A new arene-Ru based supramolecular coordination complex for efficient binding and selective sensing of green fluorescent protein

Article abstract of DOI:10.1039/c3dt53186d

A new dipyridyl ligand is encoded with 120° angularity between its coordination vectors by using a central pyridine carboxamide scaffold to orient two 4-(pyridin-4-ylethynyl)phenyl moieties. The N,N′-bis(4-(pyridin-4- ylethynyl)phenyl)pyridine-2,6-dicarboxamide ligand undergoes self-assembly with a diruthenium arene complex to furnish a [2 + 2] metallacycle with a wedge-like structure. The metallacycle binds to the enhanced green fluorescent protein (EGFP) variant of GFP, resulting in steady-state spectral changes in UV-Vis absorption and emission experiments. These studies indicate that the metallacycle induces conformation changes to the EGFP, disrupting the tripeptide chromophore. Furthermore, gel electrophoresis, circular dichroism and atomic force microscopy studies indicate that binding ultimately leads to aggregation of the protein. Computational investigations indicate a favorable interaction, predominantly between the metallacycle and the Arg168 residue of the EGFP. An interaction with Arg168 and related residues was previously observed for an emission-attenuating antibody, supporting that these interactions induce changes to the photophysical properties of EGFP by disrupting the tripeptidechromophore in a similar manner. Additionally, we have also described the quenching study of the reporter GFP protein in vivo by a new metal complex using reflected fluorescence microscopy. We anticipate that such metal complexes which can passively diffuse into the cells in vivo can serve as potential tools in molecular and drug targeting based biological studies. This journal is the Partner Organisations 2014.

Full text of DOI:10.1039/c3dt53186d

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
A new arene–Ru based supramolecular
coordination complex for efficient binding and
selective sensing of green fluorescent protein†
Cite this: Dalton Trans., 2014, 43,
6032
Anurag Mishra,‡a Sambandam Ravikumar,‡b,f Young Ho Song,a
Nadarajan Saravanan Prabhu,c Hyunuk Kim,*d Soon Ho Hong,b Seyeon Cheon,e
Jaegeun Nohe and Ki-Whan Chi*a
A new dipyridyl ligand is encoded with 120° angularity between its coordination vectors by using a central
pyridine carboxamide scaffold to orient two 4-(pyridin-4-ylethynyl)phenyl moieties. The N,N’-bis(4-
(pyridin-4-ylethynyl)phenyl)pyridine-2,6-dicarboxamide ligand undergoes self-assembly with a diruthe-
nium arene complex to furnish a [2 + 2] metallacycle with a wedge-like structure. The metallacycle binds
to the enhanced green fluorescent protein (EGFP) variant of GFP, resulting in steady-state spectral
changes in UV-Vis absorption and emission experiments. These studies indicate that the metallacycle
induces conformation changes to the EGFP, disrupting the tripeptide chromophore. Furthermore, gel
electrophoresis, circular dichroism and atomic force microscopy studies indicate that binding ultimately
leads to aggregation of the protein. Computational investigations indicate a favorable interaction, pre-
dominantly between the metallacycle and the Arg168 residue of the EGFP. An interaction with Arg168 and
related residues was previously observed for an emission-attenuating antibody, supporting that these
interactions induce changes to the photophysical properties of EGFP by disrupting the tripeptidechromo-
phore in a similar manner. Additionally, we have also described the quenching study of the reporter GFP
protein in vivo by a new metal complex using reflected fluorescence microscopy. We anticipate that such
metal complexes which can passively diffuse into the cells in vivo can serve as potential tools in molecular
and drug targeting based biological studies.
Received 11th November 2013,
Accepted 5th December 2013
DOI: 10.1039/c3dt53186d
www.rsc.org/dalton
concerning intercellular regulatory factors and their effects on
protein conformations and subsequent functions. Since
Introduction
Over the last decade, there has been growing interest in the coordination complexes may behave in a similar fashion to
interface of inorganic chemistry and biology with an emphasis these regulatory factors, non-covalent complex–protein inter-
on the non-covalent interactions of coordination complexes actions can be investigated by applying analogous techniques
with proteins.1 Studies of the interactions of metal complexes in order to probe the mechanism of action of a complex which
with proteins are facilitated by the large body of research can involve a variety of pathways including amino acid and
side-chain degradation, redox chemistry of thiol and disulfide
moieties, aggregation, and tertiary structure changes which
aDepartment of Chemistry, University of Ulsan, Ulsan 680-749, Republic of Korea.
can attenuate or halt biological activity.2
E-mail: kwchi@ulsan.ac.kr
Supramolecular coordination complexes (SCCs) are congru-
ous with such applications as they exemplify the synthetic ease
and tunability associated with metal–ligand bonding.3 Formed
by coordination-driven self-assembly, such constructs can be
rapidly generated, affording a number of different 2D and 3D
geometries with nuclearities ranging from two metals for
small rhomboids, to up to twenty-four or more metals for
larger polyhedra.4
Ruthenium-based molecules are of particular interest in
both supramolecular coordination chemistry5,6 as well as
metal complex–protein interactions due to the metal’s control-
lable coordination environment and inherent biological
bDepartment of Chemical and Bioengineering, University of Ulsan, Ulsan 680-749,
Republic of Korea
cSchool of Biotechnology, Yeungnam University, Gyeongsan 712-749, Republic of
Korea
dEnergy Materials and Convergence Research Department, Korea Institute of Energy
Research, Daejeon 305-343, Republic of Korea. E-mail: hyunuk@kier.re.kr
eDepartment of Chemistry, Hanyang University, 17 Haengdang-dong, Seongdong-gu,
Seoul 133 791, Republic of Korea
fDepartment of Pharmaceutical Science and Technology, Catholic University of
Daegu, Hayang-eup, Gyeongsan-Si, Gyeongbuk 712-702, Republic of Korea
†Electronic supplementary information (ESI) available: Spectroscopic figures,
gel electrophoresis, molecular modeling studies. CCDC 899538. For ESI and crys-
tallographic data in CIF or other electronic format see DOI: 10.1039/c3dt53186d
‡These authors contributed equally.
6032 | Dalton Trans., 2014, 43, 6032–6040
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
activity.7 Meggers and coworkers8 have demonstrated that octa-
hedral Ru(II) complexes containing bidentate pyridocarbazole
ligands are noteworthy in the design of novel transition metal-
based drugs wherein they act as inhibitors of protein kinases.
Since such inhibition is an attractive vector for developing new
anticancer agents, these species are highly relevant in drug
development. SCCs utilizing Ru are well-established and typi-
cally employ dinuclear arene–Ru(II) building blocks in which
the two metal centers are bridged by an oxalate-type ligand.
Both 2D and 3D Ru-based SCCs have been investigated in bio-
logical settings; however, the majority of this work has focused
on interactions with DNA and cell viability studies to investi-
gate potential anticancer activities rather than interactions
with other intercellular targets, such as proteins.9
Scheme 1 Synthesis of ligand 1.
Herein, we report the design of a new ditopic ligand as a
building block for SCC formation as demonstrated by a [2 + 2]
self-assembly of an arene–Ru based complex. The ligand is
well-suited to interact with amino acid residues, specifically
through its central pyridyl-2,6-dicarboxamide moiety. The
amide functionalities offer sites for hydrogen bonding without
participating in or disrupting the self-assembly process to
furnish an SCC, as evidenced by early examples of complex for-
mation with amide-based dipyridyl donors.10 Here, the ligand
has been spatially tuned via the inclusion of ethynyl groups
which offer structural rigidity while expanding the conjugated
π system to enhance photophysical properties.
The resulting product was then used under standard Sono-
gashira coupling conditions with the hydrochloride salt of
4-ethynylpyridine in DMF to generate 1 in ca. 50% overall
yield. In the IR spectrum of the product, NH and CO stretching
peaks for the amide linkage were observed at 3366 and
1
1690 cm−1 respectively, while CuC was at 2216 cm−1. The H
NMR showed one broad singlet for the NH at δ 10.18 in nitro-
methane-d3, and the six aromatic resonances were found to
integrate to twelve protons (Fig. S1†).
Synthesis and characterization of metallacycle 1
Green fluorescent protein (GFP) is a widely used biological
marker for gene expression and in situ protein interaction
studies.11 Its photophysical properties are sensitive to confor-
mational changes thus providing a spectral handle which can
be observed even by the naked eye due to visible emission
wavelengths. The enhanced GFP variant (EGFP) was selected
for these complex interaction studies due to its ∼35 fold
increase in emission intensity over wild-type GFP. EGFP is a
barrel-shaped protein which possesses a p-hydroxybenzylidene–
imidazolidone moiety as a chromophore.12 Post-translational
cyclization and oxidation of this sequence produces the
chromophore, making its formation highly dependent on
proper folding and tertiary structure, and rendering it highly
sensitive to agents which can induce conformational changes.13
As a pioneering example of the interaction of Ru macro-
cycles with GFP, we believe that this work provides an impor-
tant foundation in the design of new protein–SCC systems
which will motivate broader targets, including new thera-
peutics, novel biosensors, and strategies for the separation of
peptide mixtures using metal complexes.
The [2 + 2] self-assembly of ligand 1 with Ru-acceptor 2 was
carried out in methanol solution in the presence of AgOTf
(OTf = triflate) which furnished a wedge-shaped tetracationic
complex 3 as a triflate salt along with AgCl. For use in control
experiments, 2 was also combined with 4,4′-bipyridine and
1,3-bis(3-pyridylethynyl)benzene to generate related [2 + 2]
metallacycles 4 and 5 respectively (see Scheme 2) which lacked
the hydrogen bonding.14,15
Results and discussion
Synthesis and characterization of ligand 1
The new ligand 1 was synthesized in two steps. The first start-
ing ligand precursor N,N′-bis(4-iodophenyl)pyridine-2,6-dicar-
boxyl-amide was synthesized by coupling of pyridine-2,6-
dicarbonyl chloride and 4-iodoaniline in CH2Cl2 in the pres-
ence of triethylamine (Scheme 1).
Scheme 2 Synthesis of metallacycles 3–5.
Dalton Trans., 2014, 43, 6032–6040 | 6033
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Dalton Transactions
Fig. 2 (a) The structure of 3 reveals a [2 + 2] assembly. (b) Space filling
model of 3.
Fig. 1 Theoretical (red, top) and experimental (blue, bottom) ESI-MS
regions corresponding to [3 − 2OTf]2+ (left) and [3 − 3OTf]3+ (right)
charge states of 3.
arrangement rather than remaining in-plane as seen for 2D
rhomboid and rectangular assemblies. This conformation
reduces the strain on the Ru coordination sphere associated
with rhomboid formation using a 120° ligand. The distance
between Ru(II) centers is 5.4 × 21.5 Å2, which is the dimension
of the wedged pocket. Each Ru center reflects the coordination
sphere expected from precursor 2, specifically consisting of a
η6p-cymene ligand, two oxygen atoms from the bridging
oxalate ligand and coordination from a terminal pyridyl group
of ligand 1. The average Ru–N and Ru–O distances are 2.08 and
2.12 Å, respectively, which are comparable to those found in
other oxalato-bridged [2 + 2] assemblies.10b The N–Ru–O bite
angles range from 78.1(2)° to 87.8(2)°, indicating a noticeable
distortion from the idealized 90° angles expected in a cis-
capped octahedral coordination sphere. The internal pocket of
3 was not occupied by solvent molecules or triflate counterions.
The structural assignment of 3 as a [2 + 2] complex was
initially indicated by its 1H NMR spectrum in nitromethane-
d3, which revealed chemical shift changes relative to those
observed for 1, the most significant being the large upfield
shifts of the Hα terminal pyridyl protons (Fig. S3†). The ideal-
ized C2v symmetry of 3 results in a highly symmetric core that
1
manifests in a relatively simple aromatic region in the H and
13C NMR spectra (Fig. S3 and S4†).
Electron-spray ionization mass spectrometry (ESI-MS) was
employed to further confirm the [2 + 2] nature of 3. The spec-
trum of 3 showed peaks at m/z = 1126.9 and 768.8 which corre-
spond to [M − 2OTf]2+ and [M − 3OTf]3+, respectively, and
consistent with their theoretical isotopic distributions (Fig. 1).
The electronic absorption spectrum of 3 along with its pre-
cursors reveals that the transitions associated with the free
ligand 1 are largely preserved upon self-assembly (Fig. S5a†).
The dominant spectral features of complex 3 match closely
with those of ligand 1, albeit with minor bathochromic shifts.
As such, the large band centered at 328 nm is assigned to a
ligand-based π→π* transition, with the slight red shift (∼8 nm)
relative to the free ligand likely due to contributions from
metal-to-ligand charge transfers which are introduced upon
Ru coordination. Similar bands are maintained between 3 and
1 at higher energies, which support that these features are also
ligand-centered. While 2 shows minor bands past its shoulder
at ∼200 nm, the low signal of these bands causes them to be
lost amidst the ligand centered transitions found in 3. Despite
the normalized concentrations, the absorbance of 3 is not sig-
nificantly higher than 1, despite containing two ligands per
molecule. This suggests that the individual ligands are not
electronically isolated in 3, which is not unsurprising given
that all the organic fragments present in 3 possess π systems
to facilitate coupling.
Green fluorescent protein binding study
The binding affinity of 3 for EGFP was initially investigated
using electronic absorption spectroscopy. The UV-vis absorp-
tion spectrum of EGFP contains high energy bands at
∼280 nm which have been assigned to tryptophan (Trp) and
tyrosine (Tyr) residues.16 The central Thr65-Tyr66-Gly67 tri-
peptide sequence of the protein represents a chromophore and
provides a relatively low energy absorption band at 488 nm.17
Upon the addition of EGFP to a solution of 3 (Fig. 3a), the
350 nm band of the complex decreased commensurate with an
The strongest evidence for the structural assignment of 3 is
given by a single crystal X-ray diffraction analysis. Crystals suit-
able for structural analysis were grown by the slow diffusion of
diethyl ether into solutions of 3 in nitromethane. The resulting
structure, shown in Fig. 2, reveals that the [2 + 2] assembly of 1
and 2 produces a wedge-shaped complex with approximately
C2v symmetry (Fig. S18†). The two ligands bridge in a cofacial
Fig. 3 (a) UV-vis absorption spectra of 3 (20 µM) upon incremental
addition of EGFP (0–26 µM). (b) Emission spectra of EGFP (2 × 10−6 M)
upon gradual addition of 3 (0–20 µM).
6034 | Dalton Trans., 2014, 43, 6032–6040
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
absorption growth at 500 nm, corresponding to the low-energy other proteins (BSA & ovalbumin). Fig. S13† exhibits no signifi-
band of the protein. The significant hypochromic shift at cant change in the bands suggesting 3 as a selective sensor. To
350 nm and hyperchromic shift at 500 nm are evidence of an further probe the nature of this aggregation, the assay was run
interaction between EGFP and complex 3, as the band at once more using metallacycles 4 and 5 (Scheme 2). Both
350 nm would not otherwise decrease.
control compounds lacked amide protons and carbonyl oxygen
We next investigated the binding of the complexes to EGFP atoms and were expected to thus lack the ability to hydrogen
by a fluorescence spectroscopic method. The emission behav- bond and bind strongly to EGFP. The SDS–PAGE result from
ior of EGFP was significantly affected by the addition of 3; this assay suggested that there was a cleavage in the EGFP
however in the case of 4 and 5 (Scheme 2), no major signifi- bands but no aggregation was observed (Fig. S14 and S15†). As
cant changes were observed (Fig. S8 and S9†). Fig. 3b shows such, the amide functionalities of 3 appear necessary for
the attenuation of the emission band at 512 nm with an exci- protein interaction and subsequent aggregation to occur.
tation wavelength of 488 nm in a Tris-HCl buffer medium. The
Next, we studied the ability of 3 to induce the conformation-
steady decrease in signal intensity upon the addition of 3 indi- al changes in EGFP by circular dichroism. This result also
cates a disruption in either the tertiary or secondary structure strongly supports that 3 destabilizes the EGFP conformation,
of EGFP which renders the tri-peptide chromophore sequence but no major changes were observed on BSA. To study the
unable to form the necessary p-hydroxybenzylidene–imidazoli- specificity of denaturing ability of 3, the same assay was also
dinone moiety which is responsible for emission.
The Stern–Volmer constants (KSV) calculated for 3 were in the conformation of EGFP and BSA (Fig. 5 and S16†).
7.5 × 108 M−1 (r = 0.98) (Fig. S10†). The large binding con-
The ability of 3 to induce aggregation of EGFP was further
stants, which agree well with the Stern–Volmer constants, are investigated using atomic force microscopy (AFM). The images
evidence of strong EGFP binding with 3. clearly show a significant change in the surface morphology of
carried out with 4 and 5 which makes no significant changes
The spectral changes at 512 nm were used to quantify the EGFP after treatment with 3 (Fig. 6). In the absence of 3, the
binding constant of this interaction. The magnitude of the EGFP produces a relatively uniform coverage of the mica
binding constants (K) was calculated to be 7.394 × 108 M−1 surface with molecular spots possessing lateral dimensions of
(Fig. S11†).
10–30 nm. Upon treatment with 3, this uniformity was
Since the photophysical studies of 3 with EGFP indicated removed and the material aggregated to produce globular
that strong interactions were occurring between the complex structures with lateral dimensions of 40–100 nm. This result
and the protein, time-dependent protein cleavage assays were strongly supports that 3 induces aggregation of EGFP which is
employed to further probe the nature of the binding. As shown consistent with the binding interactions revealed by prior
in Fig. 4, the gels exposed to both UV light and Coomassie experiments. A control sample of EGFP treated with 4 does not
blue stain indicate that no cleavage of EGFP was induced by 3 exhibit aggregation and the mica surface remains uniform,
due to the lack of any bands at a lower molecular weight than
the protein; however, bands were present at higher molecular
weights, specifically at 50 and 81 kDa. Analogous assays were
also carried out with precursors 1 and 2, which did not exhibit
any cleavage, fading or smearing, nor were bands observed at
higher molecular weight to suggest aggregation (see Fig. S12†).
To further probe whether complex 3 acts as a selective
sensor for EGFP and to determine the nature of this aggrega-
tion, we carried out experiment involving interaction of 3 with
Fig. 5 Circular dichroism of different proteins in the presence of 3 and
4 (10 mM Tris-HCl buffer, pH 7.0). (a) CD spectrum of 11.2 µM EGFP
upon titration with 3 (0, 1.6, 3.3, 5.0, 6.6 and 8.3 µM). (b) CD spectra of
EGFP (11.2 µM) in the absence and presence of 4 (5.0 and 8.3 µM). (c)
Fig. 4 SDS–PAGE of EGFP (5 µM) incubated at 37 °C with 3 (10 µM) in a
5% DMSO/10 mM Tris-HCl buffer (pH = 7.0) exposed to Coomassie blue
stain (left) and UV light (right).
CD spectra of BSA (8.5 µM) in the absence and presence of 3 (5.0 and
8.3 µM). (d) CD spectra of BSA (8.5 µM) in the absence and presence of
4 (5.0 and 8.3 µM).
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 6032–6040 | 6035

View Article Online
Paper
Dalton Transactions
Fig. 6 Circular AFM images (1 μm × 1 μm) of the mica surfaces: (a)
EGFP alone, (b) EGFP after treatment with 3 and (c) EGFP after treatment
with 4. A drop of 7.5 ng µL−1 aqueous solutions on mica surfaces were
incubated for 5 min and dried under N2 gas stream.
Fig. 7 Close view of the interaction between 3 and EGFP.
further confirming that the unique moieties present in 3 are
critical for these interactions.
In vivo study
Furthermore, the interactions between 3 and EGFP were
predicted using molecular modeling studies. As the crystal
structure of our EGFP variant was not available, we attempted
to develop a 3D model structure for the EGFP based on
related, known structures. A template search for our variant
using BlastP revealed a GFP structure that differed by only a
single amino acid in the chromophore (Table 1 in ESI†). In
addition, we also analyzed two templates of EGFP co-crystal-
lized with two specific nanobodies. A recent report by Kirchho-
fer describes two camelid-derived single-domain antibodies
(nanobodies), one of which induced a 5-fold decrease in emis-
sion intensity, the other effecting a 10-fold increase.18 The
attenuating antibody, dubbed the minimizer in contrast to its
enhancer counterpart, was found to interact with the Arg168
residue of the protein which is proximal to the chromophore
tri-peptide sequence.
Synthetic molecules can serve as a potential tool in the mole-
cular and drug targeting based biological studies. One of the
innumerable applications offered by such synthetically designed
complexes is that they can passively diffuse into the cells and
deactivate the target bio-molecules at the cellular level. To inves-
tigate the interference of complex 3 with the fluorescent reporter
protein in vivo, reflected fluorescence microscopy was carried
out. Fluorescence induced E. coli cells harboring pET21a-GFP
were incubated with 1 mM of complex 3 at different time inter-
vals. In relation to in vitro studies, quenching of the EGFP was
also observed in vivo from 1 to 6 h incubation of 3 (Fig. 8). The
This interaction shifted the His148 residue away from the
chromophore, causing Arg168 to lose its interaction with
Asn146, the ultimate effect being an overall conformational
change sufficient to affect emission. One established strategy
for predicting the interactions between a novel inhibitor and a
protein is to make comparisons with the mechanism of a
known inhibitor of that same protein.19 As such, our GFP
variant was modeled using both the minimizer and enhancer
nanobody-bound EGFP structures and docking studies were
carried out with models of 3. The computational results indi-
cated that 3 gave better docking models when interacting with
the minimizer-derived conformation of EGFP, shown in Fig. 7
and S17.† Since 3 clearly stabilizes a fluorescence-attenuated
confirmation of the protein relative to the native structure and
emission-enhanced structure, it follows that the complex may
act by inducing similar conformational changes, leading to a
quenching effect.
Fig. 8 Reflected fluorescence microscopy images of an E. coli BL21
carrying pET21a-EGFP incubated in the presence and absence (a) of
complex 3 (1 mM) after induction with 1 mM of IPTG for the identifi-
cation of cellular uptake for complex 3 into these bacterial strains after
1 h (b), 2 h (c) and 6 h (d) of incubation. (e) Insets give DIC image of
complex 3 incubated for 6 h with bacterial strain.
6036 | Dalton Trans., 2014, 43, 6032–6040
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
fluorescence emitted by the GFP protein was found to be com- The host bacterium Escherichia coli (E. coli) strain XL1-blue
pletely quenched after 6 h in the incubation of the bacterial (Stratagene, CA, USA) was used for plasmid DNA preparation
cells with complex 3. Quenching of the EGFP within the cellu- in this study. The nickel nitrilotriacetic acid (Ni-NTA)
lar cytosol further indicates the membrane permeable trait of affinity column was purchased from Qiagen (Valencia, CA,
complex 3. In this respect, complex 3 mediated deactivation of USA). The egfp gene was amplified from the plasmid
the target biomolecule might become an attractive source for pROBE-NT′ and subsequently cloned into pET21a using
future studies with relevance to drug delivery, targeting and restriction enzyme. The construct pET21a-EGFP was trans-
biosensing through complex–protein interactions.
formed into E. coli BL21.
X-ray crystallography for 3
Conclusions
The diffraction data from a single crystal of 3 mounted on a
loop were collected at 100 K on an ADSC Quantum 210 CCD
diffractometer with synchrotron radiation (λ = 0.80000 Å) at
the Macromolecular Crystallography Beamline 6B1, Pohang
Accelerator Laboratory (PAL), Pohang, Korea. The raw data
were processed and scaled using the program HKL2000. The
structure was solved by direct methods, and the refinements
were carried out with full-matrix least-squares on F2 with
appropriate software implemented in SHELXTL software suite.
All the non-hydrogen atoms were refined anisotropically.
Hydrogen atoms were added to their geometrically ideal posi-
tions except solvent molecules (acetone and methanol). The
crystallographic data are summarized in Table S2.† CCDC
899538 contains the supplementary crystallographic data for
this paper.
In conclusion, we report here the first studies of protein inter-
actions of Ru-based SCCs using a novel [2 + 2] assembly
encoded with moieties capable of non-covalent interactions
with amino acid residues. A high-yielding self-assembly furn-
ished 3, as confirmed by NMR, ESI-MS and X-ray crystallo-
graphy. The roughly C2v symmetric, wedge-shaped SCC 3 was
found to strongly bind to EGFP, altering the photophysical pro-
perties of the protein, as probed by steady-state UV-vis and
emission binding experiments. This binding was further
probed by gel electrophoresis, circular dichroism and AFM
studies which indicate that EGFP undergoes aggregation in the
presence of 3, but not its precursors, nor a related SCC which
lacked hydrogen bonding functionalities. Molecular modeling
validated the proposed quenching mechanism by revealing
that 3 likely acts in a similar manner to a known emission-
attenuating antibody, disrupting the chromophore of EGFP via
conformational changes. Additionally, we also described the
UV-Vis binding study
Absorption spectra were recorded using a Cary 100 Conc
quenching study of the reporter GFP protein in vivo by a new UV-Visible spectrophotometer. The spectral changes were
metal complex, using reflected fluorescence microscopy. observed upon the addition of 0–26 μM EGFP to a 20 μM solu-
Studies to further elaborate the mechanism for protein tion of 3 and addition of a 0–32 μM solution of 3 to a 20 μM
binding and potential selectivity of the complex with other EGFP in 10 mM of Tris-HCl solution.
GFP variants are underway.
Fluorescence binding study
Fluorescence spectra of control and test proteins were recorded
on a RF-5301PC spectrofluorimeter (Shimadzu, Japan). The
Materials and methods
emission behavior of EGFP was significantly affected by the
All chemicals used in this work were purchased from commer-
addition of 3. An excitation wavelength of 488 nm shows the
cial sources and used without further purification. All solvents
attenuation of the emission band at 512 nm with in a Tris-HCl
used were distilled by standard methods. Precursors 2 and 4
buffer medium. The steady decrease in signal intensity upon
were prepared according to the literature.14 The 1H and 13C
the addition of 3 indicates a disruption in either the tertiary or
NMR spectra were recorded on a Bruker 300 MHz spectro-
secondary structure of EGFP which renders the tri-peptide
meter. The chemical shifts (δ) in the 1H NMR spectra were
chromophore sequence unable to form the necessary
reported relative to tetramethylsilane (Me4Si) as an internal
p-hydroxybenzylidene–imidazolidinone moiety which is
standard (0.0 ppm). Mass spectra were recorded on a Micro-
responsible for emission. The spectral changes which occur
mass Quattro II triple-quadrupole mass spectrometer using
when 3 is present in a solution of EGFP suggest that the
electrospray ionization (ESI) and processed using the Mas-
complex could be used as a sensor for this and related pro-
sLynx software suite. Elemental analyses were performed using
teins. Generally, a linear Stern–Volmer (SV) plot was used to a
a
Perkin-Elmer CHN analyzer. Absorption spectra were
1 : 1 binding model for fluorescence quenching to get a linear
relationship. In order to gain insight into the quenching
mechanism, the fluorescence quenching data for complex 3
were analysed by using the Stern–Volmer equation.
recorded using a Cary 100 Conc UV-Visible spectrophotometer.
Fluorescence titration studies were carried out on a RF-5301PC
spectrofluorimeter (Shimadzu, Japan). PCR reagents, T4 DNA
ligase and restriction endonucleases were purchased from
Promega (Madison, WI, USA). The isopropyl-D-thiogalactopyra-
noside (IPTG) was purchased from Sigma (St. Louis, MO, USA).
I0
¼ 1 þ Ksv½3ꢀ
ð1Þ
I
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 6032–6040 | 6037

View Article Online
Paper
Dalton Transactions
where I0 and I are the fluorescence intensities of EGFP in the Molecular modeling studies
absence and in the presence of the quencher, [3] is the concen-
tration of the quencher and Ksv is the Stern–Volmer constant.
Homology modeling. The protein sequence of the EGFP
The displayed Stern–Volmer plots showed that within the variant was translated from the nucleotide sequence using the
investigated concentrations, the results agreed with the Stern– online tool SMS2. A template search for the EGFP variant
Volmer equation.
sequence was carried out using BlastP against the PDB data-
The fluorescence quenching data obtained from the inter- base.21 From the Blast result, the EGFP protein structures
action of complex 3 with EGFP were further analyzed in order which had co-crystallized with antibodies were chosen based
to obtain binding parameters by using the equation
on how well they matched the EGFP variant used here. Models
using both the minimizer and enhancer-bound EGFP were
developed and validated using the SAVES server.
ꢀ
ꢁ
I0 ꢁ I
log
¼ log K þ n log½3ꢀ
ð2Þ
I
Molecular docking. The validated model structures were
docked with 3 using GOLD4.1.2. Prior to docking calculations,
all hydrogen atoms and charges were added to the protein
using Discovery studio2.1. The grid size was set to 10, centered
on the XYZ co-ordinate of Arg168. The Lamarckian genetic algor-
ithm was employed as a search parameter. Fifty genetic algor-
ithms were run for the docked models. From the fifty solutions
thus obtained for each model, the best solution was selected
and exported in PDB format and analyzed using Pymol.22
Circular dichroism. CD measurements were performed on a
JASCO J-715 instrument using quartz 0.1 cm cuvettes. The
machine settings were as follows: J-715 scan settings; 600 nm–
190 nm scan range, 0.2 nm pitch, continuous scanning at
100 nm min−1, 100 mdeg sensitivity, 4 s response time, 1.0 nm
band width, single scan accumulation.
where I0 and I are the fluorescence intensity in the absence
and in the presence of complex 3. Plots of log[(I0 − I)/I] versus
log [3] gave a straight line whose slope provided n, the number
of binding sites, while the intercept allowed the calculation of
the values of apparent binding constant, K, specifying the
number of 3 bound to a EGFP macromolecule, whose slope
was equal to n and the intercept on the y axis to log K. The
value of n is approximately equal to 1 which indicated the
existence of just a single binding site in EGFP for 3. The value
of binding constant K, 7.394 × 108 M−1, showed strong inter-
action between EGFP and 3.20
Protein purification
Preparation of N,N′-bis(4-iodophenyl)pyridine-2,6-dicarbox-
amide. A solution of 2,6-pyridinedicarbonyl dichloride (1.0 g,
4.90 mmol) and triethylamine (3 mL) in 30 mL of CH2Cl2 was
prepared and chilled to 4 °C in an ice-bath for 5 min, after
which 4-iodo aniline (2.25 g, 10.30) was added slowly over a
period of 10 min. The reaction mixture was stirred at room
temperature overnight. The resulting precipitate was collected
on a frit, recrystallized with MeOH–H2O, and dried in a
vacuum oven. Yield, ca. 90% (2.50 g). M.P. = 273 °C, MS (m/z):
calcd = 569.13, found = 569.16. Anal. Calcd for C19H13I2N3O2:
C, 40.10%; H, 2.30%; N, 7.38%; Found: C, 40.08%; H, 2.22%;
The above mentioned construct was cultivated in a Luria–
Bertani medium supplemented with ampicillin at 37 °C with
vigorous shaking. When the optical density of the culture
reached 0.8 to 1.0 at 600 nm, 1 mM of IPTG was added to
induce protein expression. OD600 was then measured and the
cells were harvested after 6 h of induction. Harvested cells
were stored at −70 °C until used. Cell lysis was carried out
using a Bug Buster protein extraction kit (Novagen) followed
by sonication. Cell pellets corresponding to 1 mL of culture
were resuspended in 100 μl of lysis buffer, incubated at room
temperature for 10 min, and centrifuged at 9000g at 4 °C for
20 min. The supernatant was saved as a soluble protein frac-
tion, and analyzed by SDS–PAGE (10% acrylamide gel). The
recombinant proteins were purified from the soluble protein
fraction by using an Ni-NTA affinity column. His-tag purified
samples were subjected to gel permeation chromatography
(GPC) by an AKTA Explorer FPLC system containing a Superdex
75 HR column at 4 °C. The concentration of the protein was
quantified using the Bradford assay.
1
N, 7.50%; H NMR (d6-DMSO, 300 MHz, δ, ppm): 11.07 (s, 2H,
CONH), 8.40 (d, 2H, J = 7.5 Hz, Hb), 8.30 (t, 1H, Ha), 7.79–78
(m, 8H; Hd/Hc).
Synthesis of N,N′-bis(4-(pyridin-4-ylethynyl)phenyl)pyridine-
2,6-dicarboxamide 1. In a flame-dried Schlenk flask, 4-ethynyl-
pyridine hydrochloride (0.590 g, 4.2 mmol) and triethylamine
(6 mL) were vigorously stirred for 20 min. During this time a
white precipitate formed. DMF (30 mL), Pd(PPh3)2Cl2 (0.050 g,
0.07 mmol), PPh3(0.018 g, 0.07), CuI (0.014 g, 0.07 mmol), and
N,N′-bis(4-iodophenyl)pyridine-2,6-dicarboxamide (0.800 g,
1.40 mmol) were successively added. The Schlenk flask was
Time dependent protein binding assays using SDS–PAGE
A 5 µM solution of EGFP in Tris-HCl buffer medium (pH 7.0) wrapped in aluminum foil to keep the reaction mixture in the
was incubated at 37 °C with 10 µM of 3 at time intervals dark and the solution was refluxed for 24 h. The resulting
ranging from 15 to 60 min. Aliquots of the reaction mixtures brown mixture was filtered on frit, and recrystallized with
were subjected to modified 10% SDS–PAGE (without thermal CH3OH–H2O (1 : 1). The compound was filtered and dried
denaturation). The gel was subsequently rinsed with distilled under vacuum. Yield, ca. 55% (0.400 g). M.P. = 297 °C, MS
water and exposed to a UV transilluminator to detect fluo- (m/z): calcd = 519.55, found = 519.56. Anal. Calcd for
rescence bands. The gels were then stained with Coomassie C33H21N5O2: C, 76.29; H, 4.07; N, 13.48; found: C, 76.32; H,
blue stain.
4.11; N, 13.49. IR (KBr, ν, selected peaks): 3366 (NH), 2216
6038 | Dalton Trans., 2014, 43, 6032–6040
This journal is © The Royal Society of Chemistry 2014

View Article Online
Dalton Transactions
Paper
(CuC), and 1690 (CvO) cm−1; λmax/nm = CH3OH (ε, dm3
(b) A. Vergara, G. D’Errico, D. Montesarchio, G. Mangiapia,
L. Paduano and A. Merlino, Inorg. Chem., 2013, 52, 4157;
(c) K. J. Kilpin and P. J. Dyson, Chem. Sci., 2013, 4, 1410;
(d) M. Patra and G. Gasser, ChemBioChem, 2012, 13, 1232;
(e) E. Meggers, Chem. Commun., 2009, 1001;
(f) N. P. E. Barry and P. J. Sadler, Chem. Commun., 2013, 49,
5106; (g) T. Reiner, D. Jantke, A. N. Marziale, A. Raba and
J. Eppinger, Chem. Open, 2013, 2, 50.
2 (a) N. P. E. Barry and P. J. Sadler, Chem. Soc. Rev., 2012, 41,
3264; (b) D. J. Clymer, C. R. Geren and K. E. Ebner, Bio-
chemistry, 1976, 15, 1093; (c) E. A. Permyakov,
S. E. Permyakov, G. Y. Deikus, L. A. Morozova-Roche,
V. M. Grishchenko, L. P. Kalinichenko and V. N. Uversky,
Proteins, 2003, 51, 498.
3 (a) C. K. McLaughlin, G. D. Hamblin and H. F. Sleiman,
Chem. Soc. Rev., 2011, 40, 5647; (b) R. Kieltyka,
P. Englebienne, J. Fakhoury, C. Autexier, N. Moitessier and
H. F. Sleiman, J. Am. Chem. Soc., 2008, 130, 10040;
(c) S. Shanmugaraju, A. K. Bar and P. S. Mukherjee, Organo-
metallics, 2010, 29, 2971; (d) A. K. Bar, S. Shanmugaraju,
K.-W. Chi and P. S. Mukherjee, Dalton Trans., 2011, 40,
2257; (e) S. Shanmugaraju, S. A. Joshi and P. S. Mukherjee,
Inorg. Chem., 2011, 50, 11736.
4 (a) R. Chakrabarty, P. S. Mukherjee and P. J. Stang, Chem.
Rev., 2011, 111, 6810; (b) Y.-F. Han, W.-G. Jia, W.-B. Yu and
G.-X. Jin, Chem. Soc. Rev., 2009, 38, 3419; (c) Y.-F. Han,
Y.-J. Lin, W.-G. Jia and G.-X. Jin, Organometallics, 2008, 27,
4088; (d) Y.-F. Han, Y.-J. Lin, W.-G. Jia, L.-H. Weng and
G.-X. Jin, Organometallics, 2007, 26, 5848; (e) Y.-F. Han,
Y.-J. Lin, W.-G. Jia, G.-L. Wang and G.-X. Jin, Chem.
Commun., 2008, 1807; (f) Y.-F. Han, H. Li and G.-X. Jin,
Chem. Commun., 2010, 46, 6879; (g) Y.-F. Han, W.-G. Jia,
Y.-J. Lin and G.-X. Jin, Angew. Chem., Int. Ed., 2009, 48,
6234; (h) Y.-F. Han, Y. Fei and G.-X. Jin, Dalton Trans., 2010,
39, 3976.
1
mol−1 cm−1): 320 (1500). H NMR (nitromethane-d3, 300 MHz,
δ, ppm): 10.18 (s, 2H, CONH), 8.61 (d, 4H, J = 6 Hz, Ha), 8.52
(d, 2H, J = 8.1 Hz, He), 8.34–8.29 (t, 1H, Hf), 8.02 (d, 4H, J =
6 Hz, Hb), 7.70 (d, 4H, J = 6.6 Hz, Hd), 7.48 (d, 4H, J = 5.7 Hz,
Hc); 13C NMR (DMSO-d6, 300 MHz, δ, ppm): 161.87(CvO, C10),
149.99(C11), 148.59(C13), 140.23(C13), 139.30(C9), 132.60(C7),
130.35(C3), 125.73(C2), 125.31(C12), 120.78(C8), 116.61(C6),
93.79(C5), 86.51(C4).
Synthesis of metallacycle 3
A mixture of 2 (50 mg, 0.08 mmol) and 2 equiv. of AgCF3SO3
(42 mg, 0.16 mmol) in methanol was stirred at room tempera-
ture for 2 h and filtered to remove AgCl. Ligand 1 (42 mg,
0.08 mmol) was added to the filtrate. The mixture was then
stirred at room temperature for 12 h, and the solvent was
removed under reduced pressure. The crude product thus
obtained was redissolved in nitromethane and subjected to
vapour diffusion of diethyl ether. This resulted in the crystalli-
zation of the product over the course of a week. The product
was filtered and dried under vacuum. Yield: 100 mg (91%).
Anal. Calcd for C114H98F12N10O24Ru4S4: C, 49.74; H, 3.59; N,
5.09; S, 4.66; found: C, 49.80; H, 3.72; N, 5.12; S, 4.35. FT-IR
spectrum (KBR, ν, selected peaks): 3340(NH), 2245 (CuC),
1620 (CvO) cm−1. MS (ESI) calcd for [M − 2OTf]2+ m/z 1226.6,
found 1226.9; calcd for [M − 3OTf]3+ m/z 768.8, found 768.8.
Absorption spectrum [λmax/nm, CH3OH (ε, M−1 cm−1): 328
1
(1600). H NMR (nitromethane-d3, 300 MHz, δ, ppm): 9.79 (s,
4H, CONH), 8.14 (d, 8H, J = 9.0, Ha), 8.01 (d, 8H, J = 6.6, He),
7.95 (t, 4H, Hf), 7.72 (d, 8H, J = 8.7, Hd), 7.42–7.37 (m, 16H, Hc/
Hb), 5.89 (d, 8H, J = 8.7, –C6H4), 5.72 (d, 8H, J = 8.7, –C6H4),
2.90 (m, 4H, –CH(CH3)2), 2.07 (s, 12H, –CH3), 1.39 (d, 12H,
–CH(CH3)2). 13C NMR (DMSO-d6, 300 MHz, δ, ppm): 162.22
(CvO), 153.37, 149.53, 141.10, 140.76, 136.86, 134.68, 128.71,
120.69, 117.71, 104.23, 101.46, 86.51, 83.38, 82.97(CvC),
32.48, 22.45, 18.31, 15.66.
5 A. Mishra, S. C. Kang and K.-W. Chi, Eur. J. Inorg. Chem.,
2013, 5222.
6 (a) V. Vajpayee, Y. H. Song, T. R. Cook, H. Kim, Y. Lee,
P. J. Stang and K.-W. Chi, J. Am. Chem. Soc., 2011, 133,
19646; (b) K. Severin, Chem. Commun., 2006, 3859;
(c) B. Therrien, G. Süss-Fink, P. Govindaswamy,
A. K. Renfrew and P. J. Dyson, Angew. Chem., Int. Ed., 2008,
47, 3773; (d) A. Mishra, H. Jung, J. W. Park, H. K. Kim,
H. Kim, P. J. Stang and K.-W. Chi, Organometallics, 2012,
31, 3519; (e) A. Mishra, S. Lee, H. Kim, T. R. Cook,
P. J. Stang and K.-W. Chi, Chem. Asian J., 2012, 7, 2592;
(f) B. Kilbas, S. Mirtschin, T. R.-Johannessen, R. Scopelliti
and K. Severin, Inorg. Chem., 2012, 51, 5795; (g) A. Pitto-
Barry, N. P. E. Barry, O. Zava, R. Deschenaux, P. J. Dyson
and B. Therrien, Chem.–Eur. J., 2011, 17, 1966;
(h) S. Shanmugaraju, A. K. Bar and P. S. Mukherjee, Inorg.
Chem., 2010, 49, 10235; (i) B. Kilbas, S. Mirtschin, T. Riis-
Johannessen, R. Scopelliti and K. Severin, Inorg. Chem.,
2012, 51, 5795; ( j) F. Linares, M. A. Galindo, S. Galli,
M. A. Romero, J. A. R. Navarro and E. Barea, Inorg. Chem.,
2009, 48, 7413.
Acknowledgements
This work was supported by the Basic Science Research
program through the National Research Foundation of Korea
(NRF) funded by the Ministry of Science, ICT and Future Plan-
ning (NRF-2013R1A1A2006859). The Priority Research Centers
program (2009-0093818) through the NRF is also appreciated
for financial support. X-ray diffraction experiments using syn-
chrotron radiation were performed at the Pohang Accelerator
Laboratory in Korea. We thank Prof. Seok-Kyu Kim and
Dr Yoon Jung Jang, Department of Chemistry, Yeungnam
University, for providing the circular dichroism facility.
Notes and references
1 (a) A. Vergara, I. R. Krauss, D. Montesarchio, L. Paduano
and A. Merlino, Inorg. Chem., 2013, 52, 10714;
This journal is © The Royal Society of Chemistry 2014
Dalton Trans., 2014, 43, 6032–6040 | 6039

View Article Online
Paper
Dalton Transactions
7 (a) H. Takashima, S. Shinkai and I. Hamachi, Chem. 11 (a) D. Groff, F. Wang, S. Jockusch, N. J. Turro and
Commun., 1999, 2345; (b) J. Ohkanda, R. Satohb and N. Katoa,
Chem. Commun., 2009, 6949; (c) M. H. Filby, J. Muldoon,
S. Dabb, N. C. Fletcher, A. E. Ashcroft and A. J. Wilson, Chem.
Commun., 2011, 47, 559; (d) J. Muldoon, A. E. Ashcroft and
P. G. Schultz, Angew. Chem., Int. Ed., 2010, 49, 7677;
(b) G. U. Nienhaus, Angew. Chem., Int. Ed., 2008, 47, 8992;
(c) M. Swierczewska, G. Liu, S. Lee and X. Chen, Chem. Soc.
Rev., 2012, 41, 2641.
A. J. Wilson, Chem.–Eur. J., 2010, 16, 100; (e) A. J. Wilson, 12 (a) G. H. Patterson, S. M. Knobel, W. D. Sharif, S. R. Kain
J. R. Ault, M. H. Filby, H. I. A. Philips, A. E. Ashcroft and
N. C. Fletcher, Org. Biomol. Chem., 2013, 11, 2206.
8 (a) S. Blanck, J. Maksimoska, J. Baumeister, K. Harms,
and D. W. Piston, Biophys. J., 1997, 73, 2782; (b) J. Zhang,
R. E. Campbell, A. Y. Ting and R. Y. Tsien, Nat. Rev. Mol.
Cell Biol., 2002, 3, 906.
R. Marmorstein and E. Meggers, Angew. Chem., Int. Ed., 13 J. K. Sanders and S. E. Jackson, Chem. Soc. Rev., 2009, 38,
2012, 51, 5244; (b) J. Maksimoska, L. Feng, K. Harms, C. Yi, 2821.
J. Kissil, R. Marmorstein and E. Meggers, J. Am. Chem. Soc., 14 H. Yan, G. Suss-Fink, A. Neels and H. Stoeckli-Evans,
2008, 130, 15764; (c) E. Meggers, Chem. Commun., 2009,
1001.
9 (a) M. J. Hannon, Chem. Soc. Rev., 2007, 36, 280;
J. Chem. Soc., Dalton Trans., 1997, 4345.
15 S. Shanmugaraju, A. K. Bar, S. A. Joshi, Y. P. Patil and
P. S. Mukherjee, Organometallics, 2011, 30, 1951.
(b) G. I. Pascu, A. C. G. Hotze, C. S.-Cano, B. M. Kariuki 16 D. B. Wetlaufer, Adv. Protein Chem., 1962, 17, 303.
and M. J. Hannon, Angew. Chem., Int. Ed., 2007, 46, 4374; 17 R. Heim and R. Y. Tsien, Curr. Biol., 1996, 6, 178.
(c) N. P. E. Barry, N. H. Abd Karim, R. Vilarand and 18 A. Kirchhofer, J. Helma, K. Schmidthals, C. Frauer,
B. Therrien, Dalton Trans., 2009, 10717.
S. Cui, A. Karcher, M. Pellis, S. Muyldermans,
C. S. Casas-Delucchi, M. C. Cardoso, H. Leonhardt,
K. P. Hopfner and U. Rothbauer, Nat. Struct. Mol. Biol.,
2010, 17, 133.
10 (a) A. Mishra, S. Ravikumar, S. H. Hong, H. Kim,
V. Vajpayee, H. W. Lee, B. C. Ahn, M. Wang,
P. J. Stang and K.-W. Chi, Organometallics, 2011, 30,
6343; (b) A. Mishra, V. Vajpayee, H. Kim, M. H. Lee, 19 M. H. Kubala, O. Kovtun, K. Alexandrov and B. M. Collins,
H. Jung, M. Wang, P. J. Stang and K.-W. Chi, Dalton Protein Sci., 2010, 19, 2389.
Trans., 2012, 41, 1195; (c) A. Mishra, H. Kim, 20 A. T. Poulos, V. Kuzmin and N. E. Geacintov, J. Biochem.
S. C. Lee, J. W. Min, M. H. Lee and K.-W. Chi, Inorg. Biophys. Methods, 1982, 6, 269.
Chim. Acta, 2013, 405, 77; (d) A. Mishra, H. Jung, 21 S. McGinnis and T. L. Madden, Nucleic Acids Res., 2004, 32,
M. H. Lee, M. S. Lah and K.-W. Chi, Inorg. Chem.,
2013, 52, 8573.
W20.
22 Protein Preparation Guide, Schrodinger, LLC, 2010.
6040 | Dalton Trans., 2014, 43, 6032–6040
This journal is © The Royal Society of Chemistry 2014