Copper complexes as a source of redox active MRI contrast agents

Article abstract of DOI:10.1007/s10534-015-9875-3

The study reports an advance in designing copper-based redox sensing MRI contrast agents. Although the data demonstrate that copper(II) complexes are not able to compete with lanthanoids species in terms of contrast, the redox-dependent switch between dia

Full text of DOI:10.1007/s10534-015-9875-3

Biometals
DOI 10.1007/s10534-015-9875-3
Copper complexes as a source of redox active MRI
contrast agents
.
Lynsey Dunbar Rebecca J. Sowden Katherine D. Trotter
.
.
.
Michelle K. Taylor David Smith Alan R. Kennedy
.
.
.
John Reglinski Corinne M. Spickett
Received: 15 July 2015 / Accepted: 27 July 2015
Ó Springer Science+Business Media New York 2015
Abstract The study reports an advance in designing
copper-based redox sensing MRI contrast agents.
Although the data demonstrate that copper(II) com-
plexes are not able to compete with lanthanoids
species in terms of contrast, the redox-dependent
switch between diamagnetic copper(I) and paramag-
netic copper(II) yields a novel redox-sensitive contrast
moiety with potential for reversibility.
Keywords Copper macrocycles Á Oxidation–
reduction Á Redox sensitive contrast agent Á Imaging Á
Inflammation
Introduction
MRI contrast agents capable of detecting redox status
in biological tissues would be valuable in the diagnosis
of inflammatory diseases that involve an oxidative
pathology, such as atherosclerosis or rheumatoid
arthritis (Filippin et al. 2008; Profumo et al. 2011).
Historically, contrast agents have depended on lan-
thanide chemistry (De Leon-Rodriguez et al. 2009;
Aime et al. 2009) and the development of agents for
MRI continues to focus largely on gadolinium owing
to its key qualities; i.e. it is a quadrupolar nucleus, it
has a large number of unpaired electrons and when
coordinatively unsaturated it displays rapid exchange
kinetics in aqueous solution (Table 1). However, the
synthesis of new, smart imaging agents for measuring
biochemical processes (e.g. pH and pO2) requires that
the contrast agent has added capability (Angelovski
et al. 2008; Iwaki et al. 2012; Viger et al. 2013;
Vologdin et al. 2013) A method of introducing added
capability is to modify the ligand that chelates the
lanthanoid centre (Angelovski et al. 2008; Iwaki et al.
2012; Vologdin et al. 2013). Alternatively, it is
possible to encapsulate a gadolinium species into a
polymer matrix which degrades in response to the
biological environment (Viger et al. 2013).
Electronic supplementary material The online version of
this article (doi:10.1007/s10534-015-9875-3) contains supple-
mentary material, which is available to authorized users.
L. Dunbar Á K. D. Trotter Á M. K. Taylor Á
D. Smith Á A. R. Kennedy Á J. Reglinski (&)
Department of Pure & Applied Chemistry, Strathclyde
University, 295 Cathedral St., Glasgow G1 1XL, UK
e-mail: j.reglinski@strath.ac.uk
R. J. Sowden Á C. M. Spickett (&)
Strathclyde Institute of Pharmacy and Biomedical
Sciences, Strathclyde University, 27 Taylor Street,
Glasgow G4 0NR, UK
Present Address:
M. K. Taylor
School of Science and Technology, University of New
England, Armidale NSW 2351, Australia
Present Address:
C. M. Spickett
School of Life and Health Sciences, Aston University,
Aston Triangle, Birmingham B4 7ET, UK
123

Biometals
Table 1 The properties of gadolinium and certain first row transition metal species key to their ability to perform as contrast agents
(Tobe 1972; Weast 1980)
No. unpaired electrons and
relevant oxidation state
Nuclear spin (abundance %)
Exchange rate (s-1
)
Gd
7
3/2 155Gd (14.7)
3/2 157Gd (15.7)
7/2 51V (99.8)
3/2 53Cr (9.6)
109
V
2 (III)
103
109
Cr
4 (II)
3 (III)
5 9 10-5
5 9 107
5 9 106
5 9 10-5
109
Mn
Co
7 (II)
5/2 55Mn (100)
7/2 59Co (100)
3 (II) high spin Oh
0 (III) low spin Oh
1 (II)
Cu
3/2 63Cu (69.1)
3/2 65Cu (30.9)
0 (I)
Unlike the lanthanoids, the exchange kinetics and number of unpaired electrons vary more widely and are governed by the prevailing
oxidation state and ligand field considerations. Oh denotes an octahedral geometry
Recent concerns about the nephrotoxicity of
gadolinium suggest that there would be advantages
in broadening the range of metals used as contrast
agents (Badero et al. 2008; Ledneva et al. 2009). The
clinical use of transition metals in MRI has been
mostly restricted to nanoparticles which, as large
multimetallic assemblies, are able to compensate for
the weak relaxivity of the parent elements (e.g. iron,
copper) (Di Marco et al. 2007; Longmire et al. 2008;
Kueny-Stotz et al. 2012; Min et al. 2012). One strategy
is to design modified low T2 nano-particles which
assemble in the presence of key biomarkers (e.g.
streptavidin–biotin) to form assemblies with a high T2
(Min et al. 2012).
Loving et al. 2013; Gale et al. 2014). Indeed, there has
been some success with manganese in the design of
redox sensing contrast agents (Loving et al. 2013;
Gale et al. 2014; Tsitovich et al. 2014). Copper(II) also
has some attractive properties (Table 1) (Tobe 1972;
Weast 1980). Although it only has one unpaired
electron, its ligand exchange rate is higher by two
orders of magnitude than manganese and commensu-
rate with that of gadolinium (Table 1). Furthermore,
its chemistry suggests that it should be easier to create
an available coordination site for water without the
risk of forming insoluble oxides (c.f. MnO2, Mn2O3).
Another advantage of copper over manganese is the
availability of a diamagnetic form, d10 Cu(I), which
can be sustained under biological conditions. The
availability of diamagnetic (Cu(I)) and paramagnetic
(Cu(II)) forms allows the design of agents that have
redox switch capability. Lauffer briefly discussed the
properties of copper as a platform for the design of
imaging agents 25 years ago, deeming it unsuitable
for use due to its poor relaxivity (Lauffer 1987).
However, with the advent of better imaging software,
access to higher field strengths, more sophisticated
pulse sequences and the emergence of chemical
exchange saturation transfer (CEST) as a relaxation
mechanism, the need for metals with such high
relaxivity may be coming to an end (De Leon-
Rodriguez et al. 2009; Aime et al. 2009; Hancu et al.
2010; Sherry and Wu 2013). The design of MRI
contrast agents has developed significantly over the
Certain transition metals (Table 1) have physical
properties compatible with their development as con-
trast agents (Tobe 1972; Weast 1980; Grant et al. 1997;
Thunus and Lejeune 1999; Yam and Lo 1999; Dorazio
and Morrow 2012; Drahos et al. 2012; Yu et al. 2012;
Loving et al. 2013; Gale et al. 2014; Tsitovich et al.
2014). Since transition metals display different coordi-
nation chemistry to gadolinium (lower coordination
numbers, simpler topologies) and can combine with a
wider range of ligand types (e.g. phosphines, thiolates),
it should be easier to target them to physiological
parameters such as pH and redox status.
Due to its large number of unpaired electrons,
manganese(II) has attracted the most attention as an
alternative to gadolinium (Grant et al. 1997; Thunus
and Lejeune 1999; Yam and Lo 1999; Yu et al. 2012;
123

Biometals
Preparation of N-substituted macrocycles
past 25 years and it is not uncommon for discrete
sensing moieties to be attached directly to an MRI
reporting gadolinium species. Thus, should copper
remain an unsuitable metal in its own right, there are
opportunities to bolster its poor relaxivity by combin-
ing a reporting copper complex with a more respon-
sive gadolinium species. Although the relaxivity of
copper will never supersede that of manganese or
gadolinium, its potential as a switchable redox sensing
contrast moiety makes it an interesting target and,
considering all the advances in the field of MRI over
the past 25 years, a topic worth re-visiting.
SprNen (1.04 g, 0.003 mol) was dissolved in acetonitrile
(100 mL) to which sodium bicarbonate (5.07 g,
0.06 mol) was added. The mixture was stirred and
heated to reflux. Halo-alkane (2-bromo-ethanol,
3-bromo-propan-1-ol or bromoacetic acid, 2 mol
equiv.) in acetonitrile (50 mL) was added drop-wise to
therefluxingmixtureovera6 h period. Themixturewas
then refluxed overnight. The solution was left to cool,
beforebeingfiltered. Thesolventwasremovedinvacuo,
resultinginayellowwaxysolid. Theproductswereused
as received. Yields were typically in the order of 70 %.
cis-SprNen(–(CH2)2OH)2: Mass Spec. (ESI) m/z
455.27 [M ? Na]?.
Experimental
cis-SprNen(–(CH2)3OH)2: Mass Spec. (ESI) m/z
461.14 [M ? H]?.
Unless otherwise stated, all chemicals were commer-
cially obtained and used without further purification.
Gd(DTPA) and Cu(BF4)2 were purchased from the
Sigma-Aldrich company UK. SprNen and the unmodi-
fied copper macrocycles ([Cu(II)SenNen]2?, [Cu(II)Sen-
Npr]2?, [Cu(II)SprNen]2? and [Cu(II)SprNpr]2?) were
prepared as previously reported (Taylor et al. 2008;
Trotter et al. 2009, 2010; Sowden et al. 2013). Solid
reflectance spectra (400–900 nm) were recorded on a
Photonics CCD array UV–Vis spectrophotometer.
Electrospray mass spectra were recorded using a
Thermo Finnigan LCQDuo instrument equipped with
anion trap as described previously (Taylor et al. 2008;
Trotter et al. 2009, 2010; Sowden et al. 2013). The
spectrophotometric titrations were carried out on a
Unicam UV 300 spectrometer.
cis-SprNen(–CH2COOH)2: Mass Spec. (ESI) m/z
461.13 [M ? H]?.
Preparation of N-substituted copper(II)
macrocycles
The required N-substituted macrocycle (1.04 g) was
dissolved in methanol (150 mL) and treated with an
equimolar amount of Cu(BF4)2Á6H2O (0.96 g) dis-
solved in methanol (10 mL). An immediate colour
change from pale yellow to dark green was observed.
The resulting solution was refluxed gently overnight.
The solution was cooled and filtered and the solvent
removed in vacuo to yield a dark green solid. The
crude product was re-dissolved in the minimum
amount of methanol, filtered through Celite and
recrystallised by vapour diffusion using diethyl ether
to give a dark green powder.
X-ray crystallography
Crystals were coated in mineral oil and mounted on
glass fibres. Data were collected at 123 K on a Nonius
Kappa CCD diffractometer using graphite monochro-
mated Mo-Ka radiation. The heavy atom positions
were determined by Patterson methods and the
remaining atoms located in the difference electron
density maps. Full matrix least squares refinement was
based on F2 with all non-hydrogen atoms anisotropic.
While the hydrogen atoms were mostly observed in
the difference maps, they were placed in calculated
positions riding on the parent atoms. The structure
solution and refinement used the programs SHELX-97
and the graphical interface WinGX (Sheldrick 1998;
Farrugia 1999).
cis-[Cu(II)SprNen(-(CH2)2OH)2] 2BF4
Anal. Calcd. for C23H32 N2S2O2CuB2F8: C, 39.30; H,
4.41; N, 4.36 % Found: C, 39.98; H, 4.72; N, 4.57 %.
Mass Spec (ESI) m/z 495.08 [M]?, 451.13 [M-
CH2CH2O]?. kmax (solid reflectance) 435, 640 nm.
cis-[Cu(II)SprNen(-(CH2)3OH)2] 2BF4
Anal. Calcd. for C25H36 N2S2O2CuB2F8: C, 43.04; H,
5.16; N, 4.02 % Found: C, 43.83; H, 5.46; N, 3.98 %.
Mass Spec (ESI) m/z 523.20 [M]?, kmax (solid
reflectance) 435 nm, 600 nm. XRD data: Monoclinic
123

Biometals
˚
C2/c, a = 35.7398(14) A, b = 9.1810(3) A,
˚
3
Anal. Calcd. for C23H18N4O4S2CuB2F8: C, 38.60; H,
2.53; N, 7.83 % Found: C, 40.98; H, 2.60; N, 6.59 %
Mass Spec. (ESI) m/z 541 [M]?
c = 21.0854(7) A, b = 92.385(3)o,
6912.7(4), Z = 8, robs = 0.0438, wR2 = 0.1104.
V
˚
˚
(A )
The Schiff base macrocyclic complex (10 mmol) was
dissolved in ethanol (100 ml) and an excess of NaBH4
(1 g) was added portion wise to the solution and stirred
overnight. The excess NaBH4 was quenched with
ammonium acetate. The solvent was then removed in
vacuo to yield a green oil. This was dissolved in
distilled water and extracted with dichloromethane
and dried over anhydrous sodium sulphate. The
solvent was removed in vacuo to yield [Cu(II)Spr-
NphenR]2? 2BF4- as a green solid.
cis-[Cu(II)SprNen(-CH2COOH)2] 2BF4
Anal. Calcd. for C23H28 N2S2O4CuB2F8: C, 41.94; H,
4.44; N, 3.86 % Found: C, 42.13; H, 4.42; N, 4.02 %.
Mass Spec (ESI) m/z 523.27 [M]?. kmax (solid
reflectance) 415 nm, 600 nm.
cis-[Cu(II)SprNen(-(CH2)2SPh)2] 2BF4
Anal. Calcd. for C35H40 N2S4CuB2F8: C, 49.22; H,
4.68; N, 3.28 % Found: C, 49.67; H, 4.79; N, 3.11 %.
Mass Spec (ESI) m/z 679 [M]?, 543 [M-C9H12S]?
Anal. Calcd. for C23H20N4O4S2CuB2F8: C, 38.37;
H, 4.42; N, 7.67 % Found: C, 39.03; H, 4.26; N,
7.59 %. Mass Spec. (ESI) m/z 543 [M]?. kmax (solid
reflectance) 425 nm, 610 nm.
kmax (solid reflectance) 425, 610 nm.
Relaxation measurements
Preparation of [Cu(II)SprNphenR]2? 2BF4
-
Relaxation measurements were carried out in triplicate
ona Bruker Avance I NMRspectrometer equipped with
an 11.7 T magnet. The measurements were carried out
using an inversion recovery pulse sequence under full
automation control with a 908 pulse width of 11.5 ls.
All spectra were referenced internally to the residual
proton resonance of the deuterated solvent. The relax-
ation delay was varied from 1-100 s, depending on the
relaxation characteristics of the sample. Data were
transferred for remote data processing using Topspin
(version 2.1, Bruker Biospin, Karlsruhe) on a desktop
PC running under Windows XP. Following Fourier
transformation, all data were baseline corrected. The
T1/T2 sub-routine within Topspin was used to evaluate
the relevant T1 values from the data.
5-nitro-2-chloro-benzaldehyde (1.86 g, 0.01 mol) was
dissolved in iso-propylalcohol to which o-phenylene-
diamine (0.6 g, 0.005 mol) was added. The solution
was stirred at room temperature for an hour and a
yellow precipitate formed. The solution was then
filtered and the Schiff base, as a yellow solid, collected
and dried.
1H NMR (400 MHz, DMSO; dH) 8.99 (s, 2H, –CH
= N-), 8.92 (d, 2H, arom), 8.29 (dd, 4H, arom), 7.87
(d, 2H, arom), 7.17 (d, 2H, arom), Anal. Calcd.
for C20H12Cl2N4O4: C, 54.20; H, 2.73; N, 12.64 %
Found: C, 54.34; H, 2.60; N, 12.83 % Mass Spec.
(ESI) m/z 443 [M ? H]?
Samples of the copper macrocycles, GdDTPA and
Cu(BF4)2 were prepared in 0.1 M NaCl, 100 % D2O in
a 10 mL volumetric flask at concentrations of 10, 2.5,
1, 0.1 and 0.01. A control sample of 0.1 mM NaCl in
100 % D2O was also included.
The Schiff base (10 mmol) was dissolved in
dichloromethane to which Cu(BF4)2 (0.34 g,
10 mmol) was added. The solution was heated to
reflux and then propanedithiol (0.09 g, 5 mmol)
added drop-wise. The solution was subsequently
refluxed for 48 h. The solution was then filtered and
the solvent removed in vacuo to yield a green-black
oil. The crude product was dissolved in distilled
water and extracted with dichloromethane and dried
over anhydrous sodium sulphate. The solvent was
removed to yield the copper Schiff base macrocycle
as a green-black oil.
Electrochemistry
Cyclic voltammetry in aqueous solution was carried
out on a CH Instruments 660A Electrochemical
Workstation with iR compensation, using ultra high
purity water (Elga: 0.22lm filtered, 17 MX/cm) and
acetonitrile (redistilled from calcium hydride). The
123

Biometals
electrodes were glassy carbon, platinum wire and
silver wire as the working, counter and reference
electrodes, respectively. All solutions were degassed
with argon and contained sample concentrations
2 9 10-3 M for acetonitrile, together with the
supporting electrolyte; tBu4NBF4 (0.1 M) respec-
tively. Measurements are referenced against the redox
related series of copper complexes (Taylor et al. 2008;
Trotter et al. 2009, 2010; Sowden et al. 2013). The first
series (Fig. 1) comprised four thioether-secondary
amine macrocycles where the alkyl chains linking the
donor atoms are modified so as to allow the formation
of 14, 15 or 16 membered rings. In all of these
complexes the macrocycle lies in the meridial plane of
the metal. A number of these complexes crystallise
with a solvent molecule (MeCN, H2O) coordinated to
the copper atom in an axial position (Taylor et al.
2008; Trotter et al. 2009. This structural feature
indicates that a relaxation mechanism similar to that
found in Gd(DTPA) should be available with these
copper macrocycles. A second series of macrocycles
contain tertiary amines have been prepared. Here the
pendant arms (X in Fig. 1) include a terminal donor
atom, which is introduced to increased the denticity of
the ligand and the stability of the resulting copper
complex. A third system has been synthesised that
includes electron withdrawing groups (Fig. 1, aryl and
-NO2) in an attempt to support the copper(I) form
better.
couple (E ) of ferrocene/ferrocenium.
‘
Chemical redox cycling
Reduction
A solution of the copper macrocyclic (0.3 mM)
complex and reducing agent (10 mM ascorbic acid)
was prepared in Tris-buffer (5 mM pH 7.4). A 1.0 ml
volume of the solution of the complexes was placed in
a 1.5 mL quartz cuvette (1 cm pathlength) and the
spectrum was recorded (250–800 nm). The solution
was titrated with reducing agent in 5 ll aliquots. The
cuvette was allowed to stand for 5 min at room
temperature to allow the solution to come to equilib-
rium before the spectrum was recorded. The titrations
continued until no further change was observed. Each
experiment was carried out in triplicate.
The secondary amine copper macrocycles (Fig. 1,
left) were prepared as previously described (Sowden
et al. 2013) and their relaxation properties assessed in
saline solution (0.10 M NaCl, 2H2O, 9.4T). The
performance of the species were measured against
Gd(DTPA) and Cu(BF4)2. Gd(DTPA) was included as
a reference material to demonstrate the relative ability
of these copper macrocycles to act as contrast agents.
A simple hydrated copper salt (Cu(BF4)2) was also
included in the study to gauge the maximum effect one
might expect from copper agents and in turn the
influence of the macrocyclic ligand. As expected, the
relaxation measurements (Table 2) show that hydrated
copper(II) ions have poor relaxivity compared to
gadolinium (Lauffer 1987). The hydrated copper(II)
cation has a primary coordination sphere of six waters
and a weakly bound secondary coordination sphere of
up to eight water molecules (Persson 2010). Conse-
quently, it is unsurprising that once the copper is
placed within the macrocycles four of the exchange
sites are occupied and the relaxivity falls.
Re-oxidation
A solution of the copper macrocyclic complex
(0.3 mM) was prepared in Tris-buffer (5 mM, pH 7.4)
and reduced using ascorbic acid from the data above to
generatethecopper(I)complexinsitu.A1.0 mlvolume
of the solution of the required copper complexes was
placed in a 1.5 ml quartz cuvette (1 cm pathlength) and
the spectrum was recorded (250–800 nm) against a
Tris-buffer reference placed in a matched cuvette. A
100 mM solution of sodium hypochlorite was titrated in
5 ll aliquots into the solution. The cuvette was allowed
to stand for 5 min at room temperature to allow the
solution to come to equilibrium before the spectrum was
recorded. The titrations continued until no further
change was observed in the spectra. Each experiment
was carried out in triplicate.
It has been shown previously that bovine serum
albumin (BSA) can efficiently remove the copper from
these secondary amine macrocycles (Fig. 1, top left)
in competitive binding studies (Sowden et al. 2013).
Thus, although the copper macrocycles have the
potential to act as contrast agents, in order to have
Results and discussion
As a potential platform for the design of copper-based
MRI moieties, we have synthesized and studied three
123

Biometals
Fig. 1 The copper macrocyclic frameworks studied (Taylor
et al. 2008; Trotter et al. 2009, 2010; Sowden et al. 2013).
a Secondary amine complexes, where the polymethylene chains
between the nitrogen and sulfur donor atoms are ethylene (en) or
[Cu(II)SenNpr]2?, [Cu(II)SprNen]2? and [Cu(II)SprNpr]2?. b Ter-
tiary amine complexes [Cu(II)SprNen(X)2]2?
, where X =
–(CH2)2OH, –(CH2)3OH, –CH2COOH or –(CH2)2SPh.
c [Cu(II)SprNphenR]2? where additional electron withdrawing
groups (aryl, -NO2) are included in the motif
propylene (pr) giving rise to four compounds: [Cu(II)SenNen]2?
,
Table 2 Relaxivity and electrochemical data: The relaxivity of
GdDTPA, Cu(BF4)2, [Cu(II)SenNen]2? [Cu(II)SprNen]2?
[Cu(II)SenNpr]2? [Cu(II)SprNpr]2? [Cu(II)SprNen(X)2]2?
where X = –(CH2)2OH, –(CH2)3OH, –CH2COOH, –(CH2)2-
SPh and [Cu(II)SprNphenR]2? (n = 3) in 0.1 M NaCl in H2O
at 9.4T
2
;
;
,
;
;
Compound
Relaxivity s-1mM-1
E0 V
Reference compounds
Gd(DTPA)
CuII(BF4)2
4.73
0.67
Simple macrocycles (Fig. 1a)
[Cu(II)SenNpr]2?
[Cu(II)SprNen]2?
[Cu(II)SprNpr]2?
,
0.27 0.03
0.27 0.01
0.26 0.03
0.42 0.02
0.16 0.02
0.11 ? 0.01
0.12 0.02
0.11 0.02
0.65*
0.85 (57 %)
0.85*
Pendant arm macrocycles [Cu(II)SprNen(X)2]2? (Fig. 1b)
X = –(CH2)2OH
X = –(CH2)3OH
X = –CH2COOH
X = –(CH2)2SPh
[Cu(II)SprNphenR]2?
t
1.04 (22 %)
1.23 (21 %
1.44 (24 %)
1.28 (79 %)
1.42 (9 %)
Macrocycle-EWG (Fig. 1c)
Electrochemical data for the copper macrocycles in acetonitrile (0.1 M Bu4NBF4: The values for the complexes are expressed
relative to the normal hydrogen electrode (HNE). * (Taylor et al. 2008)
any real utility the stability of these complexes needs
to be increased. Earlier studies have shown that the
[Cu(II)SprNen]2? motif was marginally more resistant
to de-metalation by BSA and this framework was
adopted for modification (Sowden et al. 2013).
Preliminary studies using a tertiary amine complex
that incorporates ethanol pendant arms (Fig. 1 top
right, X = –(CH2)2OH) indicated that this modifica-
tion prevented BSA from removing the metal from the
complex (Sowden et al. 2013). Mattes et al. have
previously synthesised a range of macrocyclic species
based on [Cu(II)SenNen]2? and [Cu(II)SenNpr]2?) that
have ethanol, acetate and pyridyl groups attached to
the secondary amines (Funkemeier and Mattes 1993;
123

Biometals
Bentfeld et al. 1995; Mattes et al. 2004). In the solid
state, these complexes are observed to be cationic, five
coordinate species in which the pendant donor oxygen
atoms are in the coordination sphere of the copper, lie
cis- to one another and are found in their protonated
form. The two donor oxygen atoms are included in the
coordination geometry at the expense of one of the
ring sulfur atoms. The [Cu(II)SprNen]2? platform has
good potential and we have therefore adapted the work
of Mattes et al. to introduce ethanol, propanol, acetate
and arylthioether pendant arms into this motif (Fig. 1).
Although all four complexes were isolated and
characterised, only [Cu(II)SprNen(–(CH2)3OH)2]2?
provided X-ray quality crystals. In contrast to the
structures reported previously, this complex was
found to adopt a five coordinate square based
pyramidal structure (Fig. 2). Although both alcohol
groups are found above the plane of the ligand and are
protonated, only one of the hydroxyl groups coordi-
nates. The structures of the corresponding ethanol,
acetate thioether adducts remain unclear at this time
but we would predict that they are also five coordinate
in the solid state (Funkemeier and Mattes 1993;
Bentfeld et al. 1995; Mattes et al. 2004).
the potential biochemistry. We have developed a more
direct method of analysis whereby the copper complex
is titrated with BSA (Sowden et al. 2013). Using this
approach, we circumvented the need to calculate each
individual formation constant and directly observed if
the protein can sequester the metal (copper). Using
this approach we can show that copper in the five
coordinate species are stable to metal sequestration by
BSA. Although the introduction of a fifth donor into
the motif has generated a suite of complexes that are
suitably bio-tolerant, the increase in stability requires
that a fifth water exchange site (c.f. Cu(BF4)2) is
occupied and consequently the relaxivity of the
complexes falls (Table 2).
The ability of the compounds to act as redox sensing
contrast agents requires that they have redox reversibil-
ity between the copper(II) and copper(I) forms. The
issue of redox reversibility has been interrogated in two
ways: electrochemically and chemically. Electro-
chemical analysis of the compounds (Table 1) shows
that the addition of the pendant arms has raised the E0
of the complexes significantly. Surprisingly, the addi-
tion of electron withdrawing groups (i.e. [Cu(II)Spr-
NphenR]2?) and soft donor moieties ([Cu(II)SprNen
((–CH2)2SPh)2]2?) into the motif has little impact on
redox potential of the copper centre and the values
obtained for these species (1.2–1.5 V) consistently
come to the higher end of the desired range (Table 2).
The voltammograms of compounds indicate that the
addition of pendant arm donor moieties renders the
complexes irreversible except in one case i.e.
([Cu(II)SprNen((–CH2)2SPh)2]2?). This compound
The fifth donor atom was introduced into the ligand
to increases the stability of the complexes. Typically
the stability of a complex is framed using its formation
constants, which are then used to hypothesise about
Fig. 2 The X-ray crystal structure of [Cu(II)SprNen(-(CH2)3-
OH)2]2?. The counter ions (BF4-) are omitted for clarity. The
copper sits in a distorted square based pyramidal geometry. The
hydrogens on the oxygens were found during refinement. The
thermal ellipsoids are shown at 50 % probability. Selected
Fig. 3 The cyclic voltammogramm of [Cu(II)SprNen((CH2)2-
SPh)2] 2BF4 (2 mM) in acetonitrile/tBuNBF4 (0.1 M) at scan
rates of 10, 40, 80, 120, 160 and 200 mVs-1. The E0 values
(Table 2) were referenced to the normal hydrogen electrode
using a ferrocene standard
o
˚
metrical parameters (d/A, ) Cu–S 2.3450(7), 2.3487(7); Cu–N
2.045(2), 2.052(2); Cu–O 2.1725(18);\S-Cu–N (o), 163.00(6),
93.48(6), 176.82(6), 93.49(6)
123

Biometals
Fig. 4 The chemical reduction of [Cu(II)SprNen((CH2)2SPh)2]2? by ascorbic acid (left) followed by the chemical re-oxidation using
hypochlorite (right). The figure is a composite of the data from two separate titrations
shows good reversibility at a range of scan rates
(Fig. 3). This indicates that an increase in the sulphur
density around the copper centre is having a beneficial
effect and that the additional soft donors are able to
support the formation of the copper(I) species during
redox cycling better.
The study does indicate that it should be possible to
generate a compound which is stable in both the ?1
and ?2 oxidation states under biological conditions
with an appropriate oxidation potential. However, it is
clear that for these compounds to have any use in MRI
they will have to be coupled with a higher relaxivity
metal such as gadolinium. Finally, it would be better if
the sensing agent was administered in its low valent
form and the challenge now is to produce cop-
per(I) complexes of this type that respond to oxidants
rather than copper(II) complexes which respond to
reductants.
In our previous studies we investigated the redox
chemistry of our compounds by subjecting them to
chemical redox cycling using ascorbic acid reduction
followed by hypochlorite oxidation (Taylor et al.
2008; Trotter et al. 2009, 2010; Sowden et al. 2013).
Repeating this approach here, we found behaviour
which paralleled the electrochemical study above.
While all the compounds could be stoichiometrically
reduced, only [Cu(II)SprNen((CH2)2SPh)2]2? could be
efficiently re-oxidised (*85 %) back to the divalent
state (Fig. 4). The ability to re-oxidise [Cu(II)Spr-
Nen((CH2)2SPh)2]2? is a significant improvement over
the previous compounds studied (Table 1) (Taylor
et al. 2008; Trotter et al. 2009, 2010; Sowden et al.
2013). However, whereas reduction remains stoichio-
metric, re-oxidation still required an excess of oxidant
to effect the redox switch.
Notes and references
It should be noted that Kupka et al. (1992) made a brief
comment on the potential use of copper(II) D-peni-
cillamine as a contrast agent in plants. To our
knowledge this work did not develop further.
Acknowledgments L.D., K.D.T., M.K.T. and D.S. would like
to thank Strathclyde University and WestChem for financial
assistance. J.R., C.M.S. and R.J.S. gratefully acknowledge the
support of the BBSRC (BBS/B/01553).
Conclusions
Here we present the first major effort to explore
copper’s ability to be used as a redox-sensitive MRI
moiety. As predicted by Lauffer (1987), the relaxivity
of these copper based species is poor compared with
gadolinium and considerable further development is
needed before copper will be able to perform
independently as a redox sensitive contrast agent.
References
Aime S, Castelli DD, Crich SG, Gianolio E, Terreno E (2009)
Pushing the sensitivity envelope of lanthanide-based magnetic
resonance imaging (MRI) contrast agents for molecular
imaging applications. Acc Chem Res 42:822–831
Angelovski G, Fouskova P, Mamedov I, Canals S, Toth E,
Logothetis NK (2008) Smart magnetic resonance imaging
123

Biometals
agents that sense extracellular calcium fluctuations.
ChemBioChem 9:1729–1734
Badero OJ, Schlanger L, Rizk D (2008) Gadolinium nephro-
toxicity: case report of a rare entity and review of the lit-
erature. Clin Nephr 70:518–522
Bentfeld R, Ehlers N, Mattes R (1995) Synthesis and charac-
terization of copper(II) complexes of a 14-membered cis-
N2S2 dibenzo macrocycle and of its bis-acetato and
bis(methylpyridyl) derivatives. Chem Ber 128:1199–1205
De Leon-Rodriguez LM, Lubag AJM, Malloy CR, Martinez
GV, Gillies RJ (2009) Sherry AD responsive MRI agents
for sensing metabolism in vivo. Acc Chem Res 42:948–957
Di Marco M, Sadun C, Port M, Guilbert I, Couvreur P, Dubernet
C (2007) Physicochemical characterization of ultrasmall
superparamagnetic iron oxide particles (USPIO) for
biomedical application as MRI contrast agents. Int J
Nanomed 2:609–622
Dorazio S, Morrow JR (2012) The Development of Iron(II)
Complexes as ParaCEST MRI Contrast Agents. Eur J Inorg
Chem 12:2006–2014
Drahos B, Lukes I, Toth E (2012) Manganese(II) Complexes as
Potential Contrast Agents for MRI. Eur J Inorg Chem
12:1975–1986
Farrugia LJ (1999) WinGX suite for small-molecule single-
crystal crystallography. J Appl Crystallogr 32:837–838
Filippin LI, Vercelino V, Marroni NP, Xavier RM (2008) Redox
signalling and the inflammatory response in rheumatoid
arthritis. Clin Exper Immunol 152:415–422
Funkemeier D, Mattes R (1993) Synthesis and structural studies
of copper(II), nickel(II) and cobalt(II) complexes of a
14-membered trans-N2S2 dibenzo macrocycle with 2
pendant pyridylmethyl groups. J Chem Soc Dalton Trans
8:1313–1319
Longmire M, Choyke PL, Kobayashi H (2008) Clearance
properties of nano-sized particles and molecules as imaging
agents: considerations and caveats. Nanomed 3:703–717
Loving GS, Mukherjee S, Caravan P (2013) Redox-activated
manganese-based MR contrast agent. J Amer Chem Soc
135:4620–4623
Mattes R, Muhlenbrock C, Leeners K, Pyttel C (2004) Metal
complexes with N2O2S2 donor set. Synthesis and charac-
terization of the cobalt(II), nickel(II), and copper(II)
complexes of a 15-and a 16-membered bis(2-hydrox-
yethyl) pendant macrocyclic ligand. Z Anorg Allg Chem
630:722–729
Min C, Shao H, Liong M, Yoon T-J, Weissleder R, Lee H (2012)
Mechanism of magnetic relaxation switching sensing. ACS
Nano 6:6821–6828
Persson I (2010) Hydrated metal ions in aqueous solution: how
regular are their structures? Pure Appl Chem 82:1901–1917
Profumo E, Buttari B, Rigano R (2011) Oxidative stress in
cardiovascular inflammation: its involvement in autoim-
mune responses, Int J Inflamm Article ID 295705
Sheldrick GM, SHELXL-97 (1998) Programs for crystal
structure analysis (release 97-2), Institut fu¨r Anorganische
¨
Chemie der Universitat Gottingen, Tammanstrasse 4,
D-3400 Gottingen, Germany
¨
Sherry AD, Wu YK (2013) The importance of water exchange
rates in the design of responsive agents for MRI. Curr Opin
Chem Biol 17:167–174
Sowden RJ, Trotter KD, Dunbar L, Craig G, Erdemli O, Spickett
CM, Reglinski J (2013) Reactions of copper macrocycles
with antioxidants and HOCl: potential for biological redox
sensing. Biometals 26:85–95
Taylor MK, Trotter KD, Reglinski J, Berlouis LEA, Kennedy
AR, Spickett CM, Sowden RJ (2008) Copper N2S2 Schiff
base macrocycles: the effect of structure on redox poten-
tial. Inorg Chim Acta 361:2851–2862
Thunus L, Lejeune R (1999) Overview of transition metal and
lanthanide complexes as diagnostic tools. Coord Chem
Revs 184:125–155
Gale EM, Mukherjee S, Liu C, Loving GS, Caravan P (2014)
Structure–Redox–Relaxivity relationships for redox
responsive manganese-based magnetic resonance imaging
probes. Inorg Chem 53:10748–10761
Grant D, Refsum H, Rummeny E, Marchal G (1997) Liver
imaging: clinical applications and future perspectives. Acta
Radiol 38:623–630
Tobe ML (1972) Inorganic reaction mechanisms. Thomas
Nelson & Sons Ltd, London
Hancu I, Dixon WT, Woods M, Vinogradov E, Sherry AD,
Lenkinski RE (2010) CEST and PARACEST MR contrast
agents. Acta Radiol 51:910–923
Iwaki S, Hanaoka K, Piao W, Komatsu T, Ueno T, Terai T,
Nagano T (2012) Development of hypoxia-sensitive Gd3?
-based MRI contrast agents. Bioorg Med Chem Lett
22:2798–2892
Kueny-Stotz M, Garofalo A, Felder-Flesch D (2012) Man-
ganese-Enhanced MRI Contrast Agents: From Small
Chelates to Nanosized Hybrids. Eur J Inorg Chem
12:1987–2005
Kupka T, Dziegielewski JO, Pasterna G, Malecki JG (1992)
Copper-D-penicillamine complex as potential contrast
agent for MRI. Mag Reson Imag 10:855–858
Lauffer RB (1987) Paramagnetic metal complexes as water
proton relaxation agents for NMR imaging: theory and
design. Chem Rev 87:901–927
Ledneva E, Karie S, Launay-Vacher V, Janus N (2009) Deray G
renal safety of gadolinium-based contrast media in patients
with chronic renal insufficiency. Radiology 250:618–628
Trotter KD, Reglinski J, Robertson K, Forgie JC, Parkinson JA,
Kennedy AR, Armstrong DR, Sowden RJ, Spickett CM
(2009) Structural studies of trans-N2S2 Copper macrocy-
cles. Inorg Chim Acta 362:4065–4072
Trotter KD, Taylor MK, Forgie JC, Reglinski J, Berlouis LEA,
Kennedy AR, Spickett CM, Sowden RJ (2010) The struc-
tural and electrochemical consequences of hydrogenating
Copper N2S2 Schiff base macrocycles. Inorg Chim Acta
363:1529–1538
Tsitovich PB, Burns PJ, Mckay AM, Morrow JR (2014) Redox-
activated MRI contrast agents based on lanthanide and
transition metal ions. J Inorg Biochem 133:143–154
Viger ML, Sankaranarayanan J, de Gracia Lux C, Chan M,
Almutairi A (2013) Collective activation of MRI agents via
encapsulation and disease-triggered release. J Am Chem
Soc 135:7847–7850
Vologdin N, Rolla GA, Botta M, Tei L (2013) Orthogonal
synthesis of a heterodimeric ligand for the development of
the Gd-III-Ga-III ditopic complex as a potential pH-sen-
sitive MRI/PET probe. Org Biomol Chem 11:1683–1690
123

Biometals
Weast RC (1980) Handbook of chemistry and physics, 60th edn.
CRC Press, Boca Raton
Yam VWW, Lo KKW (1999) Recent advances in utilization of
transition metal complexes and lanthanides as diagnostic
tools. Coord Chem Rev 184:157–240
Yu M, Beyers RJ, Gorden JD, Cross NJ, Goldsmith CR (2012) A
magnetic resonance imaging contrast agent capable of
detecting hydrogen peroxide. Inorg Chem 51:9153–9155
123