(phen)2]2+ (1), [Cu(phenD)2]2+ (2), [Cu(phenAH)2]2+ (3) and
[Cu(phenBH)2]2+ (4), their ability to convert supercoiled (S) plas-
mid DNA into open-circular/nicked (N) and linear (L) DNA forms
was monitored by agarose gel electrophoresis. Plasmid DNA
pBR322 was incubated either with the complexes at 25 mM con-
centration in the presence of the reducing agent ascorbate
(1 mM) or with the respective complexes at 50 mM concentration
in the absence of ascorbate for 2 h at 37 °C. The pH value was kept
constant at 7.4 using MOPS buffer (50 mM). We have chosen this
buffer instead of the more frequently used Tris buffer in order to
avoid potential competitive coordination of the buffer molecules
to Cu(II) [27]. Furthermore, due to the insufficient solubility of
the ligands in common solvents and in order to prevent precipita-
tion of the ligands during incubation, an aqueous CuCl2 solution
and a DMSO solution of the ligands were combined for in situ for-
mation of complexes 1–4 (Cu(II):ligand ratio of 1:2.1). An excess of
ligand was used to avoid free Cu(II). The nuclease activity of com-
plexes 1–4 in the presence and absence of ascorbate is shown in
Fig. 1.
Fig. 1. (Top) Cleavage of plasmid DNA pBR322 (0.2 mg) by CuCl2 and complexes 1, 2,
3 and 4 in MOPS buffer (50 mM, pH 7.4) after incubation for 2 h at 37 °C in the
absence or presence of ascorbate (asc, 1 mM). Lane 1: DNA ladder (S/N/L), lane 2:
DNA reference + asc, lane 3: CuCl2 (25 mM) + asc, lane 4: CuCl2 (50 mM), lane 5: 1
(25 mM) + asc, lane 6: 1 (50 mM), lane 7: 3 (25 mM) + asc, lane 8: 3 (50 mM), lane 9: 4
(25 mM) + asc, lane 10: 4 (50 mM), lane 11: 2 (25 mM) + asc, lane 12: 2 (50 mM). All
samples contained 12.5% DMSO. (Bottom) Visualization of the cleaved DNA in
percentage of total DNA. Error bars represent the standard deviation from at least
three experiments.
In general, all complexes, and also CuCl2, show better DNA
cleavage activity in the presence of ascorbate (in comparison to
its absence), which indicates that these Cu(II) species are activated
by an external reducing agent and cleave the DNA in an oxidative
pathway. Complex 1 is the best DNA cleaving agent as expected
from the literature [6–8], for which only DNA fragments were
detected under the conditions described above. The nuclease activ-
ity decreases in the order 1 > 3 > 4 > 2 (Fig. 1, lanes 5, 7, 9 and 11)
in the presence of ascorbate. Complexes 3 and 4 are able to cleave
the DNA into its linear form and fragments as well, but to a lesser
extent. For 2 mostly supercoiled and nicked DNA were observed.
Thus, the conversion of phen into a 5,6-dione (phenD) decreases
dramatically the nuclease activity of the corresponding complex.
However, an additional functionalization in position 5 to a hydra-
zone (3 and 4) increases the DNA cleavage activity again. As shown
below (Sections 3.3 and 3.4), the dione derivatization of phen influ-
ences the DNA binding and the redox behavior of the Cu(II)/Cu(I)
cycle for generation of ROS in a negative fashion. The hydrazone
functionalization reverses this effect with 3 being as active as 1
and 4 achieving an activity close to 1 and 3.
Surprisingly, in the absence of ascorbate the Cu(II) complexes
2–4 also exhibit DNA cleavage activity. This activity is remarkably
high for the hydrazone-functionalized complexes 3 and 4. Overall,
the nuclease activity of complexes decrease in the order 3 > 4 > 2 > 1
(Fig. 1, lanes 8, 10, 12 and 6) in the absence of any external
R =
n/a
complex
1
2
3
4
Chart 1. Proposed structures of the Cu(II) complexes [Cu(phen)2]2+ (1), [Cu(phenD)2]2+ (2), [Cu(phenAH)2]2+ (3) and [Cu(phenBH)2]2+ (4) with unsubstituted phen, oxidized
phenD and hydrazone-functionalized phenAH and phenBH. Anions and solvent molecules are omitted for clarity. For 3 and 4, only one of two possible isomers generated by
asymmetric substitution at positions 5 and 6 of the phenanthroline moiety is shown.
Please cite this article in press as: J. Heinrich et al., Inorg. Chim. Acta (2017), https://doi.org/10.1016/j.ica.2017.11.015

J. Heinrich et al. / Inorganica Chimica Acta xxx (2017) xxx–xxx
5
Table 1
reducing agent. Thereby, a functionalization (dione or hydrazone)
in positions 5 and 6 of the phenanthroline scaffold can have a neg-
ative influence in the presence of ascorbate, but has a positive one
in the absence of ascorbate. In both cases the complexes with
Scavenger
ROS
DMSO
NaN3
Hydroxyl radicals (ÅOH) [29]
Singlet oxygen (1O2) [30]
hydrazone-functionalized
phenanthrolines
(acetyl > benzoyl)
Pyruvate
Superoxide dismutase (SOD)
Hydrogen peroxide (H2O2) [31]
Superoxide anion radicals (OÅ2À) [32]
show an increased nuclease activity in comparison to the complex
with the oxidized form of phenanthroline (phenD).
In order to make sure that the counter ion has no influence on
the nuclease activity, we also compared the isolated complexes
(nitrate counterions) with in situ prepared complexes (chloride
counterions) (cf. S12) Indeed, there was no difference in the
observed cleavage activity. Also for the isolated complex 3, for
which the suggested composition 3(NO3)2∙(H2O)2.5∙(EtOH)0.5 does
not perfectly fit with the proposed structure (ꢀ0.5% deviation),
DNA cleavage experiments comparing it to the in situ generated
complex gave the same results (cf. S12).
3.2.2. Reactive oxygen species (ROS)
To test for an oxidative DNA cleavage mechanism in both the
presence and absence of ascorbate, a quench assay for ROS was car-
ried out. In general, Cu(II) metallonucleases with different N-donor
ligands can cleave the DNA via an oxidative cleavage mechanism
[28]. Therefore activation with an external reducing agent is gener-
ally needed to initiate the generation of ROS through the Cu(II)/Cu
(I) redox cycle. A possible pathway for the occurrence of several
ROS starting with Cu(II) and ascorbate as a reducing agent is shown
in Scheme 1. Such ROS react with DNA by hydrogen abstraction
resulting in oxidized ribose and nucleobase moieties and eventu-
ally in an irreversible DNA cleavage [12,13].
Fig. 2. (Top) Cleavage of plasmid DNA pBR322 (0.2 mg) by complex 3 (12.5 mM) in
MOPS buffer (50 mM, pH 7.4) with ascorbate (1 mM). Incubation for 2 h at 37 °C
in the absence and presence of corresponding ROS scavengers. Lane 1: DNA ladder
(S/N/L), lane 2: DNA reference, lane 3: 3, lane 4: 3 + DMSO (400 mM), lane 5:
3 + NaN3 (10 mM), lane 6: 3 + pyruvate (2.5 mM), lane 7: 3 + SOD (625 U/mL), lane 8:
3 + pyruvate (2.5 mM) + SOD (625 U/mL), lane 9: 3 + all scavengers. All samples
contained 12.5% DMSO. (Bottom) Visualization of the cleaved DNA in percentage of
total DNA. Error bars represent the standard deviation from at least three
experiments.
3.2.2.1. Presence of a reducing agent. In order to confirm ROS forma-
tion as a cleavage mechanism, quenching experiments were car-
ried out. Representatively, complex 3 (12.5 mM) was incubated
with plasmid DNA pBR322 for 2 h at 37 °C in the presence of ascor-
bate (1 mM) and ROS scavengers. In Table 1, ROS scavengers are
listed with the corresponding reactive species that is quenched
[29–32]. In Fig. 2 the nuclease activity of 3 in the absence and pres-
ence of the corresponding ROS quenchers is shown.
Compared to the nuclease activity of 3 without any ROS scav-
enger (lane 3) the activity is decreased in the presence of pyruvate
(lane 6) and SOD (lane 7) indicating the presence of hydrogen
peroxide (H2O2) and superoxide anion radicals (OÅ2À). Thus, an
oxidative DNA cleavage pathway by activating Cu(II) with an
external reducing agent was proven. The reaction of superoxide
anion radicals with SOD (lane 7) leads to hydrogen peroxide [33],
which is also a ROS. Not surprisingly, combining SOD with
pyruvate (lane 8) results in a more pronounced quenching effect.
Additionally, combining all scavengers (lane 9) gives comparable
results to the sample with pyruvate and SOD. Consequently,
hydroxyl radicals (ÅOH) and singlet oxygen (1O2) can be precluded
as ROS in this specific reaction.
be excluded that hydroxyl radicals are formed and are involved in
the DNA cleavage.
3.2.2.2. Absence of a reducing agent. DNA cleavage activity in the
absence of any external activating agent can be caused either by
a hydrolytic cleaving pathway or by self-activation through gener-
ation of ROS. In order to investigate a potential self-activating
property of the artificial metallonucleases, exemplarily, complex
3 was incubated with plasmid DNA pBR322 in the absence of
ascorbate for 2 h at 37 °C with the ROS scavengers listed above
(Fig. 3).
The quenching experiments for the artificial metallonuclease 3
gave similar results in the presence and absence of ascorbate as an
external reducing agent. In both quench assays, hydrogen peroxide
(H2O2) and superoxide anion radicals (OÅ2À) were the dominant spe-
cies, even more so in the absence of ascorbate. Thus, the Cu(II)
complex 3 initiates an oxidative DNA cleavage pathway also in
the absence of a reducing agent. It can be assumed that also the
other complexes (2 and 4) have such self-activating properties as
already indicated from the cleavage activity in the absence of
ascorbate (Fig. 1).
It has to be mentioned that all samples contained 12.5% DMSO
(corresponds to 1.75 M), a supplement also used in the quenching
assay due to its hydroxyl radical scavenging activity, however, usu-
ally at lower concentrations (in this case 400 mM). Thus, it cannot
The mechanism of self-activating metallonucleases is rarely
investigated in the literature. For instance, a hydroxy-salen Cu(II)
complex was converted into a Cu(III) species in the presence of
oxygen, which caused the generation of superoxide anion radicals
as ROS [20]. Cu(III) could also be generated in situ by reaction with
Scheme 1. Oxygen reduction by copper redox cycling in the presence of ascorbate
and generation of ROS [14,15].
Please cite this article in press as: J. Heinrich et al., Inorg. Chim. Acta (2017), https://doi.org/10.1016/j.ica.2017.11.015

6
J. Heinrich et al. / Inorganica Chimica Acta xxx (2017) xxx–xxx
Fig. 4. UV/Vis spectra of BNPP (50 mM) after incubation with either 3 (50 mM or 100
mM) or phosphodiesterase I (PDE, 0.05 U/mL). The baseline included all compounds
with their respective concentrations except for BNPP. All samples contained 25%
DMSO.
Fig. 3. (Top) Cleavage of plasmid DNA pBR322 (0.2 mg) by complex 3 (37.5 mM) in
MOPS buffer (50 mM, pH 7.4) without ascorbate. Incubation for 2 h at 37 °C in the
absence and presence of corresponding ROS scavengers. Lane 1: DNA ladder (S/N/L),
lane 2: DNA reference, lane 3: 3, lane 4: 3 + DMSO (400 mM), lane 5: 3 + NaN3 (10
mM), lane 6: 3 + pyruvate (2.5 mM), lane 7: 3 + SOD (625 U/mL), lane 8: 3 +
pyruvate (2.5 mM) + SOD (625 U/mL), lane 9: 3 + all scavengers. All samples
contained 12.5% DMSO. (Bottom) Visualization of the cleaved DNA in percentage
of total DNA. Error bars represent the standard deviation from at least three
experiments.
of BNPP at 290 nm and of the cleaved product p-nitrophenol at
400 nm (Fig. 4).
Due to the fact that after incubation of BNPP with 3 (50 mM or
100 mM) no absorption band in the range of 400 nm was observed,
the hydrolytic cleavage mechanism was excluded. Thus, the self-
activating character was corroborated, implying an oxidative
cleavage mechanism.
a redox non-innocent ligand system like the phenD used here. It
has been shown in the literature that phenD is indeed reducible
and leads to semiquinone and catecholate species (Scheme 2)
[34–36], suggesting such an activation for complex 2.
Table 2
Active Cu(II) species
E1/2 [V] of Cu(II)/Cu(I) reduction
In case of phenAH (complex 3) and phenBH (complex 4), how-
ever, UV/Vis spectra (cf. S11) indicate formation of Cu(I) species in
aqueous solution on the basis of broad and intense MLCT bands at
ꢀ430 nm and ꢀ550 nm [25]. During incubation with DNA such Cu
(I) species can then initiate the generation of ROS as described
above and thus lead to damage of this biomolecule (see 3.2.2 and
Scheme 1). In the literature, such a self-activation mechanism
has been described for Cu(II) complexes with hydrazone or tripyr-
rolic ligands [17–19]. These oxidizable ligands act as reductants for
the generation of Cu(I) species. As suggested by a literature report,
oxidation in case of phenAH and phenBH could lead to hydrazonyl
radicals [37].
1
2
3
4
À0.35
À0.92
À0.99
À0.94
3.2.3. Bis(4-nitrophenyl) phosphate (BNPP) assay
In order to disprove a hydrolytic DNA cleavage mechanism for
complexes 2–4, exemplarily Cu(II) complex 3 (50 and 100 mM)
was incubated with bis(4-nitrophenyl) phosphate (BNPP, 50 lM)
in MOPS buffer (50 mM, pH 7.4) for 2 h at 37 °C. As a positive con-
trol for the cleavage of BNPP, a model for the phosphate backbone
of DNA, the enzymatic cleavage by phosphodiesterase I (0.05 U/
mL) was used. The hydrolytic cleavage was monitored by UV/Vis
spectroscopy by means of the characteristic absorption maximum
Fig. 5. Cyclic voltammograms for the reduction processes of 1–4 in DMF with 0.1 M
NBu4PF6 as supporting electrolyte (first cycle: strong line; second cycle: pale line).
The supposed Cu(II)/Cu(I) reduction is marked with an asterisk. The ferrocene/fer-
rocenium redox couple was used as an internal reference. The scan rate was 100
mV/s and the current I was normalized.
Scheme 2. Redox equilibrium for o-quinone.
Please cite this article in press as: J. Heinrich et al., Inorg. Chim. Acta (2017), https://doi.org/10.1016/j.ica.2017.11.015

J. Heinrich et al. / Inorganica Chimica Acta xxx (2017) xxx–xxx
7
3.3. Cyclic voltammetry
the unsubstituted phen. Thus, the functionalization in 5/6 position
shifts the Cu(II)/Cu(I) reduction to more negative potentials.
Bringing the redox activity of the Cu(II)/Cu(I) reduction of 1–4
in accordance with their nuclease activity, 1 should initiate ROS
generation to a larger extent in comparison to 2, 3 and 4 due to
its favorable redox potential. Not surprisingly, the Cu(II) complex
1 has the highest nuclease activity. In contrast, the differences in
nuclease activity of the functionalized complexes 2, 3 and 4 can
rather be defined by their distinguished mode of binding interac-
tion towards DNA.
To explain the differing nuclease activity of Cu(II) complexes 1–
4, the redox behavior (by means of cyclic voltammetry) and the
extent of DNA interaction can be considered.
With this in mind, the reduction processes of the complexes 1–
4 were examined by cyclic voltammetric measurements to obtain
the half-wave potential E1/2 of the Cu(II)/Cu(I) redox couple and
therefore an insight into the connection between nuclease activity
and redox activity of the corresponding Cu(II) complexes. Fig. 5
shows the cyclic voltammograms of 1–4 by using the ferrocene/-
ferrocenium redox couple as an internal reference. The cyclic
voltammograms yield the E1/2 of the reversible Cu(II)/Cu(I) reduc-
tion against FcH/FcH+ (Table 2).
The redox potential for the Cu(II)/Cu(I) reduction decreases in
the order 1 >> 2 > 4>3. Furthermore, the Cu(II) complex 1 with
unsubstituted phen has by far the least negative redox potential
for Cu(II)/Cu(I) reduction, whereas dione- (2) and hydrazone-
derivatization (3 and 4) on the phen scaffold in positions 5 and 6
gives more negative redox potentials, but all in the same range.
This can be explained by the initial and favorable reduction of
the ligand phenD to a catecholate via a semiquinone for 2 and also
of phenAH and phenBH to a semiquinone for 3 and 4 (Fig. 5,
Scheme 2) [34–36]. Such a reduction is not possible for 1 with
3.4. CD spectroscopy
Circular dichroism spectroscopy is a commonly used optical
technique to investigate the interaction of metal complexes
towards DNA due to distinction of different modes of DNA binding
such as intercalation and groove binding. In general, CD spectra of
CT-DNA exhibit a positive band at 275 nm caused by base-stacking
of the nucleobases and a negative band at 245 nm caused by the
helicity of the B-form of DNA [38]. A change (increase or decrease)
of the positive band corresponds to intercalation of metal com-
plexes and a decrease of the negative band can suggest groove
binding interaction [39–41]. The CD spectra of CT-DNA with an
increasing amount of the Cu(II) complexes 1–4 are shown in Fig. 6.
Fig. 6. CD spectra of CT-DNA (100 mM) in MOPS buffer (50 mM, pH 7.4) with an increasing amount of Cu(II) complexes 1–4 (0–15 mM) leading to an enhanced alteration of the
positive band at 275 nm (base stacking) and the negative band at 245 nm (helicity of B-DNA). All samples contained 10% DMSO.
Please cite this article in press as: J. Heinrich et al., Inorg. Chim. Acta (2017), https://doi.org/10.1016/j.ica.2017.11.015

8
J. Heinrich et al. / Inorganica Chimica Acta xxx (2017) xxx–xxx
As expected complex 1 with the unsubstituted phen and
Appendix A. Supplementary data
thereby planar ligand scaffold caused the strongest change of the
positive band at 275 nm due to intercalation in comparison to 2–
4. Whereas complex 1 increases the positive band, complexes 2–
Supplementary data associated with this article can be found, in
the online version, at https://doi.org/10.1016/j.ica.2017.11.015.
4
with functionalized ligands phenD, phenAH and phenBH
decrease the positive band, but both indicate intercalation as the
DNA binding interaction. Overall, the extent of DNA intercalation
decreases in the order 1 >> 4 > 2>3 as derived from their CD spectra
with CT-DNA. Thus, the nuclease activity of the best DNA cleaving
agent 1 in the presence of ascorbate correlates with the DNA bind-
ing by means of intercalation and with the redox potential of the
Cu(II)/Cu(I) reduction.
References
[1] P.G. Sammes, G. Yahioglu, Chem. Soc. Rev. 23 (1994) 327.
[2] P. Barbazán, A. Hagenbach, E. Oehlke, U. Abram, R. Carballo, S. Rodríguez-
Hermida, E.M. Vázquez-López, Eur. J. Inorg. Chem. (2010) 4622.
[3] M. Carcelli, G. Corazzari, S. Ianelli, G. Pelizzi, C. Solinas, Inorg. Chim. Acta 353
(2003) 310.
[4] M. Carcelli, S. Ianelli, P. Pelagatti, G. Pelizzi, D. Rogolino, C. Solinas, M. Tegoni,
Inorg. Chim. Acta 358 (2005) 903.
In consideration of the negative band at 245 nm due to the
helicity of CT-DNA, the strongest change is caused by complex 3.
In general, the order changes to 3 > 2 > 4 > 1 by decreasing the neg-
ative band through an assumed groove binding. The complexes 2–
4 have nearly the same redox potentials for the Cu(II)/Cu(I) reduc-
tion (cf. Table 2), but the enhanced nuclease activity of 3 can be
explained by an assumed stronger groove binding in comparison
to 2 and 4. Additionally, a hydrazone functionalization with an
acetyl moiety (complex 3) at Cu(II) complex 1 seems to be better
for influencing the helicity of CT-DNA than a benzoyl moiety (com-
plex 4), which again corresponds to the respective DNA cleavage
activity.
[5] W.H. Hegazy, M.H. Marzouk, Chem. Sci. Trans. 2 (2013) 1482.
[6] D.S. Sigman, D.R. Graham, V. D’Aurora, A.M. Stern, J. Biol. Chem. 254 (1979)
12269.
[7] D.S. Sigman, C.B. Chen, Annu. Rev. Biochem. 59 (1990) 207.
[8] D.S. Sigman, A. Mazumder, D.M. Perrin, Chem. Rev. 93 (1993) 2295.
[9] G. Verma, A. Marella, M. Shaquiquzzaman, M. Akhtar, M.R. Ali, M.M. Alam, J.
Pharm. Bioallied Sci. 6 (2014) 69.
[10] K. Suntharalingam, D.J. Hunt, A.A. Duarte, A.J.P. White, D.J. Mann, R. Vilar,
Chem. Eur. J. 18 (2012) 15133.
[11] E.L. Hegg, J.N. Burstyn, Coord. Chem. Rev. 173 (1998) 133.
[12] W. Knapp Pogozelski, T.D. Tullius, Chem. Rev. 98 (1998) 1089.
[13] M.M. Meijler, O. Zelenko, D.S. Sigman, J. Am. Chem. Soc. 119 (1997) 1135.
[14] L. Guilloreau, S. Combalbert, A. Sournia-Saquet, H. Mazarguil, P. Faller,
ChemBioChem 8 (2007) 1317.
[15] C. Wende, N. Kulak, Chem. Commun. 51 (2015) 12395.
[16] N.F. König, N. Kulak, Elsevier Reference Module in Chemistry, Mol. Sci. Chem.
Eng. (2015) 1.
[17] G. Park, J.T. Tomlinson, M.S. Melvin, M.W. Wright, C.S. Day, R.A. Manderville,
Org. Lett. 5 (2003) 113.
4. Conclusion
[18] M.S. Melvin, J.T. Tomlinson, G.R. Saluta, G.L. Kucera, N. Lindquist, R.A.
Manderville, J. Am. Chem. Soc. 122 (2000) 6333.
[19] S. Rodríguez-Hermida, C. Wende, A.B. Lago, R. Carballo, N. Kulak, E.M.
Vázquez-López, Eur. J. Inorg. Chem. (2013) 5843.
[20] E. Lamour, S. Routier, J.-L. Bernier, J.-P. Catteau, C. Bailly, H. Vezin, J. Am. Chem.
Soc. 121 (1999) 1862.
[21] A. Kellett, M. O’Connor, M. McCann, M. McNamara, P. Lynch, G. Rosair, V.
McKee, B. Creaven, M. Walsh, S. McClean, A. Foltyn, D. O’Shea, O. Howe, M.
Devereux, Dalton Trans. 40 (2011) 1024.
[22] A. Kellett, M. O’Connor, M. McCann, O. Howe, A. Casey, P. McCarron, K.
Kavanagh, M. McNamara, S. Kennedy, D.D. May, P.S. Skell, D. O’Shea, M.
Devereux, Med. Chem. Commun. 2 (2011) 579.
[23] W. Guo, B.J. Engelman, T.L. Haywood, N.B. Blok, D.S. Beaudoin, S.O. Obare,
Talanta 87 (2011) 276.
[24] Z. Molphy, A. Prisecaru, C. Slator, N. Barron, M. McCann, J. Colleran, D.
Chandran, N. Gathergood, A. Kellett, Inorg. Chem. 53 (2014) 5392.
[25] O. Das, T.K. Paine, Dalton Trans. 41 (2012) 11476.
In conclusion, we synthesized the Cu(II) complexes of phen and
phenD (1 and 2) as well as the Cu(II) complexes of hydrazone-bear-
ing phenAH and phenBH (3 and 4), previously not described in the
literature. All complexes were characterized via elemental analysis,
UV/Vis spectroscopy and high resolution ESI mass spectrometry.
DNA cleavage activity was observed in the presence of ascor-
bate as a reducing agent following the order 1 > 3 > 4 > 2 and in
the absence of ascorbate decreasing in the order 3 > 4 > 2 > 1. Thus,
in both cases the hydrazone functionalization in position 5 (acetyl
> benzoyl) increases the nuclease activity. In general a derivatiza-
tion has a positive influence on DNA cleavage in the absence of
ascorbate.
The self-activating properties of 2–4 in the absence of any
reducing agent were proven by the generation of hydrogen perox-
ide and superoxide anion radicals in an oxidative DNA cleavage
mechanism through an ROS quench assay. Additionally, a hydroly-
tic DNA cleavage mechanism was excluded by means of a BNPP
assay. To the best of our knowledge, we have presented the first
Cu(II) metallonucleases on the basis of hydrazone-functionalized
phenanthrolines. The self-activating properties of the complexes
can be attributed to their redox active ligands and the occurrence
of Cu(III) in case of phenD and Cu(I) in case of phenAH and phenBH.
Furthermore, the nuclease activity of the best DNA
cleaving agent 1 in the presence of ascorbate as reducing agent
correlates with the redox potential of the Cu(II)/Cu(I) reduction
(E1/2 = À0.35 V) and the DNA binding interaction through intercala-
tion. The differences in nuclease activity of 2–4 despite having
nearly the same redox potentials for Cu(II)/Cu(I) reduction
(E1/2 = À0.92 to À0.99 V) can be explained by additional groove
binding, which is strongest for complex 3 with an acetohydrazide
functionalization at the phenanthroline.
[26] P. Hertzberg, P.B. Dervan, Biochemistry 23 (1984) 3934.
[27] J. Nagaj, K. Stokowa-Sołtys, E. Kurowska, T. Fra˛czyk, M. Jezowska-Bojczuk, W.
_
Bal, Inorg. Chem. 52 (2013) 13927.
[28] C. Wende, C. Lüdtke, N. Kulak, Eur. J. Inorg. Chem. (2014) 2597.
[29] M.K. Eberhardt, R. Colina, J. Org. Chem. 53 (1988) 1071.
[30] K. Li, L.-H. Zhou, J. Zhang, S.-Y. Chen, Z.-W. Zhang, J.-J. Zhang, H.-H. Lin, X.-Q.
Yu, Eur. J. Med. Chem. 44 (2009) 1768.
[31] E.A. Mazzio, K.F.A. Soliman, Neurochem. Res. 28 (2003) 733.
[32] J. Hormann, C. Perera, N. Deibel, D. Lentz, B. Sarkar, N. Kulak, Dalton Trans. 42
(2013) 4357.
[33] S.J. Lippard, J.M. Berg, Principles of Bioinorganic Chemistry, first ed., University
Science Books, Mill Valley, California, 1994.
[34] L. Michaelis, M.P. Schubert, R.K. Reber, J.A. Kuck, S. Granick, J. Am. Chem. Soc.
60 (1938) 1678.
[35] L. Michaelis, M.P. Schubert, Chem. Rev. 22 (1938) 437.
[36] D.M. Murphy, K. McNamara, P. Richardson, V. Sanchez-Romaguera, R.E.P.
Winpenny, L.J. Yellowlees, Inorg. Chim. Acta 374 (2011) 435.
[37] M. Harej, D. Dolenc, J. Org. Chem. 72 (2007) 7214.
[38] V.I. Ivanov, L.E. Minchenkova, A.K. Schyolkina, A.I. Poletayev, Biopolymers 12
(1973) 89.
[39] N. Shahabadi, S. Kashanian, A. Fatahi, Bioinorg. Chem. Appl. (2011) 687571.
[40] P. Uma Maheswari, M. Palaniandavar, J. Inorg. Biochem. 98 (2004) 219.
[41] C. Tong, G. Xiang, Y. Bai, J. Agric. Food Chem. 58 (2010) 5257.
Acknowledgements
We thank Christian Wende for initiating this project, Biprajit
Sarkar for useful discussions and the Kulak group for proofreading
of this article.
Please cite this article in press as: J. Heinrich et al., Inorg. Chim. Acta (2017), https://doi.org/10.1016/j.ica.2017.11.015