Cu2+ selective chelators relieve copper-induced oxidative stress: In vivo

Article abstract of DOI:10.1039/c8sc04041a

Copper ions are essential for biological function yet are severely detrimental when present in excess. At the molecular level, copper ions catalyze the production of hydroxyl radicals that can irreversibly alter essential bio-molecules. Hence, selective copper chelators that can remove excess copper ions and alleviate oxidative stress will help assuage copper-overload diseases. However, most currently available chelators are non-specific leading to multiple undesirable side-effects. The challenge is to build chelators that can bind to copper ions with high affinity but leave the levels of essential metal ions unaltered. Here we report the design and development of redox-state selective Cu ion chelators that have 10⁸ times higher conditional stability constants toward Cu²⁺ compared to both Cu⁺ and other biologically relevant metal ions. This unique selectivity allows the specific removal of Cu²⁺ ions that would be available only under pathophysiological metal overload and oxidative stress conditions and provides access to effective removal of the aberrant redox-cycling Cu ion pool without affecting the essential non-redox cycling Cu⁺ labile pool. We have shown that the chelators provide distinct protection against copper-induced oxidative stress in vitro and in live cells via selective Cu²⁺ ion chelation. Notably, the chelators afford significant reduction in Cu-induced oxidative damage in Atp7a^-/- Menkes disease model cells that have endogenously high levels of Cu ions. Finally, in vivo testing of our chelators in a live zebrafish larval model demonstrate their protective properties against copper-induced oxidative stress.

Full text of DOI:10.1039/c8sc04041a

Showcasing research from Dr. Ankona Datta’s laboratory,
Department of Chemical Sciences, Tata Institute of
Fundamental Research, Mumbai, India.
As featured in:
2+
Cu selective chelators relieve copper-induced oxidative
stress in vivo
2
+
8
Cu selective chelators exhibit 10 times higher conditional
2+
+
stability constants toward Cu over Cu and other biologically
relevant metal ions. The chelators specifically remove excess
2+
Cu ions from living cells and alleviate Cu induced oxidative
stress all the way from Menkes disease model cells of Cu overload
to a live multicellular vertebrate system. The unique selectivity
of our chelators indicates tremendous potential in minimizing
side-effects associated with non-specific Cu ion chelation based
therapy currently used treat severe neurological and metabolic
disorders like, Wilson’s disease, amyotrophic lateral sclerosis, and
Alzheimer’s disease.
See Ankona Datta et al.,
Chem. Sci., 2018, 9, 7916.
rsc.li/chemical-science
Registered charity number: 207890

Chemical
Science
View Article Online
View Journal | View Issue
EDGE ARTICLE
2
+
Cu selective chelators relieve copper-induced
oxidative stress in vivo†
Cite this: Chem. Sci., 2018, 9, 7916
a
a
b
c
All publication charges for this article Ananya Rakshit,
Kaustav Khatua,
*
Vinit Shanbhag,
Peter Comba
have been paid for by the Royal Society
a
and Ankona Datta
of Chemistry
Copper ions are essential for biological function yet are severely detrimental when present in excess. At the
molecular level, copper ions catalyze the production of hydroxyl radicals that can irreversibly alter essential
bio-molecules. Hence, selective copper chelators that can remove excess copper ions and alleviate
oxidative stress will help assuage copper-overload diseases. However, most currently available chelators
are non-specific leading to multiple undesirable side-effects. The challenge is to build chelators that can
bind to copper ions with high affinity but leave the levels of essential metal ions unaltered. Here we
8
report the design and development of redox-state selective Cu ion chelators that have 10 times higher
2+
+
conditional stability constants toward Cu compared to both Cu and other biologically relevant metal
2+
ions. This unique selectivity allows the specific removal of Cu ions that would be available only under
pathophysiological metal overload and oxidative stress conditions and provides access to effective
+
removal of the aberrant redox-cycling Cu ion pool without affecting the essential non-redox cycling Cu
labile pool. We have shown that the chelators provide distinct protection against copper-induced
2+
Received 11th September 2018
Accepted 29th September 2018
oxidative stress in vitro and in live cells via selective Cu ion chelation. Notably, the chelators afford
ꢀ/ꢀ
significant reduction in Cu-induced oxidative damage in Atp7a
Menkes disease model cells that have
DOI: 10.1039/c8sc04041a
endogenously high levels of Cu ions. Finally, in vivo testing of our chelators in a live zebrafish larval
rsc.li/chemical-science
model demonstrate their protective properties against copper-induced oxidative stress.
of cellular lipids, nucleic acids, and proteins leading to severe
1
. Introduction
9,10,12,14–16
oxidative damage (Fig. 1A and B).
Previous studies have
Redox-active metal ions are critical for biological function. shown that metal ion chelators can alleviate oxidative stress
17–23
Copper (Cu), iron (Fe), and manganese (Mn) ions act as co- associated symptoms of metal-induced disorders.
However,
factors for essential enzymes. Recent reports show that labile a major shortcoming of chelation therapy is the non-specic
or weakly bound pools of redox active metal ions also partici- removal of other biologically essential metal ions and metal
pate in neuronal signaling and removal of reactive oxygen co-factors from metallo-enzymes leading to multiple side-
1–4
12,24–28
species. While absolutely essential for biological activity both effects.
Therefore, cell-permeable metal selective chelators
in protein-bound and labile forms, excess redox-active metal that can specically remove excess metal ions responsible for
ions have been associated with severe neuro-degenerative oxidative stress would be extremely valuable motifs for chela-
5–8
disorders. A signicant route by which excess redox-active tion therapy (Fig. 1C).
metal ions affect cellular function is via the Fenton reac-
We have worked on the development of selective cell-
Cu /Cu and Fe /Fe catalyze the production of permeable chelators for redox-active metal ions with a focus
reactive hydroxyl radicals which cause permanent modication on Cu ion chelation. The motivation for our work derives from
9
–13
+
2+
2+
3+
tion.
+
2+
the fact that while Cu /Cu ions are absolutely essential for
biological activity, abnormal Cu homeostasis is implicated in
severe neurological and metabolic disorders like Alzheimer's
a
Department of Chemical Sciences, Tata Institute of Fundamental Research, 1 Homi
Bhabha Road, Colaba, Mumbai-400005, India. E-mail: ankona@tifr.res.in
b
7,29–32
33–37
38,39
Department of Biochemistry, Christopher S. Bond Life Science Center, University of disease,
amyotrophic lateral sclerosis,
cancer,
Missouri, Columbia, USA
c
5,40,41
8,42–44
Menkes disease,
and Wilson's disease.
Universit a¨ t Heidelberg, Anorganisch-Chemisches Institut, Interdisciplinary Center for
Scientic Computing, INF 270, D-69120 Heidelberg, Germany
Electronic supplementary information (ESI) available: Supporting gures and
Intra-cellular Cu ions are tightly regulated such that ‘free’
+
45–49
Cu –aqua species are almost non-existent.
Labile Cu ions
†
are mostly bound to available biological ligands like glutathione
tables, details of LC methods for stability and purity analysis of chelators,
+
and exist in the Cu oxidation state under the normal reducing
ESI-MS spectrum of copper complexes, Job's plot, absorbance data and stability
50,51
constant determination with different metal ions, EPR spectra, control physiological conditions.
These ‘natural’ ligands tune the
1
13
experiments in cells and zebrash larvae, and H NMR, C NMR are included.
reduction potential of Cu such that the physiologically available
See DOI: 10.1039/c8sc04041a
7916 | Chem. Sci., 2018, 9, 7916–7930
This journal is © The Royal Society of Chemistry 2018

View Article Online
Edge Article
Chemical Science
Fig. 1 (A) Scheme highlighting mechanisms that can increase intracellular Cu ion levels. Exposure to external agents and mis-folded peptides
can cause increased levels of reactive oxygen species (ROS) leading to oxidative stress. Increased oxidative stress leads to protein oxidation
+
releasing Cu ions from proteins and also reduces levels of glutathione (GSH) bound Cu . Mutations in Cu ion transporters lead to elevated
+
2+
2+
intracellular Cu ion levels. (B) Cu /Cu catalyzed Fenton reaction produces reactive hydroxyl radicals, HOc. Cu is reduced by cellular
ꢀ
ꢀ
2+
reductants like superoxide (O c ) and hydroascorbate (AscH ) to complete the catalytic cycle. (C) Proposed cell-permeable Cu chelators that
alleviate oxidative stress via selective Cu chelation.
2
2+
1
3,52
labile pool does not participate in Fenton chemistry and the cell conditions.
Therefore, we set forth to develop cell-permeable
2
+
+
is able to efficiently regulate Cu ions to minimize oxidative chelators that would be able to specically bind Cu over Cu .
12,13,52
damage.
However, the Cu balance gets compromised Such chelators would not only serve as scaffolds for the devel-
during disease conditions via multiple mechanisms leading to opment of therapeutics but would also act as strategic chemical
5
,13,35,36,41,43,52
Cu ion overload (Fig. 1A).
Under Cu dysregulation tools for elucidating Cu ion homeostasis.
2
+
conditions, excess Cu ions can bind to other biologically avail-
Achieving Cu selective chelation against this backdrop
2
+
able ligands and misfolded proteins forming copper complexes would also necessitate the removal of aberrant labile Cu
1
5,36,53
with nM to pM affinities.
Importantly, several of these without affecting the labile pools of other biological metal ions.
2
+
newly formed species have redox potentials that now allow Cu- The concentration of the labile pools of metal ions like Ca ,
assisted catalysis of the Fenton reaction.
15,29,36,54
2+
2+
2+
46,56
Mn , Zn and Fe range from mM to nM levels.
Since the
Since the aberrant labile Cu ion pool is also bound to ionic radii of divalent transition metal ions fall within 0.7–0.8 A˚
available ligands, strong chelators are required for effective and all of these metal ions can bind to N and O donor atom
2
+
chelation. However, such strong chelators would also affect the containing ligands, designing a selective chelator with a Cu
+
+
benign non-redox cycling labile Cu ion pool, like Cu –gluta- binding affinity that is several orders of magnitude higher
12,55
thione,
further exacerbating the pathophysiological condi- compared to other metal ions is a signicant coordination
tion. A novel approach to overcome this challenge is to develop chemistry challenge.
2
+
chelators that target the Cu labile pool. This pool is mostly
We have designed a modular synthetic strategy and devel-
55
2+
non-existent inside cells during physiological conditions, but oped mixed N- and O-donor atom containing Cu selective
would be generated as a result of redox cycling during Fenton chelators (Scheme 1). The binding constants of the novel
reaction catalysed by the aberrant Cu ion pool (Fig. 1B and C). chelators toward Cu are 10 times higher than that for other
2
+
8
2
+
+
Such a chelator would also be expected to protect against Cu
biologically relevant metal ions including Cu . Notably, the
liberated during protein mis-metallation under oxidative stress binding affinities of the chelators for other divalent metal ions
This journal is © The Royal Society of Chemistry 2018
Chem. Sci., 2018, 9, 7916–7930 | 7917

View Article Online
Chemical Science
Edge Article
59
important criterion. We, therefore, assessed the stability of the
molecules over time, using LC/ESI-MS. Both compounds 2c and
3c were stable in aqueous buffer as well as in cell-culture
medium for over 12 h as shown in Fig. S3.†
2
+
Ligands 2c and 3c bind Cu ions
2
+
The molecules 2c and 3c were taken forward for studying Cu
chelation properties. Absorption titrations for both ligands were
2
+
performed in the presence of Cu ions in aqueous buffer. The
absorption spectrum of compound 2c showed peaks at 247 nm,
277 nm, and a shoulder at 301 nm (Fig. S4†). For compound 3c,
peaks were observed at 247 nm and 295 nm (Fig. S4†).
A distinct new band was observed for both compounds 2c
2
+
ꢀ1
ꢀ1
and 3c in the presence of Cu at 420 nm (3, 736 M cm ) and
ꢀ
1
ꢀ1
4
50 nm (3, 832 M cm ), respectively (Fig. 2a and b, Table
2
+
S2†). Upon titrating Cu to both ligands, the new bands at
20 nm and 450 nm showed increase in intensity with a sharp
transition and saturation at 2 equivalents of ligand relative to
4
a
Scheme 1 Synthesis of chelators 2c and 3c. Reagents and conditions:
2
+
ꢁ
Cu ions (Fig. 2a and b, inset). Other spectral changes were also
(
a) sodium acetate buffer, pH 4.5 and methanol, 100 C; (b) sodium
2
+
ꢁ
observed upon Cu addition to both ligands. For example, an
increase in the intensity of the 275 nm peak was observed for
ligand 2c. Further, the 247 nm peak shied to 237 nm with
concomitant intensity increase (Fig. 2a). Compound 3c afforded
an increase in intensity for both the 247 nm and 295 nm peaks
borohydride in dry methanol, 0 C.
4
6,56
are below the concentrations of their respective labile pools
2
+
and the chelators do not remove protein bound Cu . The
chelators are cell-permeable and do not affect cell-viability
under applied concentrations. These novel molecules
(
Fig. 2b). A visible color change was observed from colorless to
2
+
greenish upon Cu addition to both ligands (Fig. S5†). Closer
2
+
inspection of the absorption spectrum revealed new bands at
suppress Cu triggered oxidative stress both in live cells and in
vivo. Importantly, the chelators were also able to suppress Cu-
ꢀ
1
ꢀ1
ꢀ1
ꢀ1
6
23 nm (3, 173 M cm ) and 626 nm (3, 275 M cm ),
ꢀ
/ꢀ respectively for 2c and 3c, which explain the observed color
induced oxidative stress in Menkes disease model Atp7a
57
change (Fig. 2c and inset, Table S2†).
cells which have endogenously high levels of Cu ions. Hence,
2
+
Area normalized absorption spectra for Cu titration into
both ligands showed isosbestic points indicating the presence
our ligands stand out not only in affording exquisite selectivity
2
+
for Cu but also for their applicability to biological systems
which we have proven all the way from in cell disease models to
a live multicellular vertebrate system.
2
+
of distinct Cu complexes (Fig. 2d and S6†). The stoichiometry
2
+
of the complexes was conrmed to be 1 : 2 Cu : ligand by the
method of continuous variation from Job's plots (Fig. 2e and
2
+
S7†). The formation of 1 : 2 Cu : ligand complexes was further
conrmed by ESI/MS analysis (Fig. S8 and S9†). The sharp
transitions in absorption titrations (Fig. 2a and b, inset) and
2. Results and discussion
Chelator synthesis and stability in physiological buffer
2+
Job's plots for 2c and 3c with Cu (Fig. 2e and S7†) indicated the
2
+
Mixed N- and O-donor atom containing ligands were synthe- formation of Cu complexes with high binding affinity toward
sized via a modular strategy as shown in Scheme 1. The basic the metal ion. We, therefore, proceeded to determine the
design criterion was to keep the number of synthetic steps stability constants of the ligands with Cu . Since, the absorp-
2
+
minimal. Further, S-donor atoms were avoided to introduce tion titrations for both 2c and 3c indicated sharp transitions
2
+
+
58
2+
selectivity for Cu over Cu . The precursors 2b and 3b were we determined the lower limits for the Cu conditional
synthesized in a single step by a Schiff base condensation stability constants (b) by simulating binding curves using a 1 : 2
reaction between the amine 1 and either aldehyde 2a or 3a metal : ligand binding model (see ESI†). As shown in Fig. 2f, the
2
+
(
Scheme 1). Compounds 2b and 3b were obtained in 93% and data for the titration of Cu to 2c tted to simulated curves
5
4% yields, respectively. An advantage of this strategy is that the generated by considering log b values $12. Similarly, the log b
2
+
substituent on the aromatic aldehyde could be varied to value for Cu binding to 3c was determined to be $13
generate ligands with potentially different metal ion binding (Fig. S10†).
affinities. Compounds 2b and 3b were reduced in the presence
of sodium borohydride. The reduced novel ligands 2c and 3c spectroscopy of the Cu –2c complex in order to obtain insight
were obtained in 77% and 57% yields respectively starting from into its coordination geometry (Fig. S21†). The spin Hamilto-
We also performed electron paramagnetic resonance (EPR)
2
+
2
b and 3b (Scheme 1, Fig. S1 and S2†).
Since the chelators would be nally applied in living cells spectrum of the Cu –2c complex indicate that the Cu ion
and biological systems, stability in aqueous media was an has a tetragonal geometry (pseudo-square planar; g
¼ 2.2220;
nian parameters obtained by simulating the experimental
2
+
2+
k
7918 | Chem. Sci., 2018, 9, 7916–7930
This journal is © The Royal Society of Chemistry 2018

View Article Online
Edge Article
Chemical Science
2
+
Fig. 2 Absorbance response of ligands 2c and 3c with Cu . (a) Ultraviolet (UV)-visible titration of 2c (50 mM) with CuCl
2
(0–60 mM) in buffer.
2
+
2+
Black arrows indicate spectral changes upon addition of Cu . The inset shows the plot of absorbance at 420 nm versus concentration of Cu
.
2+
(
b) UV-visible titration of 3c (50 mM) with CuCl
2
(0–60 mM) in buffer. Black arrows indicate spectral changes upon addition of Cu . The inset
shows the plot of absorbance at 450 nm versus concentration of Cu . (c) Visible-near infra-red (NIR) absorbance spectra for titration of 2c with
Cu . The inset shows visible-NIR absorbance spectra for titration of 3c with Cu . (d) Area normalized absorbance spectra representing titration
2+
2+
2+
2+
of CuCl
2
to 3c. Black arrows indicate isosbestic points. (e) Job's plot depicting absorbance versus mole fraction for the Cu –3c system. (f)
2+
Absorption binding plot for titration of Cu to 2c (50 mM, black squares) with simulated fits (fitted to equation: conditional stability constant b ¼
2
[
ML
2
]/[M][L] ), using log b as 9 (green dashed line), 11 (purple dashed line), 12 (blue dashed line), 13 (red dashed line), 14 (light green dashed line).
Buffer used 20 mM HEPES (100 mM NaCl, pH 7.0).
2
+
gt ¼ 2.0502). The value of g /A is 120.6 cm. This value is used Computed structures of the Cu complexes
k
zz
to assess the degree of geometric distortion and values between
The UV-visible-NIR titration experiments indicate that 2c and 3c
2
+
1
(
05–135 cm are typical of Cu complexes with tetragonal
2+
2+
coordinate to Cu in 1 : 2 Cu : ligand stoichiometry. The
isosbestic points suggest that the 1 : 2 complexes have a higher
stability than the corresponding 1 : 1 complexes (Fig. 2d and
60
square planar) geometry. Further, the EPR data indicate that
neither un-complexed Cu ions nor other Cu complexes like
2
+
2+
2
+
di-nuclear Cu species are present.
This journal is © The Royal Society of Chemistry 2018
Chem. Sci., 2018, 9, 7916–7930 | 7919

View Article Online
Chemical Science
Edge Article
S6†). The UV-visible-NIR spectra of the two complexes indicate
2
+
that phenolate donors are coordinated to Cu , when compared
2
+
to other ligand systems exhibiting Cu –phenolate coordina-
6
1,62
ꢀ1
ꢀ1
tion.
The electronic transitions at 420 nm (3, 736 M cm )
ꢀ
1
ꢀ1
and 450 nm (3, 832 M cm ), respectively, can be assigned to
phenolate-based charge-transfer (CT) transitions, while those
ꢀ
1
ꢀ1
with signicantly lower intensity at 623 nm (3, 173 M cm ),
ꢀ1
ꢀ1
and 626 nm (3, 275 M cm ) respectively, are due to d–
d transitions. The EPR data indicate the formation of a tetrag-
onal (pseudo square-planar) complex. Taken together the UV-
visible-NIR titration and the EPR data indicate the formation
of a 1 : 2 Cu : ligand complex with a pseudo-square-planar
geometry, where two N-and phenolate O-donor atoms each are
coordinated to the central metal ion.
2
+
In order to obtain further information about the coordina-
tion geometries and explore reasons for the formation of 1 : 2
2
+
Cu : ligand complexes, we used computational methods to
predict the corresponding structures. The two ligands 2c and 3c
are potentially penta-dentate and an important question to
answer is: why do the ligands only use either two or three donor
2
+
2+
2
atoms each to coordinate to Cu ? Since ligand-imposed steric Fig. 3 MOMEC optimized structures for (a) [Cu (2c)(OH
2
)
4
] (1 ; 1 : 1)
2+
2
2 2 2
and (b) [Cu (2c) (OH ) ] (1 ; 1 : 2) complexes. Ligand 2c is in the
strain seems to be a major reason, force eld calculations are an
ideal approach to get possible answers to this question.
In order to obtain strain energies for comparison between
different structures, we have optimized possible 1 : 1 complexes
phenolate form. In the structures, carbon, nitrogen, oxygen and
hydrogen atoms are represented in light brown, blue, red and white
colors, respectively. Copper ions are represented in gold.
2
3
3
with 2c as an O,N bidentate (1 ), an O,N,N (2 ,3 ) or N,N,N tri-
3
4
4
2+
dentate (4 ), an O,N,N,O (5 ) or an O,N,N,N tetradentate (6 ) and Competition experiment of 2c and 3c with a high affinity Cu
a pentadentate ligand (7 ; not all possible isomers were chelator and protein-bound Cu
5
2+
considered, for details see Fig. S34†). We have used the well-
established MOMEC program and force eld for the calcula-
We next performed a control experiment in which ethylene-
diamine tetra-acetic acid (EDTA) was titrated into a solution
containing 1 : 2 Cu : ligand 3c. The absorption band at
6
3–67
2+
tions.
Additional parameters involving Cu –Ophenolate
2+
6
1,62
bonding were obtained from known structures
(see ESI,
450 nm decreased with increasing concentrations of EDTA and
2
+
Table S5†). All Cu centers were computed as Jahn–Teller
elongated hexa-coordinate geometries with water molecules
completing the coordination spheres, and the conformational
space was evaluated deterministically. Details of the computa-
tional analysis have been included in the ESI.†
the band completely attenuated at ꢂ0.75 equivalent of EDTA
2+
68
(
Fig. S11†). The Cu –EDTA (1 : 1) complex has a log b of 15.8.
This control experiment indicated that the ligands would not be
2+
2+
able to chelate out Cu from essential enzymes that have Cu
50
binding affinities in the aM–fM regime. To further conrm
that the chelators will not affect enzyme bound copper, we
titrated the chelators to Cu –azurin complex which has a Cu
dissociation constant of 25 fM. Cu –azurin complex has
a characteristic absorption peak at 628 nm due to cysteine–Cu
The structure with 2c coordinated as an O,N bidentate ligand
1 ) afforded the lowest strain energy. The only other structures
2
(
2+
2+
that might compete in terms of steric strain were those with 2c as
69
2+
a tridentate or tetradentate ligand involving amino-salicylate
2+
3
4
donor groups (2 and 5 ), and these structures were 15 and
70
charge-transfer. The absorption of the peak remained unaf-
ꢀ
1
2
1
1 kJ mol higher in energy, respectively than the 1 structure
fected even upon addition of upto 12 equivalents of the chela-
tors 2c and 3c (Fig. S12†). Therefore, the chelators 2c and 3c will
be able to remove excess labile copper without affecting enzyme
bound copper.
(Fig. S34 and Table S6†). However, these structures involve rela-
tively exible 8-membered chelate rings and therefore are
entropically disfavored. Due to severe van der Waals repulsion by
the phenyl substituent of the tertiary amine donor, all other
ꢀ
1
isomers are destabilized by an additional 15–20 kJ mol . The
2
+
2
+
Selectivity of the chelators toward Cu
optimized structures of the lowest energy 1 : 1 and 1 : 2 Cu : 2c
complexes are presented in Fig. 3 and ESI Video.† Hence, we infer Encouraged by the Cu binding data we evaluated the condi-
2
+
2
+
that these ligands coordinate to Cu in a bidentate mode with tional stability constants of 2c and 3c toward other biologically
2
+
high 1 : 2 metal : ligand Cu binding constants. Although we relevant metal ions that can bind to ligands with mixed N- and
cannot completely rule out the possibility of the formation of O-donor atoms. Absorption titrations for both ligands were
+
2+
3+
2+
2+
2+
minor ternary complexes with available ligands from the media, performed in the presence of Cu , Ca , Fe , Fe , Ni , Zn ,
2
+
2+
2+
we note that the structure of the 1 : 2 Cu : 2c complex is in Mn and Co metal ions (Fig. S13–S20†). The titration curves
agreement with the UV-visible-NIR spectra, the spectrophoto- indicated weak binding and the data could be tted to a 1 : 1
metric titration, the ESI-MS spectra, and the EPR spectrum.
binding model. The conditional stability constants ranged
7920 | Chem. Sci., 2018, 9, 7916–7930
This journal is © The Royal Society of Chemistry 2018

View Article Online
Edge Article
Chemical Science
between log b values of 2.53 to 4.48 (Table 1) indicating that the ions. We next proceeded to test the ability of our chelators to
2
+
8
stability constants of the ligands toward Cu were at least 10
alleviate Cu-induced oxidative stress.
+
2+
2+
times higher than the stability constants toward Cu , Ca , Zn ,
and other transition metal ions. The binding data indicated that
c and 3c were indeed highly selective Cu chelators (Table 1).
2
+
Chelators reduce Cu ion catalyzed hydroxyl radical production
in vitro
2
2
+
None of the existing commercial Cu chelators that are used
for either therapy of Cu-overload diseases or for mechanistic
72
First, an in vitro deoxyribose-based assay was performed to test
studies in elucidating Cu regulation afford such distinct selec-
whether the chelators could suppress Cu ion catalyzed forma-
tion of hydroxyl radicals. In this assay, deoxyribose reacts with
in situ generated hydroxyl radicals to produce malondialdehyde.
Malondialdehyde reacts with thiobarbituric acid to produce
a pink chromophore with absorbance maxima at 532 nm
2
+
27,71
tivity toward Cu (Table S3†).
In fact, when we compared the
reported binding constants of these commercial ligands for
2
+
Cu versus other metal ions we found that most of them bind
3
+
more strongly to Fe and also have very high affinities toward
2
+
2+ 27,71
2
+
Zn and Ni .
We also compared our chelators to the few
(Fig. S24†). If a ligand is able to chelate Cu from the reaction
23,68
other previously reported reduced salen chelators
(Table
mixture then the production of the chromophore would be
2
+
S4†). Unlike our chelators the metal binding constants of these
reduced salen chelators for most biologically relevant metal
ions had not been reported. We found that our ligands stood
reduced. Importantly, Cu
chelation would reduce the
production of hydroxyl radicals provided the metal complex
does not promote further Fenton reaction. The absorbance of
the chromophore at 532 nm is therefore used as a readout for
Cu ion catalyzed production of hydroxyl radicals.
2
+
out in their selectivity toward Cu relative to the majority of this
group as well.
Since the spectroscopic characterization and computed
Both ligands 2c and 3c suppressed the production of
hydroxyl radicals as seen from the reduction of the absorbance
at 532 nm (Fig. 4). Desferrioxamine mesylate (DFO) was used as
a positive control since DFO is a known chelator for redox active
metal ions like iron and copper. A related ligand desferriox-
amine B which does not have the mesylate groups has a stability
2
+
structures of the Cu –2c complex indicate that only half of the
ligand comprising a phenolate O atom and the secondary N
atom coordinate to the metal ion we investigated if a minimal
control binding unit, 2-((benzylamino)methyl)phenol, would be
sufficient to explain the observed selectivity. Absorption titra-
2
+
2
+
71
tion data of Cu to the control molecule (performed under
constant log b value of 14.1 for a 1 : 1 Cu : ligand complex.
same experimental conditions as 2c and 3c, Fig. S22†) indicated
The plots for absorbance at 532 nm versus chelator concentra-
tion for ligands 2c and 3c followed a similar trend as that of
DFO with greater than 90% reduction in absorbance at
concentrations above 50 mM (Fig. 4). On the other hand, sali-
2
+
2+
that the ligand can bind to Cu ions in 1 : 2 Cu : ligand
stoichiometry with a log b value of 7.0. However, interestingly,
the ligand did not afford a change in its absorption spectrum in
the presence of other essential metal ions under identical
experimental conditions, suggesting that the stability constants
of the ligand toward other metal ions were very low (Fig. S23†).
This data indicates that the control molecule, although
2
+
71
cyaldehyde, a weaker chelator (log b for Cu 7.4) only partially
reduced the formation of hydroxyl radicals (Fig. 4). The deoxy-
ribose assay clearly indicated that the ligands 2c and 3c would
2
+
be able to chelate Cu and suppress Cu ion induced hydroxyl
2
+
a weaker Cu binder (log b 7.0) compared to 2c and 3c (log b $
2
+
1
2), retains its selectivity for Cu binding. This result suggests
that the other half of the molecules 2c and 3c would be
2
+
important in pre-orienting the binding unit for Cu coordina-
tion thereby increasing the binding affinity. Further, the
computational studies show that the steric strain provided by
the central N-phenyl moiety drives the binding toward a biden-
tate mode and disfavors the use of other donor atoms in the
2
+
ligand. This is a key feature that leads to Cu selectivity since
the resultant pseudo square-planar (tetragonal) geometry
2
+
adopted by the Cu complex is a highly preferred ligand
9
2+
orientation for d Cu systems when compared to other metal
Table 1 Conditional stability constant (log b) for the chelators with
different metal ions
2+
+
2+
2+
2+
2+
2+
3+
2+
Compound Cu
Cu
Mn
Co
Zn
Ca
Fe
Fe
Ni
Fig. 4 Effect of chelators 2c and 3c on deoxyribose degradation via
Cu ion mediated hydroxyl radical formation. Salicylaldehyde (SAL) and
desferroxamine mesylate (DFO) were used as controls. Ratio of values
of absorbance in the presence of chelator (A) versus absorbance in the
a
a
2
3
c
c
$12 4.19 n.b
3.97 3.93 n.b 3.63 3.39 2.53
a
a
$13 4.48 4.20 n.b
3.90 n.b 4.05 3.89 3.10
a
absence of chelator (A ) <1 signify protection against hydroxyl radical
mediated deoxyribose degradation. Conditions: CuSO (10 mM), 2-
4
No binding was detected. Conditional stability constant, b1n ¼ [ML
n
]/
o
n
2+
[
M][L] ; n ¼ 2 for Cu and n ¼ 1 for other metal ions. All the absorbance
measurements were carried out in 20 mM HEPES buffer (100 mM NaCl, deoxyribose (15 mM), H
2
O
2
(100 mM), ascorbic acid (2 mM), at pH 7.4 in
pH 7.0).
phosphate buffer.
This journal is © The Royal Society of Chemistry 2018
Chem. Sci., 2018, 9, 7916–7930 | 7921

View Article Online
Chemical Science
Edge Article
radical formation. Importantly, unlike chelators such as DFO
which have extremely high binding affinities for most metal
7
1,73
2+
ions,
our chelators are highly selective for Cu . The in vitro
2
+
assay also indicated that the Cu complexes of 2c and 3c would
not promote further Fenton chemistry. This is an important
criterion for Cu chelation targeted toward assuaging Cu-
induced oxidative stress since there are multiple examples of
2
+
chelators that can bind to Cu and promote the production of
52,74–77
hydroxyl radicals.
Hence, our chelators can be potentially
useful for treating Cu ion overload associated disorders.
2
+
Chelators remove Cu ions from Cu exposed living cells
As a rst step toward evaluating the chelators in living systems,
we wanted to test whether the chelators could remove Cu ions
2
+
from Cu overloaded cells. Phen Green FL diacetate is a cell
permeable uorescent dye that is quenched in the presence of
Cu ions, and has been used effectively to image intracellular Cu
2
+
78–80
ions in Cu overloaded biological systems.
Human embry-
onic kidney cells (HEK293T) were used for the experiments.
Cells were incubated with Phen Green FL and imaged in
a confocal microscope. In comparison to control untreated cells
2
+
the Cu -treated cells showed a signicant decrease in uores-
2
+
cence intensity (Fig. 5a and c, ***p # 0.001). When the Cu -
treated cells were incubated with the chelators 2c and 3c for
3
0 min the Phen Green uorescence intensities afforded
2
+
signicant recovery and were similar to the Cu untreated
controls (Fig. 5b and c). The recovery of Phen Green uores-
cence upon chelator treatment in Cu overloaded cells along
with the distinct Cu selectivity of our chelators indicate that
these chelators will be able to efficiently remove Cu ions from
2
+
2
+
2
+
Fig. 5 Confocal fluorescence images of live HEK293T cells to show
2+
the effect of Cu chelators using a Cu ion responsive dye Phen Green
FL. All cells were incubated with Phen Green FL dye (5 mM) initially for
cells under oxidative stress conditions.
3
0 min before any further treatments. (a) Left to right: control cells;
cells treated with CuCl (20 mM) for 30 min. (b) Left to right: cells
treated with CuCl (20 mM) for 30 min followed by 2c (80 mM) for
Chelators suppress Cu induced oxidative stress in living cells
2
2
Our in vitro data had shown that the chelators could attenuate Cu 30 min; CuCl (20 mM) for 30 min followed by 3c (80 mM) for 30 min.
2
induced production of hydroxyl radicals which is a major Cells were washed three times with PBS buffer after each addition and
then imaged (lex/em: 488/498ꢀ600 nm). Fluorescence intensity
analyses of confocal images were carried out by using ImageJ soft-
ware. (c) Bar plots represent the intensity analyses results. Intensity
data were normalized to intensity of control untreated cells. Error bars
pathway by which Cu ions cause oxidative damage. We have also
2
+
validated that our chelators would be able to remove Cu ions
from live cells. The next key step was to test the ability of the
2
+
chelators to alleviate Cu induced oxidative stress in living cells. denote SEM, n ¼ 3. Statistical analyses were performed using an
2
+
HEK293T cells were incubated with Cu and the level of reactive unpaired, two-tailed Student's t-test (**p # 0.01, ***p # 0.001).
Images were acquired using a 40ꢃ objective; scale bar: 20 mm.
oxygen species (ROS) in the cells was monitored using CellROX®
81
Deep Red which emits in the red (lex: 633 nm, lem: 641–700 nm)
upon oxidation with ROS (Fig. 6 and S25†). Fig. 6 depicts confocal (Fig. 6 and S25†). Moreover, pre-treating the cells with the
2
+
images of live HEK293T cells. The basal level of ROS in cells was chelators prior to Cu incubation also protected the cells from
determined by using control cells that had not been treated with Cu-induced oxidative stress (Fig. S26†).
2
+
2+
Cu . Addition of the chelators to cells did not cause any increase
in the emission intensity of CellROX indicating that the chelators chelators being Cu specic should not alleviate Fe-induced
Fe ions also catalyse Fenton reaction. However, our
2
+
2
+
alone did not induce any oxidative stress (Fig. 6). Cu treated oxidative stress. A control experiment in the presence of Phen
cells induced a signicant increase in ROS over control untreated Green which also responds to Fe indicated that the chelators
2
+
2
+
cells (****p # 0.0001, Fig. 6). Both CuCl
2
(Fig. 6) and CuSO
4
could not remove Fe ions from cells treated with an Fe salt
(
Fig. S25†) showed similar CellROX responses indicating that the (Fig. S27†). Further, CellROX uorescence also increased in
2
+
2+
oxidative stress was due to Cu , rather than the anion. Finally, cells incubated with Fe indicating increased oxidative stress.
2
+
when the Cu treated cells were incubated with the chelators However, the chelators could not reduce CellROX uorescence
a signicant decrease in the CellROX emission intensity was (Fig. S28†) conrming the selectivity of our chelators toward
2
+
observed indicating that both ligands 2c and 3c were able to Cu in live cells.
2
+
rescue the cells from Cu induced oxidative stress conditions
7922 | Chem. Sci., 2018, 9, 7916–7930
This journal is © The Royal Society of Chemistry 2018

View Article Online
Edge Article
Chemical Science
Fig. 6 Confocal fluorescence images of live HEK293T cells in presence of chelators and a ROS sensitive dye. (a) Left to right: control cells; cells
treated with chelator 2c (40 mM) for 30 min; with CuCl (20 mM) for 30 min; with CuCl (20 mM) for 30 min followed by 2c (40 mM) for 30 min. (b)
Left to right: control cells; cells treated with 3c (40 mM) for 30 min; CuCl (20 mM) for 30 min; CuCl (20 mM) for 30 min followed by 3c (40 mM) for
0 min. Cells were stained with CellROX (5 mM) for 30 min at the final step before imaging. (lex/em: 633/641–700 nm). Cells were washed three
2
2
2
2
3
times with PBS buffer after each addition and then imaged. Lower panels of (a) and (b) shows bright field images of cells overlaid with confocal
images. Fluorescence intensity analyses of confocal images were carried out by using ImageJ software. Bar plots represent the intensity analyses
results for (c) chelator 2c and (d) chelator 3c. Intensity data were normalized to intensity of CuCl
Statistical analyses were performed using an unpaired, two-tailed Student's t-test (****p # 0.0001). Images were acquired using a 40ꢃ objective;
scale bar: 20 mm.
2
treated cells. Error bars denote SEM, n ¼ 3.
This journal is © The Royal Society of Chemistry 2018
Chem. Sci., 2018, 9, 7916–7930 | 7923

View Article Online
Chemical Science
Edge Article
2
7,82
As a positive control, D-penicillamine (D-Pen)
a cell- Both chelators 2c and 3c had no effect on LPS induced oxidative
2
+
permeable non-selective Cu chelator (Table S3†) also sup- stress in live cells demonstrating that the chelators function via
pressed Cu induced oxidative stress to a similar extent as Cu
2
+
2+
chelation in alleviating Cu-induced oxidative stress
chelators 2c and 3c (Fig. 6a–d and S29a, b†). However, our (Fig. S29c and d†).
2
+
chelators are highly selective toward Cu and hence have
Further, we simultaneously imaged Cu ions and ROS in live
2
+
a signicant advantage over other non-specic Cu chelators. cells to rstly show that regions of high ROS in Cu-overloaded
In order to further conrm that the chelators 2c and 3c indeed cells colocalize with Cu ions; and secondly to show that the
2
+
2+
2+
reduce Cu -induced oxidative stress via Cu chelation and not chelators can access these regions, chelate out Cu , and effi-
via the removal of ROS through reactions with the phenolic ciently reduce the ROS burden. Phen Green FL and CellROX
83,84
moieties in the chelators,
we performed an additional Deep Red were used to image Cu ions and ROS respectively.
control experiment by monitoring the effect of the chelators on Cu untreated cells showed high Phen Green intensity and low
lipopolysaccharide induced oxidative stress. Lipopolysaccha- CellROX intensity indicating low Cu and low ROS levels,
rides generate ROS via the activation of toll-like receptors which respectively (Fig. 7A, B and C, a panels). Cells treated with Cu
2
+
2
+
2
+
85,86
in turn activate NADPH oxidase leading to oxidative stress.
showed low Phen Green intensity due to quenching of the Phen
Fig. 7 Simultaneous imaging of copper levels and oxidative stress in live HEK293T cells in the presence and absence of chelators 2c and 3c. Phen
Green FL was used to image Cu ions and CellROX was used to image ROS. (A) Top to bottom (a–d): confocal images of control untreated cells;
cells treated with CuCl
2
(20 mM) for 30 min; cells treated with CuCl
20 mM) for 30 min followed by 3c (80 mM) for 30 min. All cells were incubated with Phen Green FL (5 mM) for 30 min prior to any other treatment
ex/em: 488/498–600 nm) and incubated with CellROX (5 mM) for 30 min at the final step before imaging (lex/em: 633/641–700 nm). Cells were
2 2
(20 mM) for 30 min followed by 2c (80 mM) for 30 min; cells treated with CuCl
(
(
l
washed three times with PBS buffer after each addition and then imaged. Number of washings was identical for all experiments. Left to right:
HEK293T cells showing Phen Green FL emission, CellROX emission, and overlaid confocal images. (B) The white lines in the overlaid images were
used to calculate the fluorescence intensity of Phen Green FL (green) and CellROX (blue) in the line scan along the direction of 1 to 2. (C) Bar plots
represent the intensity analyses results for panel (A, a–d). Intensity data were normalized to intensity of control untreated or CuCl
Error bars denote SEM, n ¼ 3. Images were acquired using a 40ꢃ objective; scale bar: 20 mm.
2
treated cells.
7924 | Chem. Sci., 2018, 9, 7916–7930
This journal is © The Royal Society of Chemistry 2018

View Article Online
Edge Article
Chemical Science
Green uorescence by Cu ions throughout the cells (Fig. 7A, B Chelators alleviate oxidative stress in vivo in a zebrash larval
and C, b panels). CellROX intensity was high in the cytoplasm model
indicating oxidative stress conditions (Fig. 7A, B and C,
b panels). Upon treatment with the chelators, the Phen Green
intensity recovered with a simultaneous decrease in the Cell-
To evaluate the efficacy of 2c and 3c compounds in vivo, we
investigated their ability to modulate pathways of redox
signaling using a zebrash model. Zebrash larvae have
emerged as a robust vertebrate model system for toxicological
ROX intensity from the cytoplasm indicating that the chelators
2
+
can distinctly reduce Cu induced oxidative stress in live cells
studies aimed at understanding the effects and mechanisms
2
+
via Cu chelation (Fig. 7A, B and C, c and d panels).
88–90
underlying oxidative stress.
Pathways related to redox
Finally, we checked whether the chelators would affect cell-
viability. More than 93% cells were viable up to 4 h of incuba-
tion with 100 mM of the chelators (Fig. S30†). Hence, the
chelators would have minimal toxicity when applied up to micro
molar levels (Fig. S30†). All of the confocal imaging experi-
ments, intensity analyses of the images, and cell-viability data
taken together highlight the efficacy of the novel chelators
toward alleviating Cu ion induced oxidative stress conditions
and indicate the potential of the chelators toward therapeutic
applications. We, therefore, took the chelators a next step
forward and tested their applicability in a cellular disease model
representing endogenous Cu ion overload.
signaling and lipid metabolism in zebrash are highly analo-
91,92
gous to those in humans.
zebrash have an enhanced oxidative stress response.
According to literature reports
93–95
One
of the major factors responsible for the increased oxidative
stress response is the presence of comparatively high levels of
unsaturated fatty acids that help sh maintain the uidity of
92
cell membranes at lower temperatures. The optical trans-
parency of zebrash larvae make them suitable for optical
81
imaging with ROS sensitive optical probes like CellROX. We,
therefore, tested the efficacy of the chelators toward alleviating
2+
Cu induced oxidative stress in live zebrash larvae.
3
.5 days post fertilization (dpf) wildtype zebrash larvae were
2
+
exposed to Cu by incubating the larvae in a medium con-
taining copper salts. The larvae treated with copper salts
exhibited increased oxidative stress as observed by increased
intensity of CellROX. Highest intensities were observed in the
tail muscles, gut, lateral ns, and around the eyes of the larvae
ꢀ
/ꢀ
Effect of chelators on oxidative stress in an Atp7a
cell
model of Menkes disease
2
+
In order to ascertain the efficacy of our Cu selective chelators
in a disease model of copper overload, we tested the ligands 2c
and 3c in mouse embryonic broblast cells lacking the Atp7a
(Fig. 9a). The swimming motion of zebrash larvae are well-
established where the larvae perform C-start motions during
57
gene, which is mutated in Menkes disease (Fig. S31† and 8).
ATP7A is a P-type ATPase required for copper export. Loss of
96,97
which the tail n is bent into a C-shape.
This movement
helps the sh propel itself forward and dictates the speed of
4
1,87
ATP7A function causes copper hyper-accumulation,
previous ICP-MS studies have shown that the Atp7a
and
96,97
swimming.
The C-start movement is an energy-demanding
ꢀ
/ꢀ
mouse
process for the muscles in the tail n implying high mito-
chondrial activity in this region. Further, the lateral ns help
the sh maintain balance during swimming which is also
embryonic broblast cells accumulate 6-fold more copper in the
57
cytoplasm compared to isogenic wildtype cells. Higher Cu
levels would induce oxidative stress provided the excess Cu ions
are bound to cellular ligands that promote redox cycling leading
to Cu ion catalyzed ROS production. Therefore, we decided to
98
associated with increased muscle activity in these regions. The
gut undergoes spontaneous peristaltic movements starting
from 2 dpf. These peristaltic movements are driven by the gut
muscles which are controlled by the autonomous nervous
ꢀ
/ꢀ
check the level of ROS in the Atp7a
mouse embryonic
ꢀ/ꢀ

broblast cells. The Atp7a
cells showed signicantly higher
99
system. Finally, the zebrash larvae actively scan their lateral
oxidative stress compared to the wildtype cells (Fig. 8a–d le

eld of view as they swim to visualize their environment which
panels, and Fig. 8e and f). This data implied that the increased
100
requires extra-ocular muscles that help in eye movement. In
summary, we noted that the anatomical regions that showed
increased oxidative stress in the zebrash larvae upon Cu
ꢀ/ꢀ
endogenous level of Cu ions in the Atp7a
cells were capable
of promoting oxidative stress. Importantly, both 2c and 3c were
ꢀ
/ꢀ
able to reduce the levels of oxidative stress in the Atp7a
cells
exposure were regions of high muscle activity and hence would
(Fig. 8a–d right panels, Fig. 8e and f). This data demonstrates
81,101
be associated with greater mitochondrial density.
Since,
the efficacy of 2c and 3c against endogenous Cu ion overload
hydrogen peroxide production increases with increased mito-
induced oxidative stress.
102,103
chondrial activity,
these regions would be more susceptible
2
+
The fact that our chelators reduce oxidative stress via Cu
to metal induced oxidative stress via the Fenton pathway in
accordance with our observation (although other regions might
also accumulate metal ions).
We were pleased to note that the level of oxidative stress in
the Cu-treated larvae signicantly decreased in the presence of
the chelators to levels similar to untreated larvae (Fig. 9b and d).
The 2c and 3c chelators alone did not induce any oxidative
stress in the larvae (Fig. S32†), and their ability to suppress
copper-stimulated ROS was similar to D-Pen (Fig. S33†). Control
experiments with LPS treated larvae showed that 2c and 3c
ion chelation has been established both directly (Fig. 7A–C) and
through control experiments in the presence of LPS (Fig. S29c
and d†). These results when combined with the data on the
ꢀ/ꢀ
Atp7a
cells, indicate that Cu ion overload is indeed associ-
ated with increased oxidative stress catalyzed by Cu species that
can perform redox cycling to generate ROS. Our chelators catch
the Cu ions in the +2 oxidation state and generate a chelated
species that blocks the redox cycling to alleviate Cu-induced
oxidative stress. Hence, we decided to check the efficacy of the
chelators in a multicellular, vertebrate model system.
This journal is © The Royal Society of Chemistry 2018
Chem. Sci., 2018, 9, 7916–7930 | 7925

View Article Online
Chemical Science
Edge Article
Fig. 8 Confocal fluorescence imaging in mouse embryonic fibroblast wildtype (MEF WT) and mouse embryonic fibroblast Atp7a knockout (MEF
ꢀ/ꢀ
Atp7a ) cells in presence of chelators and a ROS sensitive dye. (a) MEF WT cells: left to right, untreated cells; cells treated with chelator 2c (40
ꢀ/ꢀ
mM) for 60 min. (b) MEF Atp7a
untreated cells; cells treated with chelator 3c (40 mM) for 60 min. (d) MEF Atp7a
40 mM) for 60 min. Cells were stained with CellROX (5 mM) for 30 min at the final step before imaging (lex/em: 633/641–700 nm). Cells were
cells: left to right, untreated cells; cells treated with chelator 2c (40 mM) for 60 min. (c) MEF WT cells: left to right,
ꢀ
/ꢀ
cells: left to right, untreated cells; cells treated with chelator 3c
(
washed three times with PBS buffer after each addition and then imaged. Lower panels of (a)–(d) shows bright field images of cells overlaid with
confocal images. Fluorescence intensity analyses of confocal images were carried by using ImageJ software. Bar plots represent the intensity
ꢀ/ꢀ
analyses of fluorescence images for chelators (e) 2c and (f) 3c. Fluorescence intensities were normalized to intensity of MEF Atp7a
cells. Error
bars denote SEM, n ¼ 3. Statistical analyses were performed using an unpaired, two-tailed Student's t-test (**p # 0.01, ****p # 0.0001). Images
were acquired using a 40ꢃ objective; scale bar: 20 mm.
7926 | Chem. Sci., 2018, 9, 7916–7930
This journal is © The Royal Society of Chemistry 2018

View Article Online
Edge Article
Chemical Science
Fig. 9 Confocal fluorescence images depicting the effect of chelators on live 3.5 dpf zebrafish larvae treated with copper salts. (a–c) Upper
panels depict Z stacked confocal fluorescence images and lower panels depict bright field images overlaid with Z stacked confocal fluorescence
images. (a) Control Cu-untreated larvae (left) and larvae treated with CuCl
2
(30 mM) for 30 min (right). (b) Larvae treated with CuCl
2
(30 mM) for
3
0 min followed by 2c (60 mM) for 30 min (left) and larvae treated with CuCl
2
(30 mM) for 30 min followed by 3c (60 mM) for 30 min (right). (c)
ꢀ1
ꢀ1
Larvae treated with LPS (10 mg mL ) for 30 min (left) and larvae treated with LPS (10 mg mL ) for 30 min followed by 2c (60 mM) for 30 min (right).
White arrows indicate regions exhibiting high oxidative stress. Larvae were stained with CellROX (10 mM) for 30 min in the final step before
imaging. (d) Bar plots representing average fluorescence intensities obtained from intensity analysis of confocal images of the zebrafish larvae
2
experiments for which representative images are shown in panels (a and b). Intensity data were normalized to intensity of CuCl treated larvae. (e)
Quantification of fluorescence intensities of confocal images of zebrafish larvae for which representative images are shown in panel (c). Intensity
data were normalized to LPS treated larvae. For both panels (d and e) intensity analyses were carried out on Z-stacked confocal images of
zebrafish larvae. ImageJ software (NIH, USA) was used for image analysis. For fluorescence quantification, a region of interest (ROI) was created,
and fluorescence intensity was measured. The reported fluorescence intensity was determined by averaging the measured intensity of at least
fifteen different ROIs from experiments on at least six different zebrafish larvae. Error bars denote SEM, n ¼ 6. Statistical analyses were performed
using an unpaired, two-tailed Student's t-test (**p # 0.01, ***p # 0.001). 20ꢃ objective was used for imaging; scale bar: 200 mm.
This journal is © The Royal Society of Chemistry 2018
Chem. Sci., 2018, 9, 7916–7930 | 7927

View Article Online
Chemical Science
Edge Article
could not rescue LPS induced oxidative stress (Fig. 9c and e). approved by the institutional animal ethics committee of the
Therefore ligands 2c and 3c suppress Cu ion induced oxidative Tata Institute of Fundamental Research.
2
+
stress via Cu chelation in a multicellular, vertebrate organism
indicating the in vivo applicability of the chelators.
Conflicts of interest
There are no conicts to declare.
3. Concluding remarks
The regulation of redox active metal ions involves the intricate
interplay of multiple proteins and biological ligands and
Acknowledgements
56,104
exhibits many shades of subtle homeostatic balance.
While
A. D. acknowledges nancial support from TIFR and Depart-
the concept of chelation therapy has a long history, the appre-
ciation of the balance between the labile and the protein bound
pool and the regulatory factors that maintain this balance is
ment of Atomic Energy, India. The authors acknowledge Prof.
ꢀ/ꢀ
Michael J. Petris for generously sharing MEF Atp7a
cells; Dr
Marion Kerscher for her help with the illustration of MOMEC
optimized structures and Dr Sreelaja Nair for helpful discus-
sions on zebrash larvae results. The authors acknowledge Cell
culture and Imaging facility, Department of Chemical Sciences,
TIFR, India; National NMR facility, TIFR, India; Electron spin
resonance laboratory, Department of Chemical Sciences, TIFR;
Biophotonics Laboratory, Department of Chemical Sciences,
TIFR; and Mass Laboratory, Chemistry Department, Indian
Institute of Technology, Bombay, India.
105
very recent. Hence, most chelation therapy based approaches
for treating metal-overload disorders are based on chelators
27
that are non-specic. These non-specic chelators are also
frequently used for mechanistic studies in unravelling metal
ion regulation, leading to equivocal information on metal ion
104,105
homeostasis.
The development of selective chelators for
a specic oxidation state of a redox-active metal ion remains
a signicant challenge, but is extremely worthwhile given the
wealth of mechanistic information that can be obtained by the
judicious use of such chelators and the tremendous potential in
minimizing the side-effects associated with chelation therapy.
By analyzing the literature on Cu-induced oxidative stress
Notes and references
1
2
C. J. Chang, Nat. Chem. Biol., 2015, 11, 744–747.
5
6,105
and Cu ion speciation in cells,
we realized that chelating out
M. J. Daly, E. K. Gaidamakova, V. Y. Matrosova,
A. Vasilenko, M. Zhai, A. Venkateswaran, M. Hess,
M. V. Omelchenko, H. M. Kostandarithes, K. S. Makarova,
L. P. Wackett, J. K. Fredrickson and D. Ghosal, Science,
2
+
Cu ions which would mostly be generated under oxidative
stress conditions inside cells would provide access to selective
Cu ion chelation without affecting the essential intracellular
2
+
metal pool. An important criterion was that the chelated Cu
2004, 306, 1025–1028.
should not further promote Fenton chemistry. Using our redox-
state selective chelation strategy, we have therefore developed
3
4
K. Barnese, E. B. Gralla, J. S. Valentine and D. E. Cabelli,
Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 6892–6897.
Y. Hatori, Y. Yan, K. Schmidt, E. Furukawa, N. M. Hasan,
N. Yang, C.-N. Liu, S. Sockanathan and S. Lutsenko, Nat.
Commun., 2016, 7, 10640.
2
+
novel chelators that exhibit unique Cu selectivity and can
alleviate metal overload induced oxidative stress both in vitro
and in live cells. Importantly, we were able to show that the
2
+
chelators reduce oxidative stress via Cu chelation rather than
by direct ROS scavenging, and were able to suppress ROS
production in a disease model of cellular Cu overload. Finally,
the chelators were able to reduce Cu-induced oxidative stress in
zebrash larvae indicating their ability to function in vivo.
5
M. L. Schlief, T. West, A. M. Craig, D. M. Holtzman and
J. D. Gitlin, Proc. Natl. Acad. Sci. U. S. A., 2006, 103,
14919–14924.
6
7
A. I. Bush, C. L. Masters and R. E. Tanzi, Proc. Natl. Acad. Sci.
U. S. A., 2003, 100, 11193–11194.
E. Gaggelli, H. Kozlowski, D. Valensin and G. Valensin,
Chem. Rev., 2006, 106, 1995–2044.
2
+
+
The high selectivity of our chelators for Cu over Cu and
other biologically relevant metal ions makes these scaffolds
valuable chemical tools for exploring pathways of Cu ion regu-
8
9
S. Lutsenko, Curr. Opin. Chem. Biol., 2010, 14, 211–217.
I. Fridovich, Science, 1978, 201, 875–880.
2
+
lation. Moreover, these highly specic Cu chelators can be
applied as chemical tools to modulate Cu homeostatic pathways
and Cu induced oxidative stress response in neurodegenerative
conditions like Alzheimer's disease and in Cu-overload disor-
ders like Wilson's disease. Finally, the molecular design of the
1
0 Q. Liu, U. Berchner-Pfannschmidt, U. M ¨o ller, M. Brecht,
C. Wotzlaw, H. Acker, K. Jungermann and T. Kietzmann,
Proc. Natl. Acad. Sci. U. S. A., 2004, 101, 4302–4307.
1
1
1 B. Halliwell and J. M. Gutteridge, Biochem. J., 1984, 219,
2
+
Cu binders, synthesized via a simple modular strategy, can
serve as scaffolds for design and development of molecules
aimed at chelation therapy for Cu-overload induced diseases.
1
–14.
2 S. J. Stohs and D. Bagchi, Free Radical Biol. Med., 1995, 18,
21–336.
3
1
1
3 K. Jomova and M. Valko, Toxicology, 2011, 283, 65–87.
4 L. Macomber and J. A. Imlay, Proc. Natl. Acad. Sci. U. S. A.,
Animal ethics statement
2009, 106, 8344–8349.
Zebrash were bred and maintained and experiments were
15 M. F. Poyton, A. M. Sendecki, X. Cong and P. S. Cremer,
performed in accordance with guidelines and protocols
J. Am. Chem. Soc., 2016, 138, 1584–1590.
7928 | Chem. Sci., 2018, 9, 7916–7930
This journal is © The Royal Society of Chemistry 2018

View Article Online
Edge Article
Chemical Science
1
1
1
6 B. C. Dickinson and C. J. Chang, Nat. Chem. Biol., 2011, 7,
04–511.
7 L. K. Charkoudian, D. M. Pham and K. J. Franz, J. Am. Chem.
Soc., 2006, 128, 12424–12425.
8 T. Storr, M. Merkel, G. X. Song-Zhao, L. E. Scott, D. E. Green,
M. L. Bowen, K. H. Thompson, B. O. Patrick, H. J. Schugar
and C. Orvig, J. Am. Chem. Soc., 2007, 129, 7453–7463.
9 Y. Wei and M. Guo, Angew. Chem., Int. Ed., 2007, 46, 4722–
40 Y. Wang, S. Zhu, G. A. Weisman, J. D. Gitlin and M. J. Petris,
PLoS One, 2012, 7, e43039.
5
41 A. Bhattacharjee, H. Yang, M. Duffy, E. Robinson,
A. Conrad-Antoville, Y.-W. Lu, T. Capps, L. Braiterman,
M. Wolfgang, M. P. Murphy, L. Yi, S. G. Kaler, S. Lutsenko
and M. Ralle, J. Biol. Chem., 2016, 291, 16644–16658.
42 R. Linz and S. Lutsenko, J. Bioenerg. Biomembr., 2007, 39,
403–407.
43 E. V. Polishchuk, M. Concilli, S. Iacobacci, G. Chesi,
N. Pastore, P. Piccolo, S. Paladino, D. Baldantoni,
S. C. D. van Ijzendoorn, J. Chan, C. J. Chang,
A. Amoresano, F. Pane, P. Pucci, A. Tarallo, G. Parenti,
N. Brunetti-Pierri, C. Settembre, A. Ballabio and
R. S. Polishchuk, Dev. Cell, 2014, 29, 686–700.
44 M. L. Graper, D. Huster, S. G. Kaler, S. Lutsenko,
M. L. Schilsky and D. J. Thiele, Ann. N. Y. Acad. Sci., 2014,
1314, v–vi.
45 L. A. Finney and T. V. Halloran, Science, 2003, 300, 931–936.
46 A. W. Foster, R. Pernil, C. J. Patterson, A. J. P. Scott,
L.-O. Palsson, R. Pal, I. Cummins, P. T. Chivers, E. Pohl
and N. J. Robinson, Nat. Chem. Biol., 2017, 13, 409–414.
47 A. Hong-Hermesdorf, M. Miethke, S. D. Gallaher, J. Kropat,
S. C. Dodani, J. Chan, D. Barupala, D. W. Domaille,
D. I. Shirasaki, J. A. Loo, P. K. Weber, J. Pett-Ridge,
T. L. Stemmler, C. J. Chang and S. S. Merchant, Nat.
Chem. Biol., 2014, 10, 1034–1042.
48 L. Banci, I. Bertini, S. Cio-Baffoni, T. Kozyreva, K. Zovo and
P. Palumaa, Nature, 2010, 465, 645–648.
49 A. K. Boal and A. C. Rosenzweig, Chem. Rev., 2009, 109,
4760–4779.
50 J. J. A. Cotruvo, A. T. Aron, K. M. Ramos-Torres and
C. J. Chang, Chem. Soc. Rev., 2015, 44, 4400–4414.
51 L. Yang, R. McRae, M. M. Henary, R. Patel, B. Lai, S. Vogt
and C. J. Fahrni, Proc. Natl. Acad. Sci. U. S. A., 2005, 102,
11179–11184.
52 M. C. Linder, Mutat. Res., 2001, 475, 141–152.
53 B. Alies, E. Renaglia, M. R ´o zga, W. Bal, P. Faller and
C. Hureau, Anal. Chem., 2013, 85, 1501–1508.
1
2
4725.
0 M. R. Jones, E. Mathieu, C. Dyrager, S. Faissner,
Z. Vaillancourt, K. J. Korshavn, M. H. Lim,
A. Ramamoorthy, V. Wee Yong, S. Tsutsui, P. K. Stys and
T. Storr, Chem. Sci., 2017, 8, 5636–5643.
1 K. J. Franz, Curr. Opin. Chem. Biol., 2013, 17, 143–149.
2 M. Eckshtain-Levi, R. Lavi, D. S. Yut, B. Daniel, O. Green,
O. Fleker, M. Richman, S. Rahimipour, A. Gruzman and
L. Benisvy, Chem. Commun., 2016, 52, 2350–2353.
3 A. Robert, Y. Liu, M. Nguyen and B. Meunier, Acc. Chem.
Res., 2015, 48, 1332–1339.
2
2
2
2
2
4 A. Roberts Eve and L. Schilsky Michael, Hepatology, 2008,
47, 2089–2111.
5 V. Prachayasittikul, S. Prachayasittikul, S. Ruchirawat and
V. Prachayasittikul, Drug Des., Dev. Ther., 2013, 7, 1157–
1
178.
6 P. Leanderson and C. Tagesson, Carcinogenesis, 1996, 17,
45–550.
7 X. Ding, H. Xie and Y. J. Kang, J. Nutr. Biochem., 2011, 22,
01–310.
8 A. J. Baxter and E. P. Krenzelok, Clin. Toxicol., 2008, 46,
083–1084.
2
2
2
2
3
3
3
3
3
5
3
1
9 C. Cheignon, P. Faller, D. Testemale, C. Hureau and
F. Collin, Metallomics, 2016, 8, 1081–1089.
0 M. A. Greenough, A. Ramirez Munoz, A. I. Bush and
C. M. Opazo, Metallomics, 2016, 8, 831–839.
1 P. Faller, C. Hureau and O. Berthoumieu, Inorg. Chem.,
2
013, 52, 12193–12206.
2 D. Pramanik, C. Ghosh and S. G. Dey, J. Am. Chem. Soc.,
011, 133, 15545–15552.
3 J. P. Taylor, R. H. Brown Jr and D. W. Cleveland, Nature,
016, 539, 197–206.
4 S. Sangwan, A. Zhao, K. L. Adams, C. K. Jayson,
M. R. Sawaya, E. L. Guenther, A. C. Pan, J. Ngo,
D. M. Moore, A. B. Soriaga, T. D. Do, L. Goldschmidt,
R. Nelson, M. T. Bowers, C. M. Koehler, D. E. Shaw,
B. G. Novitch and D. S. Eisenberg, Proc. Natl. Acad. Sci. U.
S. A., 2017, 114, 8770–8775.
2
54 S. Rabizadeh, E. B. Gralla, D. R. Borchelt, R. Gwinn,
J. S. Valentine, S. Sisodia, P. Wong, M. Lee, H. Hahn and
D. E. Bredesen, Proc. Natl. Acad. Sci. U. S. A., 1995, 92,
3024–3028.
55 P. Delangle and E. Mintz, Dalton Trans., 2012, 41, 6359–
6370.
2
56 E. L. Que, D. W. Domaille and C. J. Chang, Chem. Rev., 2008,
108, 1517–1549.
3
3
3
3
5 J. S. Valentine and P. J. Hart, Proc. Natl. Acad. Sci. U. S. A.,
57 S. Jia, K. M. Ramos-Torres, S. Kolemen, C. M. Ackerman and
C. J. Chang, ACS Chem. Biol., 2017, 13, 1844–1852.
58 K. Harry and J. D. Spivack, US Pat., No. US2928876, 1960.
59 Q. Wang and K. J. Franz, Bioorg. Med. Chem. Lett., 2017, 27,
4165–4170.
60 E. Faggi, R. Gavara, M. Bolte, L. Fajar ´ı , L. Juli ´a , L. Rodr ´ı guez
and I. Alfonso, Dalton Trans., 2015, 44, 12700–12710.
61 M. Tarnai, PhD thesis, Heidelberg University, 2006.
62 S. Y. Ebrahimipour, I. Sheikhshoaie, J. Castro, W. Haase,
M. Mohamadi, S. Foro, M. Sheikhshoaie and S. Esmaeili-
Mahani, Inorg. Chim. Acta, 2015, 430, 245–252.
2
003, 100, 3617–3622.
6 N. Fujimaki, F. Kitamura and H. Takeuchi, Biochemistry,
013, 52, 5184–5194.
2
7 P. M. Roos, O. Vesterberg, T. Syversen, T. P. Flaten and
M. Nordberg, Biol. Trace Elem. Res., 2013, 151, 159–170.
8 S. Ishida, P. Andreux, C. Poitry-Yamate, J. Auwerx and
D. Hanahan, Proc. Natl. Acad. Sci. U. S. A., 2013, 110,
1
9507–19512.
9 A. Gupte and R. J. Mumper, Cancer Treat. Rev., 2009, 35, 32–
6.
3
4
This journal is © The Royal Society of Chemistry 2018
Chem. Sci., 2018, 9, 7916–7930 | 7929

View Article Online
Chemical Science
Edge Article
6
3 P. Comba, T. W. Hambley and B. Martin, in Molecular
Modeling of Inorganic Compounds, Wiley-VCH Verlag
GmbH & Co. KGaA, 3rd edn with a tutorial, based on
MOMEC3, 2009, pp. 215–292.
83 A. Barzegar and A. A. Moosavi-Movahedi, PLoS One, 2011, 6,
26012.
84 Y. Aboul-Enein Hassan, I. Kruk, A. Kładna, K. Lichszteld
and T. Michalska, Biopolymers, 2007, 86, 222–230.
85 G. Y u¨ cel, Z. Zhao, I. El-Battrawy, H. Lan, S. Lang, X. Li,
F. Buljubasic, W.-H. Zimmermann, L. Cyganek, J. Utikal,
U. Ravens, T. Wieland, M. Borggrefe, X.-B. Zhou and
I. Akin, Sci. Rep., 2017, 7, 2935.
86 A. Ngkelo, K. Meja, M. Yeadon, I. Adcock and P. A. Kirkham,
J. Inammation, 2012, 9, 1.
87 C. White, J. Lee, T. Kambe, K. Fritsche and M. J. Petris,
J. Biol. Chem., 2009, 284, 33949–33956.
6
6
6
6
6
4 P. Comba, M. Str ¨o hle and T. W. Hambley, Helv. Chim. Acta,
1
995, 78, 2042–2047.
5 P. Comba, B. Martin and A. Sanyal, J. Comput. Chem., 2013,
4, 1598–1608.
3
6 J. E. Bol, C. Buning, P. Comba, J. Reedijk and M. Str ¨o hle,
J. Comput. Chem., 1998, 19, 512–523.
7 P. V. Bernhardt and P. Comba, Inorg. Chem., 1992, 31, 2638–
2644.
8 M. G. D. Leed, N. Wolkow, D. M. Pham, C. L. Daniel,
J. L. Dunaief and K. J. Franz, J. Inorg. Biochem., 2011, 105,
88 M. Priyadarshini, L. A. Orosco and P. J. Panula, PLoS One,
2013, 8, 81851.
1
161–1172.
89 R. Zhang, J. Zhao, G. Han, Z. Liu, C. Liu, C. Zhang, B. Liu,
C. Jiang, R. Liu, T. Zhao, M.-Y. Han and Z. Zhang, J. Am.
Chem. Soc., 2016, 138, 3769–3778.
6
7
7
7
7
7
9 J. Marks, I. Pozdnyakova, J. Guidry and P. Wittung-
Stafshede, J. Biol. Inorg. Chem., 2004, 9, 281–288.
0 O. Farver and I. Pecht, Coord. Chem. Rev., 2011, 255, 757–
90 S. Wu, G. Ji, J. Liu, S. Zhang, Y. Gong and L. Shi, Environ.
Toxicol., 2016, 31, 1241–1249.
773.
1 L. G. Sillen and A. E. Martell, Stability Constants of Metal Ion
Complexes, Chemical Society, 17th edn, 1964.
2 B. Halliwell, J. M. C. Gutteridge and O. I. Aruoma, Anal.
Biochem., 1987, 165, 215–219.
91 J. L. Anderson, J. D. Carten and S. A. Farber, in Methods.
Cell. Biol., ed. H. W. Detrich, M. Westereld and L. I. Zon,
Academic Press, 2016, vol. 133, pp. 165–178.
92 L. Fang and Y. I. Miller, Free Radical Biol. Med., 2012, 53,
1411–1420.
93 M. E. Hahn, A. G. McArthur, S. I. Karchner, D. G. Franks,
M. J. Jenny, A. R. Timme-Laragy, J. J. Stegeman,
B. R. Woodin, M. J. Cipriano and E. Linney, PLoS One,
2014, 9, 113158.
´
´
3 E. Farkas, E. A. Enyedy and H. Csoka, Polyhedron, 1999, 18,
2391–2398.
4 A. I. B. Romo, D. S. Abreu, T. d. F. Paulo, M. S. P. Carepo,
E. H. S. Sousa, L. Lemus, C. Aliaga, A. A. Batista,
O. R. Nascimento, H. D. Abru n˜ a and I. C. N. Di ´o genes,
Chem.–Eur. J., 2016, 22, 10081–10089.
94 Q. Wang, X. Tan, S. Jiao, F. You and P.-J. Zhang, PLoS One,
2014, 9, 102492.
75 N. Berthet, V. Martel-Frachet, F. Michel, C. Philouze,
S. Hamman, X. Ronot and F. Thomas, Dalton Trans.,
95 M. Hagedorn, M. McCarthy, V. L. Carter and S. A. Meyers,
PLoS One, 2012, 7, 39397.
2013, 42, 8468–8483.
7
7
7
6 Y. Ma, T. Ogino, T. Kawabata, J. Li, K. Eguchi and S. Okada,
Free Radical Biol. Med., 1999, 27, 227–233.
7 Y. Ma, L. Cao, T. Kawabata, T. Yoshino, B. B. Yang and
S. Okada, Free Radical Biol. Med., 1998, 25, 568–575.
8 J. Zheng, J. R. Lou, X.-X. Zhang, D. M. Benbrook,
M. H. Hanigan, S. E. Lind and W.-Q. Ding, Cancer Lett.,
96 A. Roberts, B. Bill and D. Glanzman, Front. Neural Circuits,
2013, 7, 126.
97 R. C. Eaton, R. K. K. Lee and M. B. Foreman, Prog.
Neurobiol., 2001, 63, 467–485.
98 M. H. Green, R. K. Ho and M. E. Hale, J. Exp. Biol., 2011,
214, 3111–3123.
2
010, 298, 186–194.
99 H. A. Field, K. A. Kelley, L. Martell, A. M. Goldstein and
F. C. Serluca, Neurogastroenterol. Motil., 2009, 21, 304–312.
7
9 P. A. Castro, A. Ramirez, F. J. Sep u´ lveda, C. Peters, H. Fierro,
J. Waldron, S. Luza, J. Fuentealba, F. J. Mu n˜ oz, G. V. De 100 W. Mo, F. Chen, A. Nechiporuk and T. Nicolson, BMC
Ferrari, A. I. Bush, L. G. Aguayo and C. M. Opazo, Front.
Aging Neurosci., 2014, 6, 319.
Neurosci., 2010, 11, 110.
101 V. Romanello and M. Sandri, Front. Physiol., 2015, 6, 422.
8
8
8
0 P. Chavez-Crooker, N. Garrido and G. A. Ahearn, J. Exp. 102 E. Cadenas, Mol. Aspects Med., 2004, 25, 17–26.
Biol., 2001, 204, 1433–1444.
1 V. Mugoni, A. Camporeale and M. M. Santoro, J. Visualized
Exp., 2014, 89, 51328.
103 E. J. Anderson and P. D. Neufer, Am. J. Physiol. Cell Physiol.,
2006, 290, C844–C851.
104 C. M. Ackerman and C. J. Chang, J. Biol. Chem., 2018, 13,
4628–4635.
2 R. Squitti, P. M. Rossini, E. Cassetta, F. Moffa,
P. Pasqualetti, M. Cortesi, A. Colloca, L. Rossi and 105 C. J. Chang, Acc. Chem. Res., 2017, 50, 535–538.
A. Finazzi-Agro', Eur. J. Clin. Invest., 2002, 32, 51–59.
7930 | Chem. Sci., 2018, 9, 7916–7930
This journal is © The Royal Society of Chemistry 2018