Detection of nitric oxide and nitroxyl with benzoresorufin-based fluorescent sensors

Article abstract of DOI:10.1021/ic302793w

A new family of benzoresorufin-based copper complexes for fluorescence detection of NO and HNO is reported. The copper complexes, CuBRNO1-3, elicit 1.5-4.8-fold emission enhancement in response to NO and HNO. The three sensors differ in the nature of the metal-binding site. The photophysical properties of these sensors are investigated with assistance from density functional theory calculations. The fluorescence turn-on observed upon reaction with HNO is an unexpected result that is discussed in detail. The utility of the new sensors for detecting HNO and NO in HeLa cells and RAW 264.7 macrophages is demonstrated.

Full text of DOI:10.1021/ic302793w

Article
pubs.acs.org/IC
Detection of Nitric Oxide and Nitroxyl with Benzoresorufin-Based
Fluorescent Sensors
Ulf-Peter Apfel,† Daniela Buccella,†,‡ Justin J. Wilson,† and Stephen J. Lippard*,†
†Department of Chemistry, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, United States
‡Department of Chemistry, New York University, New York, New York 10003, United States
S
* Supporting Information
ABSTRACT: A new family of benzoresorufin-based copper
complexes for fluorescence detection of NO and HNO is
reported. The copper complexes, CuBRNO1−3, elicit 1.5−4.8-
fold emission enhancement in response to NO and HNO. The
three sensors differ in the nature of the metal-binding site. The
photophysical properties of these sensors are investigated with
assistance from density functional theory calculations. The
fluorescence turn-on observed upon reaction with HNO is an
unexpected result that is discussed in detail. The utility of the
new sensors for detecting HNO and NO in HeLa cells and RAW 264.7 macrophages is demonstrated.
Scheme 1. Small-Molecule Metal-Containing Sensors for NO
(Cu[FL1]) and HNO (Cu[BOT1]) as Well as Their
Reactions with Analytes30,33
INTRODUCTION
■
Reactive nitrogen (RNS) and oxygen species (ROS) have
numerous biological consequences.1 Among these species,
nitric oxide (NO), initially identified as an endothelial-derived
relaxation factor,2 has a broad variety of biological regulatory
and signaling functions.3−8 Biological NO is generated by
oxidation of L-arginine to L-citruline by a class of enzymes
known as nitric oxide synthases (NOS).9 Among its many
functions, NO plays important roles in the control of smooth
muscle relaxation and vasodilation,10 platelet aggregation in
vascular endothelial cells,11 neurotransmission,12 and regulation
of the immune response by macrophages.13 Recently, it was
suggested that nitroxyl (HNO), the one-electron-reduced
congener of NO, is formed by NOS via oxidative degradation
of L-arginine.14,15 Some recent in vitro studies using HNO-
releasing molecules demonstrated that HNO increases the
contractility of heart cells,16 leads to vasorelaxation in muscle
cells,17 and decreases platelet aggregation.18 Taken together,
these findings suggest that HNO also plays a pivotal role in
biology.
In order to advance our insight into the physiological and
pathological roles of HNO and NO, our group has focused on
designing fluorescent sensors that selectively respond to these
small molecules and afford both spatial and temporal
information regarding the natural occurrence of these species
at the cellular level. Various NO sensors, including o-
diaminofluorescein, o-diaminonaphthalene, o-diaminocya-
nine,19−21 luminescent lanthanide complexes,22 and 5-amino-
1-naphthonitrile (NO550),23 detect NO in the presence of
oxygen. Sensors for HNO include metalloporphyrins,24 thiols,25
and phosphines.26 In contrast to these reagents, copper-based
fluorescent probes like CuFL1,27,28 CuFL2E,29 and CuSNFL30
developed by our group (Scheme 1) and others31,32 provide
direct, selective, and fast detection of NO both in vitro and in
vivo. We have also recently described a selective HNO sensor,
CuBOT1.33,34 The emission intensity of these copper-based
sensors is modulated by the oxidation state of the copper ion,
which, in turn, can be modified by the RNS of interest. The
CuFL and CuSNFL sensors react with NO to form the
emissive nitrosamines FL-NO and SNFL-NO, with concom-
itant formation of CuI ions that dissociate from the
complex.30,35 The copper complex CuBOT1, on the other
hand, reacts selectively with HNO, leading to reduction of the
paramagnetic CuII ion and fluorescence enhancement.33 The
different reactivity of CuBOT1 versus CuFL and CuSNFL
presumably arises from the lack of a secondary amine in the
Received: December 19, 2012
Published: March 5, 2013
© 2013 American Chemical Society
3285
dx.doi.org/10.1021/ic302793w | Inorg. Chem. 2013, 52, 3285−3294

Inorganic Chemistry
Article
former, which is the preferred site of attack on CuFL by NO.35
A limitation of both fluorescein- and BODIPY-based sensors is
their high-energy absorption and emission as well as their
moderate Stokes shifts of about 30 nm. These features can
cause problems associated with light scattering, which may lead
to a low signal-to-noise ratio. Additionally, for biological
imaging purposes, it is desirable to have sensors that emit far
into the red, because low-energy radiation penetrates tissue
more effectively.
In our continuing efforts to design improved sensors for NO
and HNO, we have functionalized a benzoresorufin fluorophore
with a Cu-binding site that contains a secondary amine, and we
investigated its ability to detect ROS and RNS. Resorufin dyes
emit at wavelengths >600 nm with Stokes shifts up to 60 nm.36
Accordingly, sensors based on resorufin dyes should exhibit less
background absorption and emission from biological samples
and provide an advantage over previously reported green
emitters in terms of tissue penetration depth. The synthesis and
characterization of three benzoresorufin-based sensors are
described herein. As with our previous systems, the CuII/CuI
redox couple is used to modulate the emission response of the
sensors to NO and HNO. Despite the availability of a
secondary amine, a potential site of attack of NO, we found
that the benzoresorufin-based probes show a better turn-on
response for HNO than for NO.
The solid was dried under vacuum, yielding 1.06 g (65%) of an orange
1
crystalline material. H NMR (CDCl3, 300 MHz, ppm): δ 8.71−8.67
(1H, m), 8.30−8.27 (1H, m), 7.78−7.73 (2H, m), 7.68 (1H, d, 3J = 9
Hz), 7.07 (1H, d, 3J = 9 Hz), 6.46 (1H, s), 2.39 (3H, s), 2.27 (3H, s).
13C NMR (CDCl3, 75 MHz, ppm): δ 184.1, 169.0, 151.2, 149.8, 132.4,
132.3, 132.0, 131.4, 130.9, 130.2, 127.8, 126.1, 124.9, 119.2, 118.8,
117.1, 107.7, 21.0, 9.1. ESI-MS. Calcd for [C19H13NO4 + H+]+ 320.1.
Found: 320.2. Mp: 234−235 °C.
6-Bromo-8-bromomethyl-5-oxo-benzo[a]phenoxazin-9-yl
acetate (3). Compound 2 (2.62 g, 8.2 mmol), 1,3-dibromo-5,5′-
dimethylhydantoin (5.16 g, 0.018 mol), and VAZO88 (0.72 g, 3.0
mmol) were dissolved in dry chlorobenzene (300 mL) under an N2
atmosphere. Acetic acid (700 μL) was added, and the solution was
1
heated at 60 °C until H NMR spectroscopy revealed full conversion
to the final molecule (6−10 days). The hot solution was then washed
with hot water (60 °C, 2 × 50 mL) and brine (100 mL) and dried with
sodium sulfate. Evaporation of the organic fractions to dryness
afforded an orange solid, which was washed with diethyl ether/pentane
(1:1). The solid was collected by filtration and dried under vacuum to
1
afford 3.6 g (92%) of an orange solid. H NMR (CDCl3, 300 MHz,
ppm): δ 8.71−8.69 (1H, m), 8.38−8.35 (1H, m), 7.86−7.76 (3H, m),
7.26−7.23 (1H, m), 4.73 (2H, s), 2.46 (3H, s). ESI-MS. Calcd for
[C19H11Br2NO4 + Na+]+: 497.9. Found: 497.8. Mp: 221 °C.
6-Bromo-8-formyl-9-hydroxy-5-benzo[a]phenoxazone (4).
Compound 3 (1.5 g, 3.16 mmol) and sodium bicarbonate (1.8 g, 21
mmol) were dissolved in dimethyl sulfoxide (DMSO; 50 mL) and
heated at 150 °C for 2 h. After cooling to room temperature, the blue
solution was poured into 300 mL of HCl (4 M), and the resulting
mixture was stirred for 1 h. The obtained brown precipitate was
collected by filtration, washed with water, and dried under vacuum.
The crude material was purified by silica gel column chromatography
(dichloromethane to dichloromethane/methanol 100:2). The ob-
tained residue was washed with 20 mL of diethyl ether/pentane (1:1)
to afford 382 mg (29%) of a dark-orange solid. 1H NMR (CDCl3, 300
MHz, ppm): δ 12.10 (1H, s), 10.71 (1H, s), 8.70−8.67 (1H, m),
8.40−8.37 (1H, m), 8.01 (1H, d, 3J = 9 Hz), 7.85−7.74 (2H, m), 7.02
MATERIALS AND METHODS
Synthetic Materials and Methods. All reactions were carried out
under an N2 atmosphere using standard Schlenk techniques. H and
■
1
13C{1H} NMR spectra were recorded with a Varian Mercury 300
NMR or a Varian Inova 500 NMR spectrometer at room temperature.
Peaks were referenced to residual 1H signals from the deuterated
solvent and are reported in parts per million (ppm). 2-Methyl-4-
nitrosoresorcinol was synthesized by a literature procedure.37 All other
compounds were obtained from commercial vendors and used without
further purification. Mass spectra were obtained with either an Agilent
5973 Network mass-selective detector connected to an Agilent 689N
Network GC system or an Agilent 1100 Series LC/MSA trap. High-
resolution mass spectral analyses were carried out at the Massachusetts
Institute of Technology (MIT) Department of Chemistry Instrumen-
tation Facility (DCIF). UV−vis spectra were recorded with a Varian
Cary 1E spectrometer at 25 °C. Fluorescence spectra were obtained on
a Quanta Master 4 L-format scanning spectrofluorimeter (Photon
Technology International) at 25 or 37 °C. X-band electron
paramagnetic resonance (EPR) spectra were collected with a Bruker
EMX spectrometer equipped with an ER4199HS cavity and a Gunn
diode microwave source. All solvents were dried prior to use according
to standard methods. Silica gel 60 321 (0.015−0.040 mm) was used
for column chromatography. Thin-layer chromatography was
performed using Merck TLC aluminum sheets, silica gel 60 F254.
9-Hydroxy-8-methyl-5-benzo[a]phenoxazone (1). 2-Methyl-
4-nitrosoresorcinol (1.31 g, 7.68 mmol) and 1,3-dihydroxynaphthalene
(1.23 g, 7.68 mmol) were dissolved in n-butanol (20 mL) and heated
to 50 °C. To this solution was added concentrated sulfuric acid (2.6
mL), and the mixture was stirred for 15 min at 50 °C. The mixture was
allowed to cool to room temperature, and after 12 h, a dark precipitate
formed. The precipitate was collected by filtration, washed with a
mixture of ethanol and n-butanol (1:1, 10 mL) and then water/ethanol
(1:1, 50 mL), and dried under vacuum. This material (2.11 g, 96%
yield) was used without further purification. ESI-MS. Calcd for
[C17H11NO3]−: 277.1. Found: 277.0.
3
(1H, d, J = 9 Hz). ESI-MS. Calcd for [C17H8BrNO4+H+]+: 370.0.
Found: 369.9. Mp: 254 °C.
6-Bromo-9-hydroxy-8-[[(2-methylquinolin-8-yl)amino]-
methyl]-5-benzo[a]phenoxazone (BRNO1). Compound 4 (25 mg,
0.068 mmol) and 8-amino-2-methyl-quinoline (11 mg, 0.07 mmol)
were dissolved in dry methanol (10 mL) and stirred at room
temperature for 1 h. The solution was cooled to 0 °C, and sodium
cyanoborohydride (102 mg, 1.62 mmol) was added in one portion.
The mixture was stirred for 36 h at room temperature and then poured
into 10 mL of aqueous 5 M HCl. The resulting precipitate was filtered
off and washed with copious amounts of water and then dichloro-
methane/hexane (10 mL, 1:1). Subsequently, the solid was dried
1
under vacuum to afford 14.2 mg (41%) of a dark-red compound. H
NMR (CD3OD, 500 MHz, ppm): δ 12.33 (1H, s), 9.41 (1H, d, 3J = 6
Hz), 9.03 (1H, d, 3J = 6 Hz), 8.95−8.91 (1H,m), 8.71−8.68 (1H, m),
3
8.64−8.61 (1H, m), 8.55 (1H, d, J = 6 Hz), 8.20 (1H, d, 3J = 6 Hz),
3
8.14−8.11 (1H, m), 7.95 (1H, d, J = 6 Hz), 7.90 (1H, d, 3J = 6 Hz),
7.87 (1H, d, 3J = 6 Hz), 5.59 (2H, s), 3.46 (3H, s). ESI-HRMS. Calcd
for [C27H18N3BrO3 − H+]−: 510.0453. Found: 510.0439. Mp: 184 °C
(dec).
6-Bromo-9-hydroxy-8-[(quinolin-8-ylamino)methyl]-5-
benzo[a]phenoxazone (BRNO2). Compound 4 (20 mg, 0.054
mmol) and 8-aminoquinoline (9.4 mg, 0.065 mmol) were dissolved in
dry methanol (10 mL) and stirred at room temperature for 2 h. The
solution was then cooled to 0 °C, and sodium cyanoborohydride (50
mg, 0.79 mmol) was added in one portion. The mixture was stirred for
72 h at room temperature and poured into a 20 mL solution of a
saturated aqueous sodium bicarbonate solution. The resulting
precipitate was filtered off and washed with water. Drying under
vacuum afforded 10.6 mg (39%) of a dark-purple solid. 1H NMR
(DMSO-d6, 500 MHz, ppm): δ 8.70 (1H, d, 3J = 6 Hz), 8.57 (1H, d, 3J
8-Methyl-5-oxo-benzo[a]phenoxazin-9-yl acetate (2). Com-
pound 1 (1.41 g, 5.09 mmol) was dissolved in acetic anhydride (20
mL) and pyridine (2 mL). The mixture was heated for 3 h at 100 °C
and then left to stand at room temperature for 24 h. The resulting
dark-red crystalline solid was collected by filtration and washed with a
small amount of acetic anhydride and then a large volume of water.
3
3
= 6 Hz), 8.22 (1H, d, J = 6 Hz), 8.18 (1H, d, J = 6 Hz), 7.83−7.80
(1H, m), 7.72−7.69 (1H, m), 7.61 (1H, d, 3J = 6 Hz), 7.48−7.44 (1H,
m), 7.36−7.33 (1H, m), 7.17 (1H, d, J = 6 Hz), 7.05 (1H, d, J = 6
3
3
3286
dx.doi.org/10.1021/ic302793w | Inorg. Chem. 2013, 52, 3285−3294

Inorganic Chemistry
Article
3
Hz), 6.83 (1H, d, J = 6 Hz), 4.73 (2H, s), 2.62 (3H, s). ESI-HRMS.
optimized structures and a summary of the lowest-energy singlet
excited states.
Calcd for [C26H16N3BrO3 − H+]−: 496.0297. Found: 496.0293. Mp:
256 °C (dec).
Spectroscopic Materials and Methods. Piperazine-N,N′-bis(2-
ethanesulfonic acid) (PIPES; Calbiochem) and potassium chloride
(99.999%, Aldrich) were used to prepare buffered solutions (50 mM
PIPES, 100 mM KCl, pH 7.0) in deionized water with resistivity ≥18
MΩ cm−1, obtained using a Milli-Q water purification system. Nitric
oxide (NO) was purchased from Airgas and purified as previously
described.48 S-Nitroso-N-acetylpenicillamine (SNAP), S-nitroso-L-
glutathione (GSNO), sodium peroxynitrite, and Angeli’s salt were
purchased from Cayman Chemical and stored at −80 °C when not in
use. NO and the other RNS were injected into buffered solutions via a
gastight syringe. CuCl2·2H2O (99+%, Alfa Aesar) was used to prepare
7 mM CuCl2 stock solutions in DMSO. Dye stock solutions were
prepared in DMSO (5 mmol/L) and stored at −80 °C when not in
use. Measurements were performed under inert-atmosphere con-
ditions. Quantum yields of BRNO1−3 were determined in 50 mM
PIPES buffer (100 mM KCl, pH 7) using resorufin (λem = 585 nm, λex
= 572 nm, and Φ = 0.7449) as the reference in water (pH 9.5).
Cell Culture. HeLa cells and Raw 264.7 murine macrophages were
cultured in Dulbecco’s modified Eagle medium (DMEM; Cellgro,
MediaTek, Inc.), supplemented with 10% fetal bovine serum (FBS;
HyClone), 1% penicillin−streptomycin, 1% sodium pyruvate (Cellgro,
MediaTek, Inc.), 1% MEM nonessential amino acids (Sigma), and 1%
L-glutamine. For imaging studies, cells were grown to confluence,
passaged, and plated onto poly-D-lysine-coated plates. The plates,
containing 2 mL of DMEM, were incubated at 37 °C with 5% CO2 for
at least 12 h. The media were removed, the cells were washed with 5
mL of PBS buffer, and solutions of the fluorescent probes in 2 mL of
fresh DMEM were added. For all cell studies, the CuII complexes were
generated in situ by combining a stock solution of the fluorescent
sensor and CuCl2 in a 1:1 molar ratio 1 h prior to addition to the cells.
Plates were prepared with identical volumes from the same cell stock
solution to provide an equal number of cells in each plate. For NO
detection studies, NO production by iNOS was induced in Raw 264.7
murine macrophages with 1.6 μg/mL lipopolysaccharide (LPS; Sigma)
and 495−4950 U/mL of recombinant mouse interferon-γ (IFN-γ; BD
Biosciences). Cells were then incubated with 2.5 μM CuBRNO1 or 2.5
μM CuBRNO3 for 30 min. Then 5 μM HOECHST 33258 (Sigma)
was added, and the cells were incubated for a further 30 min. Prior to
imaging, cells were washed with 2 mL of PBS and then bathed with 1.5
mL of dye-free DMEM (Sigma). NO detection studies were
performed by addition of 1.25 mM GSNO. For HNO imaging
studies, HeLa and Raw 264.7 cells were treated as described above
before addition of 1.25 mM Angeli’s salt. Localization studies were
performed in both HeLa and Raw 264.7 cells in the presence of 2.5
μM CuBRNO1 or 2.5 μM CuBRNO3 and 1.25 mM Angeli’s salt.
Prior to fluorescence imaging, the cells were incubated with either 3
μM ER-Tracker Blue Blue-White DPX (Invitrogen), 13 μM
MitoTracker Green FM (Invitrogen), or CellLight Reagents BacMam
2.0 (Invitrogen) (10 parts per cell/16 h incubation) for 30 min.
Fluorescence Imaging. Fluorescence images were acquired on a
Zeiss Axiovert 200 M inverted epifluorescence microscope equipped
with an EM-CCD camera (Hamamatsu) and an X-Cite 120 metal
halide lamp (EXFP). Differential interference contrast (DIC) and
fluorescence images were obtained using an oil immersion 63×
objective lens with exposure times ranging from 50 ms to 2 s. The
microscope was operated with the Volocity 6.01 software (Improvi-
sion), and images were analyzed with the Volocity 6.01 software. All
fluorescent images were deconvoluted and background-corrected.
Images were measured before and after addition of Angeli’s salt or
NO-releasing agent. The image before addition of Angeli’s salt or NO-
releasing agent was taken as the background level.
6-Bromo-9-hydroxy-8-[[(pyridin-2-ylmethyl]amino]methyl)-
5-benzo[a]phenoxazone (BRNO3). Compound 4 (20 mg, 0.054
mmol) and 2-(aminomethyl)pyridine (11 μL, 0.11 mmol) were
dissolved in dry methanol (5 mL), and the resulting solution was
stirred at room temperature for 1 h. The solution was then cooled to 0
°C, and sodium cyanoborohydride (66 mg, 0.10 mmol) was added in
one portion. The mixture was stirred for 48 h at room temperature,
and a saturated aqueous solution of sodium bicarbonate (20 mL) was
added. The resulting precipitate was filtered off, washed with water (10
mL), and dried under vacuum to afford 21.4 mg (43%) of a purple
solid. 1H NMR (CD3OD, 500 MHz, ppm): δ 8.72 (1H, d, 3J = 6 Hz),
3
8.41 (1H, d, 3J = 6 Hz), 8.34 (1H, d, J = 6 Hz), 7.78−7.73 (2H, m),
7.67−7.63 (1H, m), 7.60 (1H, d, 3J = 6 Hz), 7.51 (1H, d, 3J = 6 Hz),
3
7.25−7.24 (1H, m), 6.79 (1H, d, J = 6 Hz), 4.23 (2H, s), 4.03 (2H,
s). ESI-HRMS. Calcd for [C23H16N3BrO3 − H+]−: 460.0297. Found:
460.0293. Mp: 256 °C (dec).
Headspace EI-MS Studies. BRNO1 (0.95 mg, 1.86 μmol) was
dissolved in acetonitrile (1 mL), and 0.8 equiv of CuCl2·2H2O (0.2
mg, 1.18 μmol, in 1 mL acetonitrile) was added to this solution in a
custom-made, gastight cell in an inert-atmosphere glovebox. The
mixture was maintained at room temperature for 1 h, prior to addition
of 10 mg of Angeli’s salt. The cell was connected to a He gas-flow inlet
tube and to the mass spectrometer. The connecting copper tubing was
purged thoroughly with He prior to analysis of the reaction headspace.
Headspace analysis was performed with the mass spectrometer
operating in selective ion mode.
Cyclic Voltammetry. Cyclic voltammograms were measured in a
three-electrode cell with a 2.0-mm-diameter glassy carbon working
electrode, a platinum auxiliary electrode, and a Ag/AgNO3 reference
electrode in acetonitrile. The solvent contained n-Bu4N(PF6) (0.05 M)
as the supporting electrolyte. The measurements were performed at
room temperature with a VersaSTAT3 (AMETEK) galvanostat.
Deoxygenation of the samples was accomplished by passing a stream
of N2 through the solutions for 5 min prior to the measurements, and
the solutions were kept under N2 for the duration of the study. All data
were referenced to the Fc/Fc+ couple as an internal standard (E1/2
=
+405 mV vs Ag/AgNO3 reference electrode).
X-ray Data Collection and Structure Solution Refinement.
Crystals of 2 and 3 were grown by slow evaporation of solutions of the
compounds dissolved in chloroform. Single crystals suitable for X-ray
analysis were coated with Paratone-N oil, mounted on a fiber loop, and
placed in a cold, gaseous N2 stream on a Bruker APEX CCD X-ray
diffractometer performing φ and ω scans at 100(2) K. Diffraction
intensities were measured using graphite-monochromated Mo Kα
radiation (λ = 0.71073 Å). Data collection, indexing, initial cell
refinements, frame integration, and final cell refinements were
accomplished with the program APEX2.38 Absorption corrections
were applied using the program SADABS.39 The structure was solved
by direct methods using SHELXS40 and refined against F2 on all data
by full-matrix least squares with SHELXL-9741 following established
refinement strategies. Crystallographic data collection and refinement
parameters are presented in Tables S1 and S2 in the Supporting
Information (SI).
Density Functional Theory (DFT) Calculations. All calculations
were performed with the Gaussian 03 program package42 using the
B3LYP functional.43,44 Geometry optimizations were carried out in the
gas phase using the 6-31g(d,p) basis set.45 Frequency calculations were
carried out at the same level of theory to ensure that geometries
converged to true local minima on the potential energy surface. The
30 lowest-energy singlet excited states were computed with time-
dependent DFT (TDDFT) calculations. For these calculations, the
larger 6-311++g(d,p) basis set was utilized. Solvent effects were
modeled with the conductor-like polarizable continuum model for
water.46 Electron-density difference maps (EDDMs) and calculated
UV−vis absorbance spectra were generated with the program
GaussSum2.2.47 Tables S3−S12 in the SI contain the coordinates of
RESULTS AND DISCUSSION
■
Design and Synthesis of Benzoresorufin-Based
Sensors. An ongoing goal of our laboratory is the development
of fluorescent sensors for NO and HNO. We found that an
effective strategy for designing such species is to couple a metal-
3287
dx.doi.org/10.1021/ic302793w | Inorg. Chem. 2013, 52, 3285−3294

Inorganic Chemistry
Article
Scheme 2. Synthetic Scheme for the Synthesis of Benzoresorufin-Based Sensors BRNO1−3
binding site with a fluorophore.50 In the “off” state of the
sensor, a paramagnetic ion in the metal-binding site quenches
the emission of the fluorescent reporter by photoinduced
electron transfer (PET). Upon reaction with NO, the
paramagnetic ion is reduced, displaced, or both, thus
eliminating the PET quenching pathway and triggering an
emission turn-on response.32,35,51−54 Our most successful NO
sensors, which operate nicely in cellular environments, contain
a 2-methyl-8-aminoquinoline metal-binding site and either a
fluorescein (FL series) or a seminaphthofluorescein (SNFL
series) dye as fluorescent reporters.27−30 CuII is used as the
paramagnetic fluorescence quencher. The methyl group in the
2 position of the aminoquinoline gives rise to only moderate
CuII-binding affinities, characterized by Kd values of approx-
imately 1.5 μM. In this study, we sought to change both the
fluorescent reporter and metal-binding site. Benzoresorufin,
which emits in the red, was chosen as the fluorophore, and in
addition to our standard 2-methyl-8-aminoquinoline CuII-
binding site, we investigated 8-aminoquinoline and 2-
(methylamino)pyridine.
The overall synthetic strategy for preparing the desired
benzoresorufin-based sensors is shown in Scheme 2. This
synthetic approach is general and therefore offers a viable
pathway to a wide variety of monofunctionalized benzoresor-
ufin dyes. Following a similar procedure for the synthesis of
naphthophenoxazones,55 the reaction of 2-methyl-4-nitro-
resorcinol37 and 1,3-dihydroxynaphthalene afforded 1, which
is readily purified in high yield by formation of the acetyl ester
derivative 2. The molecular structure of this compound, as
determined by single-crystal X-ray diffraction, is presented in
Figure S1 in the SI. To enable functionalization of the
benzoresorufin moiety, we sought to selectively brominate the
benzylic position. The bromination of compound 2 was carried
out with 1,3-dibromo-5,5′-dimethylhydantoin and VAZO88,
yielding dibrominated species 3 in high yield. The position of
the two bromine atoms was unambiguously verified by single-
crystal X-ray analysis (Figure S2 in the SI). The formation of 3
from 2 was monitored by 1H NMR spectroscopy, revealing that
bromination occurs first on the quinoid position, thus
preventing isolation of a singly brominated species (Figure S3
in the SI). Because the exocyclic nitrogen atoms of amino-
quinolines are only weak nucleophiles, the direct SN2 reaction
on bromide of 3 with 8-aminoquinolines was not a viable
pathway to the target molecule. An alternative approach of
reductive amination was therefore sought, involving oxidation
of dibromide 3 to aldehyde 4. This aldehyde is a versatile
intermediate that may be used for functionalization of the dyes
with a variety of metal-binding sites, thus allowing fine-tuning
of the metal-binding properties and the electronic properties of
the sensors. Three different sensors with varying metal-binding
sites were synthesized by reductive amination. In all cases, a
Schiff-base adduct was initially formed by reaction of aldehyde
4 with 2-methyl-8-aminoquinoline, 8-aminoquinoline, or 2-
(aminomethyl)pyridine followed by reduction with sodium
cyanoborohydride to afford the benzoresorufin-based dyes
BRNO1, BRNO2, and BRNO3, respectively, in modest yield.
For the more nucleophilic 2-(aminomethyl)pyridine, direct
substitution of the benzylic bromine in 3 could be achieved in
methanol with potassium carbonate as the base, affording
BRNO3 with yield similar to that of the two-step route.
Photophysical Properties. The photophysical properties
of BRNO1−3 were investigated in 50 mM PIPES buffer (100
mM KCl, pH 7) to reflect physiological conditions. The results
are summarized in Table 1. At pH 7, BRNO1 and BRNO2
exhibit a very broad band in their absorption spectra, with a
maximum at 470 nm and a lower-energy shoulder at 570 nm.
3288
dx.doi.org/10.1021/ic302793w | Inorg. Chem. 2013, 52, 3285−3294

Inorganic Chemistry
Article
quenching effect caused by the nitrogen lone pair of the amine
in the metal-binding group. BRNO3, on the other hand,
exhibits stronger emission. Upon excitation at 563 nm, an
emission band at 625 nm is observed, and the photo-
luminescence quantum yield is 0.25. The significantly higher
quantum yield and extinction coefficient of BRNO3 in
comparison to the other derivatives demonstrate that the
metal-binding site can modify the electronic properties of the
ground and excited states of the dye. This observation suggests
that modifying the metal-binding sites of metal-based sensors is
a viable strategy for adjusting not only the binding affinity but
also the photophysical properties of the system.
As for the fluorescein-based FL dyes, the absorbance and
emission spectra of BRNO1−3 are sensitive to changes in pH
because of the proton-accepting properties of the metal-binding
sites and the dye itself. The pH dependence of the absorbance
and emission properties of the metal-free dyes was therefore
investigated. Under basic conditions (pH 11), the three
compounds all display a strong broad absorbance feature
between 550 and 610 nm. This band, presumably arising from
the deprotonated sensors, is more intense than those observed
at pH 7. At low pH values, the absorption decays with
concomitant formation of a weaker feature centered at 470 nm
(Figure 1). The emission properties of the fluorophores also
vary with pH. Maximum emission intensity was observed
between pH 7.5 and 9.5 for BRNO1 and BRNO2, whereas for
BRNO3, maximum emission occurs between pH 7.0 and 9.0.
The slightly different ranges for the maximum emission
intensity of the aminoquinoline-based (BRNO1 and BRNO2)
and (aminomethyl)pyridine-based (BRNO3) sensors most
likely reflect disparate pKa values associated with the different
metal-binding sites. Analysis of the integrated emission data as a
function of pH returned apparent pKa values of 6.5 and 9.6 for
BRNO1, 6.5 and 9.7 for BRNO2, and 5.9 and 9.6 for BRNO3
(Figure 2 and Table 1). These values are significantly higher
Table 1. Photophysical Properties of the BRNO Derivatives
BRNO1
BRNO2
BRNO3
absorption
λmax (nm),
ε (M−1 cm−1
470,
5590 300
470,
4390 570
470, 5030 630
)
570,
2280 230
570,
2160 210
563,
11900 1100
emission
λmax (nm),
Φ (%)
623, 5.5 0.2 623, 4.8 0.4 625, 25
4
brightness
(Φε, M−1 cm−1
acid/base constants
pKa,1
)
125
103
2974
6.46 0.05
9.57 0.14
6.49 0.03
9.75 0.15
5.93 0.05
pKa,2
10.29 0.37
Figure 2. Fluorescence spectrum of BRNO3 at different pH values
(left) and the pH dependence of the fluorescence intensity (right).
than the pKa values reported for SNFL1 (4.9, 6.3, and 7.5),30
indicating that the benzoresorufin confers a higher degree of
basicitity to the molecules. The pH-dependent emission
properties of the compounds reveal that all dyes described
here exhibit high fluorescence under physiological conditions,
with less than a 10% change in fluorescence within 2 pH units
from pH 7.
DFT Calculations. To gain insight into the pH-dependent
photophysical properties of BRNO1−3, DFT calculations were
employed. Because BRNO1 and BRNO2 contain similar
aminoquinoline metal-binding sites, DFT calculations were
only carried out for BRNO1 and BRNO3 to compare how their
different metal-binding groups affect the properties of the
sensors. The geometries of BRNO1 and BRNO3, in both
Figure 1. UV−vis absorption spectra of BRNO1 (top), BRNO2
(center), and BRNO3 (bottom) at different pH values.
The extinction coefficients for the 470 nm maximum are only
5590 M−1 cm−1 for BRNO1 and 4390 M−1 cm−1 for BRNO2.
In contrast, the absorption spectrum of the 2-(aminomethyl)-
pyridine derivative BRNO3 is marked by a lower-energy band
centered at 563 nm of greater intensity (ε563 = 11900 M−1
cm−1; Figure 1). Upon excitation of BRNO1 and BRNO2 at
570 nm, a weak emission band (ϕBRNO1 = 0.055
0.002;
ϕBRNO2 = 0.048 0.004) is observed at 625 nm. These low
luminescence quantum yields are similar to those reported for
previously described NO sensors FL1 (ϕ = 0.077 0.002)27
and SNFL1 (ϕ = 0.027
0.004)30 and suggest a common
3289
dx.doi.org/10.1021/ic302793w | Inorg. Chem. 2013, 52, 3285−3294

Inorganic Chemistry
Article
neutral protonated and anionic deprotonated forms, were
optimized in the gas phase at the B3LYP/6-31g(d,p) level of
theory. Using these geometries, TDDFT calculations with an
implicit water solvation model were employed to simulate the
UV−vis absorbance spectra and to investigate the nature of the
corresponding excited states. For the deprotonated anionic
form of BRNO1, a transition (S2) having a large oscillator
strength (f = 0.736) is predicted at 528 nm. This calculated
transition energy is 0.12 eV (30 nm) greater than that of the
experimentally observed absorbance, which has a maximum at
560 nm (Figure S4 in the SI). An EDDM depicting the charge
redistribution in this excited state is shown in Figure 3, where
measured experimentally from the absorbance spectrum of
BRNO3 in pH 11 aqueous solution (Figure S6 in the SI). The
EDDM of this excited state (Figure 4) reveals it to be a
Figure 4. EDDMs of the two lowest-energy singlet excited states of
BRNO3 in both the anionic deprotonated (left) and neutral
protonated (right) forms. Green lobes represent holes, whereas purple
lobes are electrons.
benzoresorufin π−π* transition. The pure π−π* character of
the lowest-energy transition for BRNO3 is in contrast to the
lowest-energy excited state of BRNO1, which is characterized
in part as a LLCT. The lack of this charge transfer, which is
expected to favor nonradiative emission, in the S1 state of
BRNO3 is consistent with the higher photoluminescent
quantum yield of this sensor (Φ = 25%) compared to those
of BRNO1. The large oscillator strength (f = 0.5469) of the S1
state suggests that this transition is allowed, whereas the smaller
oscillator strength (f = 0.0639) of S2 indicates a smaller degree
of allowed character. The S1 and S2 excited states of the neutral,
protonated form of BRNO3 are separated by only 0.028 eV.
The computed absorbance maximum of 469 nm is in good
agreement with the value of 470 nm experimentally determined
for BRNO3 at pH 3 in aqueous solution (Figure S7 in the SI).
The nature of the S1 and S2 excited states is given by their
EDDMs, shown in Figure 4. The S1 state is primarily a
benzoresorufin π−π* transition, whereas the S2 state originates
from charge transfer from the amine lone pair to the
benzoresorufin π* orbital (n−π*). Of these two excited states,
only the S1 state is expected to be emissive because it involves
only the fluorescent benzoresorufin unit. The small energy
difference (0.058 eV) between the emissive S1 and dark S2
states renders these two states thermally accessible at room
temperature (kBT = 0.0256 eV at 25 °C). Therefore, the
significant emission quenching at low pH of BRNO3 (Figure 2)
arises from thermal population of the dark S2 state. At higher
pH values, where BRNO3 is largely deprotonated, the
nonemissive S2 state (Figure 4) lies 0.52 eV above the S1
state and is not accessible at room temperature.
Figure 3. EDDMs of the two lowest-energy singlet excited states of
BRNO1 in both the anionic deprotonated (left) and neutral
protonated (right) forms. Green lobes represent holes, whereas purple
lobes are electrons.
green lobes correspond to holes and purple lobes to electrons.
On the basis of the EDDM, the excited state is assigned to a
mixed transition with components of both benzoresorufin
π−π* and aminoquinoline-to-benzoresorufin ligand-to-ligand
charge transfer (LLCT) character. The mixture of LLCT into
this excited state may give rise to the low observed emission
quantum yield. The S1 state of this deprotonated anion occurs
at 616 nm with a low oscillator strength (f = 0.015). On the
basis of its EDDM, the nature of the S1 state is identical with
that of the allowed S2 state (Figure 3). For the neutral
protonated form of BRNO1, the main allowed transition (S2; f
= 0.531) blue shifts relative to that of the deprotonated form,
appearing at 471 nm. This calculation is in good agreement
with the experimentally observed transition centered at 470 nm
for BRNO1 measured at pH 3 in aqueous solutions (Figure S5
in the SI), The EDDM of this excited state reveals it to be
primarily benzoresorufin π−π* in character. The S1 excited
state of the protonated species exhibits a low oscillator strength
(f = 0.0003) and is therefore not expected to be an allowed
transition. The EDDM (Figure 3) of the S1 state indicates that
it corresponds to an aminoquinoline-to-benzoresorufin LLCT.
This dark charge-transfer excited state provides a plausible
nonradiative decay pathway for the π−π* S2 state. This
hypothesis is consistent with the significant decrease in
emission intensity of BRNO1 observed upon adjustment of
the solution pH from 11 to 3 (Figure 2). The S1 excited state of
the deprotonated anionic form of BRNO3, which contains an
(aminomethyl)pyridine metal-binding site, is computed to
occur at 529 nm, with a large oscillator strength (f = 0.8182).
This calculated value is 0.23 eV higher in energy than that
Metal-Binding Properties. Metal ions are endogenous to
biological systems and can potentially interact with and
influence the emissive properties of metal-based sensors. We
therefore investigated the emission response of BRNO1−3 in
the presence of several such metal ions. Upon addition of 1000
equiv of alkali or alkaline-earth metals, no significant change in
fluorescence intensity was observed for all three dyes (Figure
5). When late-first-row transition metals were added, a small
decrease of fluorescence intensity was observed. The largest
3290
dx.doi.org/10.1021/ic302793w | Inorg. Chem. 2013, 52, 3285−3294

Inorganic Chemistry
Article
BRNO2. This methyl group presumably destabilizes Cu2+
binding due to steric crowding. This destabilization is reflected
by a dissociation constant of BRNO1 that is 1 order of
magnitude higher than that of BRNO2. BRNO3, with its 2-
(aminomethyl)pyridine group, binds Cu2+ most effectively. The
dissociation constant of the Cu2+-BRNO3 complex is more
than 1 order of magnitude smaller than that of BRNO2. The
larger Cu2+ affinity of BRNO3 may arise from the greater
flexibility as well as the higher basicity of the secondary
nitrogen atom of the 2-(aminomethyl)pyridine unit by
comparison to the more rigid aminoquinoline metal-binding
sites in the other two sensors.
Reactivity with ROS and RNS. The response of our
copper-based sensors (CuBRNO1−3), assembled by treatment
of BRNO1−3 with 1 equiv of CuCl2 under anaerobic
conditions, to different RNS and ROS was investigated. The
CuBRNO probes were treated with 500 equiv of the RNS or
ROS, and the emission response was recorded after 60 min
(Figure 6). The reaction with the oxidizing agents sodium
Figure 5. Fluorescence response (F/F0) of the BRNO sensors upon
addition of 1000 equiv of metal ions.
decrease in emission intensity occurred following addition of
Cu2+ or Fe3+. Addition of the diamagnetic metal ions, Zn2+ and
Cd2+, induced small increases in the emission of BRNO1 and
BRNO2, whereas BRNO3 displayed a slight decrease in
emission. The small response of the BRNO sensors to Zn2+
is somewhat surprising because analogous monotopic fluo-
rescein-based derivatives (QZ1) exhibit a 42-fold turn-on upon
Zn2+ addition.56 For fluorescein-based zinc sensors, the zinc ion
serves to lower the energy of the nitrogen lone pairs in the
metal-binding sites, thereby inhibiting them from quenching
the fluorescence by a PET mechanism.57 The DFT calculations
presented above are consistent with the inability of the BRNO
sensors to turn-on with Zn2+. For BRNO1 and BRNO2, the
effect of zinc coordination on the energy of the the nitrogen
lone pair would have little impact on quenching charge transfer
from the aromatic π orbitals of the quinoline that is predicted
for the S1 and S2 excited states. For BRNO3 in the
deprotonated anionic form, the n−π* transition is significantly
higher in energy than the emissive benzoresorufin π−π* and
therefore unable to quench emission, even in the metal-free
form. The lack of a turn-on response upon interaction with
Zn2+ is an advantageous property of these sensors over first-
generation analogues because no false-positive response of the
sensors to highly abundant endogenous Zn2+ will occur.
Figure 6. Comparison of the selectivity for RNS/ROS for
[CuBRNO1]+, [CuBRNO2]+, and [CuBRNO3]+ in 50 mM PIPES
buffer (100 mM KCl, pH 7, 37 °C, 60 min) after addition of 500 equiv
of RNS/ROS. Excitation wavelength: 570 nm, BRNO1 and BRNO2;
563 nm, BRNO3. integrated fluorescence emission from 590 to 800
nm.
hypochlorite and hydrogen peroxide induced no significant
change in emission for any of the three sensors. Peroxynitrite, a
strong oxidant, decreases the emission, probably owing to the
formation of nonfluorescent oxidized resazurin derivatives.
Unexpectedly, the reaction with NO only led to 2.6- (BRNO1),
1.5- (BRNO2), and 1.7-fold (BRNO3) fluorescence increases.
A similar response was observed when the NO-releasing SNAP
was added. Addition of Angeli’s salt, Na2N2O3, an HNO source,
generated a larger turn-on response, with a 4.8-fold increase in
fluorescence intensity for CuBRNO1. Angeli’s salt decomposes
in aqueous solutions to release HNO and nitrite ions.58 To
verify that nitrite was not causing the increase in emission, this
ion was added to the probes. No increase in fluorescence was
observed, proving that HNO rather than NO2− gives rise to the
turn-on response. On the basis of emission, these probes are
more effective for sensing HNO than NO. Furthermore, they
exhibit selectivity over other ROS and RNS tested here.
Because Fe3+, like Cu2+, quenches the emission of the BRNO
sensors, we investigated the response of the FeIIIBRNO
complexes to ROS and RNS. These studies revealed little
Previously reported NO and HNO sensors utilize the Cu2+/
Cu+ redox couple to modulate the fluorescence response. In the
2+ oxidation state, the paramagnetic copper ion quenches the
emission of the fluorophore.27,33 Upon reduction of the copper
ion to the diamagnetic 1+ oxidation state by NO or HNO, the
emission is restored. To investigate the viability of this sensing
mechanism for the BRNO compounds, more detailed metal-
binding and photophysical studies were carried out in the
presence of Cu2+. Titration of the sensors with CuCl2 revealed
a 1:1 binding stoichiometry (Figure S8 in the SI). Addition of 1
equiv of CuCl2 resulted in a decrease in fluorescence intensity
(Figure S8 in the SI), from which the Cu2+ dissociation
constants (Kd) were determined by emission titrations as 4470
50 (BRNO1), 400
60 (BRNO2), and 18
2 nM
(BRNO3) (Figures S9 and S10 in the SI). These values are
consistent with the nature of the different metal-binding sites in
the three sensors. The chelating groups of BRNO1 and
BRNO2 differ only by the presence, in BRNO1, of a methyl
group ortho to the nitrogen atom of the quinoline ring in
3291
dx.doi.org/10.1021/ic302793w | Inorg. Chem. 2013, 52, 3285−3294

Inorganic Chemistry
Article
utility of the FeIIIBRNO system for sensing RNS and ROS
(Figure S11 in the SI).
hypothesized that HNO reduces the paramagnetic CuII center
in CuBRNO1 to restore the emission of BRNO1. To
investigate this possibility, the reaction of CuBRNO1 with
Angeli’s salt was monitored by EPR spectroscopy. For
CuBRNO1, an axial signal is observed in the EPR spectrum
due to the S = 1/2 CuII ion (Figure S15 in the SI). Addition of
Angeli’s salt to this solution led to the disappearance of this
signal (Figure S15 in the SI). This observation is consistent
with reduction of CuII by HNO to form a diamagnetic, EPR-
silent CuI ion. Although this EPR study verifies that CuII is
reduced by HNO, it does not establish whether the newly
formed CuI ion remains bound and whether or not this ion can
affect the photophysical properties of the sensor. An ESI-MS
spectrum of the reduced EPR solution revealed only the
presence of free BRNO1, suggesting that reduction of copper
leads to its dissociation. A 1:1 mixture of the CuI source
[Cu(CH3CN)4]PF6 and BRNO1 in acetonitrile, however, does
show evidence for the formation of a CuI complex judging by
the observation of a molecular ion peak corresponding to
[CuBRNO + H]+ in the ESI-MS spectrum (m/z 575.9; calcd
m/z 576.0). The fluorescence spectrum of this mixture in
aqueous buffer does not show CuI-induced changes in the
emission intensity of BRNO1. Therefore, the CuI ion either
does not bind strongly to BRNO1 or, if it does, it has little
effect on the emission intensity, as for other diamagnetic metal
ions tested (Figure 5).
Mechanism of HNO and NO Sensing. We previously
reported mechanistic investigations for a related NO-sensing
probe, CuFL1, which utilizes a 2-methyl-8-aminoquinoline
Cu2+-binding site like that in BRNO1 but with fluorescein as
the fluorophore.35 The reaction of CuFL1 with NO reduces
copper from the 2+ to 1+ oxidation state, forming an N-
nitrosated fluorescein derivative, FL1-NO, in the process. In
contrast to FL1 and CuFL1, FL1-NO is highly fluorescent. The
sensing mechanism of our previously reported HNO sensor,
CuBOT1, has also been investigated.33 Although this
compound does not react with NO, HNO reduces the
emission-quenching paramagnetic CuII ion and induces a
turn-on response. As in these previous studies, we sought to
investigate the mechanisms by which CuBRNO1 responds to
both NO and HNO.
The reaction of BRNO1 with 1 equiv of CuCl2 decreases the
emission intensity to one-fifth of its original value, most likely
due to quenching effect of the photoexcited state by the
paramagnetic CuII ion. An ESI-MS spectrum of freshly prepared
CuBRNO1 shows the molecular ion peak of [CuIIBRNO]+ at
m/z 573.1 (calcd m/z 573.0). Addition of NO or SNAP to
CuBRNO1 induces a 2.6-fold emission turn-on, and the ESI-
MS spectrum revealed the presence of a molecular ion peak at
m/z 538.9, corresponding to the N-nitrosated species BRNO1-
NO (calcd m/z 539.0 for [BRNO1 + NO − H]+). Reaction of
NOBF4 with BRNO1 in acetonitrile afforded BRNO1-NO, as
evidenced by the same molecular ion peak in the ESI-MS
spectrum. Treatment of BRNO1 with NO gas in the absence of
copper, on the other hand, gave no reaction, thereby
demonstrating the importance of the CuII ion as an electron
acceptor. An aliquot from the reaction mixture of NOBF4 and
BRNO1 in MeCN was diluted into PIPES buffer after 1 h, and
the emission spectrum was recorded. Unexpectedly, a 1.8-fold
decrease in fluorescence intensity compared to that of BRNO1
was observed (Figure S12 in the SI). Therefore, in contrast to
FL1-NO and FL1, BRNO1-NO is less emissive than BRNO1.
The relative order of emission intensity is BRNO1 > BRNO1-
NO > CuBRNO1. The low observed turn-on response induced
upon treatment of CuBRNO1 with NO is therefore a
consequence of the intrinsically low emission intensity of
BRNO1-NO. This result is not predicted by TDDFT
calculations. The lowest-energy singlet excited state of
BRNO1-NO is a benzoresorufin π−π* transition with a large
oscillator strength of 0.701 (Figure S13 in the SI). The lack of
charge-transfer character in this excited state suggests that it
should be highly emissive, or at least more so than BRNO1.
Therefore, additional nonradiative decay pathways must
operate in BRNO1-NO. Decay by internal conversion, perhaps
through the newly introduced NNO vibrational mode, is one
possible explanation.
Because the CuII/CuI redox couple is crucial for mediating
the detection of HNO in this system, the electrochemical
properties of CuBRNO1−3 were investigated by cyclic
voltammetry in acetonitrile. Quasi-reversible reduction features
occur at 90, 150, and 10 mV vs Fc/Fc+ for CuBRNO1−3,
respectively (Figure S16 in the SI). These values are close to
that of the CuII/CuI couple of CuCl2 under the same conditions
(130 mV vs Fc/Fc+). The reduction potential corresponding to
the HNO/NO couple is −0.36 V in water.59 In acetonitrile,
conditions similar to those used for our copper complexes, NO
is reduced irreversibly with an onset potential of approximately
−1 V versus the Fc/Fc+ couple. Therefore, all three sensors are
thermodynamically capable of oxidizing HNO to NO. To
confirm the formation of NO gas, EI-MS measurements were
carried out on the headspace of a reaction mixture of
CuBRNO1 and Angeli’s salt. These measurements confirmed
the formation of NO gas (Figure S17 in the SI), consistent with
our hypothesis that HNO reduces the CuBRNO sensors.
NO and HNO Detection in Living Cells. Fluorescence
imaging studies were carried out to investigate whether the
CuBRNO sensors can detect NO and HNO in living cells, as
they do in cuvettes. Only CuBRNO1 and CuBRNO3 were
used for these studies because they contain significantly
different metal-binding sites and were therefore expected to
exhibit different intracellular properties. Human cervical cancer
cells, HeLa, and murine macrophage cells, Raw 264.7, were
investigated. The Raw 264.7 cells allow for stimulation of
endogenous NO production by iNOS following treatment with
LPS and INF-γ.60 Treatment of both cell types with 2.5 μM
CuBRNO1 or CuBRNO3 resulted only in negligible emission
in the absence of an analyte. In contrast, treatment of both cell
lines with 500 equiv of Angeli’s salt in the presence of
CuBRNO1 or CuBRNO3 produced a distinct enhancement in
emission after 15 min of incubation (Figure 7). In all cases, the
observed increase in emission was comparable to that found
when similar experiments were performed in a cuvette,
demonstrating the suitability of the sensors to detect HNO
As discussed above, treatment of CuBRNO1 with Angeli’s
salt as an HNO source induces a 4.8-fold increase in emission
after 2 min (Figure S14 in the SI). This 4.8-fold turn-on
effectively corresponds to restoration of the fluorescence of
copper-free BRNO1 because removal of copper from
CuBRNO1 with ethylenediaminetetraacetic acid generates the
same response. After the 4.8-fold turn-on, subsequent additions
of Angeli’s salt produced no further increases in emission
intensity. Additionally, the emission properties of copper-free
BRNO1 do not change in response to Angeli’s salt. By analogy
to the mechanism of HNO sensing for CuBOT1,33 we
3292
dx.doi.org/10.1021/ic302793w | Inorg. Chem. 2013, 52, 3285−3294

Inorganic Chemistry
Article
both CuBRNO1 and CuBRNO3 localize to the endoplasmic
reticulum (ER). The spatial emission overlap between the
BRNO sensors and the ER tracker dye is marked by a Pearson
correlation coefficient62 of 0.61 for CuBRNO1 and 0.74 for
CuBRNO3 (Figures S19 and S20 and Table S13 in the SI).
This observation is in good agreement with literature reports
that resorufin dyes localize in the ER.63 In addition to the ER, a
significant degree of membrane staining was observed.
CONCLUSION
■
We have presented three novel benzoresorufin-based NO and
HNO fluorescent sensors. The use of benzoresorufin as the
fluorophore gives rise to emission at 625 nm, a region favorable
for biological imaging studies. In their copper-free forms,
BRNO1 and BRNO2 exhibit only weak fluorescence. BRNO3,
which contains a 2-(aminomethyl)pyridine metal-binding site,
has a significantly higher photoluminescent quantum yield
owing to the absence of low-energy charge-transfer excited
states. In the CuII-bound form, these complexes serve as
selective sensors for NO and HNO over other ROS and RNS
with modest increases in emission intensity. The observation
that higher turn-on responses for these sensors occur for HNO
rather than NO is an unexpected result based on the structural
similarities of the metal- and NO-reacting sites with respect to
those in previously reported NO sensors in the FL and SNFL
series. These results suggest that, in addition to the chemistry
occurring at the metal-binding site, the energy levels of the
fluorophore unit also play a significant role in modulating the
emission response and must be considered carefully. Moreover,
they indicate that a secondary amine in the copper-binding site
is necessary but not sufficient for selectively sensing NO,
whereas the best HNO sensor is obtained by blocking this
secondary amine and further tuning the electronic properties.
Mechanistic studies revealed that CuII reduction is necessary for
both HNO or NO turn-on. The fluorescence turn-on observed
with the CuIIBRNO probes reveals that the redox potentials of
the complexes are largely influenced by the fluorescent dye,
which participates in metal coordination, and are only slightly
influenced by the different amine chelating moieties. The metal-
binding affinity influences fluorescence turn-on after addition of
NO or HNO. There is a lower fluorescence turn-on for both
HNO and NO with tighter binding of CuII. Despite the small
turn-on response to NO, these sensors are quite effective at
detecting NO and HNO in both the cuvette and living cells.
Figure 7. Quantified fluorescence turn-on in cells after treatment with
Angeli’s salt as well as NO-releasing GSNO and stimulation of iNOS
by LPS/INF-γ in the presence of in situ generated [CuBRNO1]+ and
[CuBRNO3]+.
in live cells. To assess their ability to detect intracellular NO, we
induced NO production by iNOS in Raw 264.7 cells by adding
LPS and INF-γ.60 Raw 264.7 macrophages generate micromolar
concentrations of NO following treatment with endotoxins and
cytokines.61 Accordingly, these cells are ideal for studying the
response of NO probes to endogenously produced NO. After
stimulation of NO production, 1.4- and 1.2-fold increases in
emission intensity were observed for CuBRNO1 and
CuBRNO3, respectively (Figure 7). As a control experiment,
Raw 264.7 macrophages were treated with 500 equiv of GSNO.
This experiment provides a direct comparison of the turn-on
response of CuBRNO1 and CuBRNO2 induced by endoge-
nous and exogenous NO sources within the same cell line. In
agreement with our findings for endogenously produced NO,
treatment of cells containing our sensors with GSNO resulted
in a similar, small enhancement of emission for both sensors
(Figure 7). Representative cell images are presented in Figures
8 and S18 in the SI. Colocalization experiments, performed
with the sensor in the presence of Angeli’s salt, indicated that
ASSOCIATED CONTENT
■
S
* Supporting Information
NMR data of all compounds, as well as crystallographic data of
compounds 2 and 3, additional DFT results, CuII-binding
studies, RNS/ROS studies with [FeBRNO]2+, time dependency
study of HNO turn-on, EPR data, cyclic voltammetry data,
headspace EI-MS data, and intracellular localization experi-
ments. This material is available free of charge via the Internet
at http://pubs.acs.org.
AUTHOR INFORMATION
■
Figure 8. Fluorescence imaging of HNO in HeLa and Raw 264.7 cells.
For each set, the top image corresponds to treatment of cells with the
fluorescent probe. The bottom image corresponds to cells treated with
Angeli’s salt. (left) DIC images. (right) Fluorescence images. Scale bar
= 25 μm.
Corresponding Author
*E-mail: lippard@mit.edu.
Notes
The authors declare no competing financial interest.
3293
dx.doi.org/10.1021/ic302793w | Inorg. Chem. 2013, 52, 3285−3294

Inorganic Chemistry
Article
(32) Khin, C.; Lim, M. D.; Tsuge, K.; Iretskii, A.; Wu, G.; Ford, P. C.
Inorg. Chem. 2007, 46, 9323.
(33) Rosenthal, J.; Lippard, S. J. J. Am. Chem. Soc. 2010, 132, 5536.
(34) Royzen, M.; Wilson, J. J.; Lippard, S. J. J. Inorg. Biochem. 2013,
118, 162.
(35) McQuade, L. E.; Pluth, M. D.; Lippard, S. J. Inorg. Chem. 2010,
49, 8025.
(36) Klein, C.; Batz, H.-G.; Hermann, R. U.S. Patent 4,954,630, 1990.
(37) Raphael, R. A.; Ravenscroft, P. J. Chem. Soc., Perkin Trans. 1988,
1823.
ACKNOWLEDGMENTS
■
This work was supported the National Science Foundation
(Grant CHE-0611944 to S.J.L.). Instrumentation in the MIT
DCIF is maintained with funding from NIH Grant
1S10RR13886-01. U.-P.A. thanks the Alexander von Humboldt
Foundation for a postdoctoral fellowship. J.J.W. is a grateful
recipient of a David H. Koch Graduate Fellowship. Daniel J.
Graham and Dr. Amit Majumdar are thanked for assistance
with headspace EI-MS experiments and X-ray crystallography,
respectively.
(38) APEX2 v2009; Bruker AXS: Madison, WI, 2009.
(39) SADABS: Area-Detector Absorption Correction; University of
Gottingen: Gottingen, Germany, 2001.
̈
̈
(40) Sheldrick, G. M. Acta Crystallogr., Sect. A 1990, 46, 467.
(41) Sheldrick, G. Acta Crystallogr., Sect. A 2008, 64, 112.
(42) Gaussian 03, revision D.01; Gaussian, Inc.: Wallingford, CT,
2004.
REFERENCES
■
(1) Darley-Usmar, V.; Halliwell, B. Pharm. Res. 1996, 13, 649.
(2) Palmer, R. M. J.; Ferrige, A. G.; Moncada, S. Nature 1987, 327,
524.
(3) Kroencke, K.-D.; Fehsel, K.; Kolb-Bachofen, V. Clin. Exp.
Immunol. 1998, 113, 147.
(43) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785.
(44) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(45) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys. 1972, 56,
2257.
(46) Cossi, M.; Rega, N.; Scalmani, G.; Barone, V. J. Comput. Chem.
2003, 24, 669.
(4) Hobbs, A. J.; Higgs, A.; Moncada, S. Annu. Rev. Pharmacol.
Toxicol. 1999, 39, 191.
(5) Wang, Y.; Ruby, E. G. Cell. Microbiol. 2011, 13, 518.
(6) Tennyson, A. G.; Lippard, S. J. Chem. Biol. 2011, 18, 1211.
(7) Toledo, J. C.; Augusto, O. Chem. Res. Toxicol. 2012, 25, 975.
(8) Fukuto, J. M.; Carrington, S. J.; Tantillo, D. J.; Harrison, J. G.;
Ignarro, L. J.; Freeman, B. A.; Chen, A.; Wink, D. A. Chem. Res.
Toxicol. 2012, 25, 769.
(9) Marletta, M. A. Cell 1994, 78, 927.
(10) Muenzel, T.; Feil, R.; Muelsch, A.; Lohmann, S. M.; Hofmann,
F.; Walter, U. Circulation 2003, 108, 2172.
(11) Moncada, S.; Higgs, E. A. Br. J. Pharmacol. 2006, 147, S193.
(12) Bult, H.; Boeckxstaens, G. E.; Pelckmans, P. A.; Jordaens, F. H.;
Van Maercke, Y. M. Nature 1990, 345, 346.
(13) Butler, A. R.; Williams, D. L. H. Chem. Soc. Rev. 1993, 22, 233.
(14) Fukuto, J. M.; Dutton, A. S.; Houk, K. N. ChemBioChem 2005,
6, 612.
(15) Irvine, J. C.; Ritchie, R. H.; Favaloro, J. L.; Andrews, K. L.;
Widdop, R. E.; Kemp-Harper, B. K. Trends Pharmacol. Sci. 2008, 29,
601.
(16) Feelisch, M. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 4978.
(17) Fukuto, J. M.; Chiang, K.; Hszieh, R.; Wong, P.; Chaudhuri, G.
J. Pharmacol. Exp. Ther. 1992, 263, 546.
(18) Bermejo, S. E.; Saenz, D. A.; Alberto, F.; Rosenstein, R. E.; Bari,
S. E.; Lazzari, M. A. Thromb. Haemostasis 2005, 94, 469.
(19) Miles, A. M.; Wink, D. A.; Cook, J. C.; Grisham, M. B. In
Methods in Enzymology; Lester, P., Ed.; Academic Press: New York,
1996; Vol. 268, p 105.
(20) Nagano, T.; Yoshimura, T. Chem. Rev. 2002, 102, 1235.
(21) Sasaki, E.; Kojima, H.; Nishimatsu, H.; Urano, Y.; Kikuchi, K.;
Hirata, Y.; Nagano, T. J. Am. Chem. Soc. 2005, 127, 3684.
(22) Terai, T.; Urano, Y.; Izumi, S.; Kojima, H.; Nagano, T. Chem.
Commun. 2012, 48, 2840.
(47) O’Boyle, N. M.; Tenderholt, A. L.; Lagner, K. M. J. Comput.
Chem. 2008, 29, 839.
(48) Lim, M. D.; Lorkovic, I. M.; Ford, P. C. Nitric Oxide; Elsevier:
San Diego, 2005; Vol. 396, Part E, p 3.
(49) Bueno, C.; Villegas, M. L.; Bertolotti, S. G.; Previtali, C. M.;
Neumann, M. G.; Encinas, M. V. Photochem. Photobiol. 2002, 76, 385.
(50) Tonzetich, Z. J.; McQuade, L. E.; Lippard, S. J. Inorg. Chem.
2010, 49, 6338.
(51) Sarma, M.; Kalita, A.; Kumar, P.; Singh, A.; Mondal, B. J. Am.
Chem. Soc 2010, 132, 7846.
(52) Kalita, A.; Kumar, P.; Deka, R. C.; Mondal, B. Inorg. Chem.
2011, 50, 11868.
(53) Mondal, B.; Kumar, P.; Ghosh, P.; Kalita, A. Chem. Commun.
2011, 47, 2964.
(54) Sarma, M.; Mondal, B. Inorg. Chem. 2011, 50, 3206.
(55) Fischer, O.; Hepp, E. Ber. Dtsch. Chem. Ges. 1903, 36, 1807.
(56) Nolan, E. M.; Jaworski, J.; Okamoto, K.-I.; Hayashi, Y.; Sheng,
M.; Lippard, S. J. J. Am. Chem. Soc. 2005, 127, 16812.
(57) Kowalczyk, T.; Lin, Z.; Voorhis, T. V. J. Phys. Chem. A 2010,
114, 10427.
(58) Dutton, A. S.; Fukuto, J. M.; Houk, K. N. J. Am. Chem. Soc.
2004, 126, 3795.
(59) Bartberger, M. D.; Liu, W.; Ford, E.; Miranda, K. M.; Switzer,
C.; Fukuto, J. M.; Farmer, P. J.; Wink, D. A.; Houk, K. N. Proc. Natl.
Acad. Sci. U.S.A. 2002, 99, 10958.
(60) Miwa, M.; Stuehr, D. J.; Marletta, M. A.; Wishnok, J. S.;
Tannenbaum, S. R. Carcinogenesis 1987, 8, 955.
(61) Noda, T.; Amano, F. J. Biochemistry 1997, 121, 38.
(62) French, A. P.; Mills, S.; Swarup, R.; Bennet, M. J.; Pridmore, T.
P. Nat. Protocols 2008, 3, 619.
(23) Yang, Y.; Seidlits, S. K.; Adams, M. M.; Lynch, V. M.; Schmidt,
C. E.; Anslyn, E. V.; Shear, J. B. J. Am. Chem. Soc. 2010, 132, 13114.
(24) Dobmeier, K. P.; Riccio, D. A.; Schoenfisch, M. H. Anal. Chem.
2008, 80, 1247.
(63) Hakamata, W.; Machida, A.; Oku, T.; Nishio, T. Bioorg. Med.
Chem. Lett. 2011, 21, 3206.
(25) Wong, P. S. Y.; Hyun, J.; Fukuto, J. M.; Shirota, F. N.;
DeMaster, E. G.; Shoeman, D. W.; Nagasawa, H. T. Biochemistry 1998,
37, 5362.
(26) Reisz, J. A.; Zink, C. N.; King, S. B. J. Am. Chem. Soc. 2011, 133,
11675.
(27) Lim, M. H.; Xu, D.; Lippard, S. J. Nat. Chem. Biol. 2006, 2, 375.
(28) Lim, M. H.; Wong, B. A.; Pitcock, W. H.; Mokshagundam, D.;
Baik, M.-H.; Lippard, S. J. J. Am. Chem. Soc. 2006, 128, 14364.
(29) McQuade, L. E.; Lippard, S. J. Inorg. Chem. 2010, 49, 7464.
(30) Pluth, M. D.; Chan, M. R.; McQuade, L. E.; Lippard, S. J. Inorg.
Chem. 2011, 50, 9385.
(31) Tsuge, K.; DeRosa, F.; Lim, M. D.; Ford, P. C. J. Am. Chem. Soc.
2004, 126, 6564.
3294
dx.doi.org/10.1021/ic302793w | Inorg. Chem. 2013, 52, 3285−3294