Ratiometric fluorescent probes for detection of intracellular singlet oxygen

Article abstract of DOI:10.1021/ol401421r

We have developed a series of molecular probes for the fluorescent detection of singlet dioxygen (¹O₂). The probes, based on asymmetrically substituted 1,3-diarylisobenzofurans, undergo the [2 + 4] cycloaddition reaction with ¹O₂, producing ratiometric fluorescent responses. Two-photon fluorescence microscope experiments demonstrated the biological utility of the probes for the visualization of endogenous ¹O₂ in macrophage cells.

Full text of DOI:10.1021/ol401421r

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
Ratiometric Fluorescent Probes for
Detection of Intracellular Singlet Oxygen
Dayoung Song,† Somin Cho,† Yejee Han,† Youngmin You,*,‡ and Wonwoo Nam*,†
Department of Bioinspired Science, Ewha Womans University, Seoul 120-750, Korea,
and Department of Advanced Materials Engineering for Information and Electronics,
Kyung Hee University, Yongin-si, Gyeonggi-do 446-701, Korea
odds2@khu.ac.kr (YY); wwnam@ewha.ac.kr (WN)
Received May 20, 2013
ABSTRACT
We have developed a series of molecular probes for the fluorescent detection of singlet dioxygen (1O2). The probes, based on asymmetrically substituted
1,3-diarylisobenzofurans, undergo the [2 þ 4] cycloaddition reaction with 1O2, producing ratiometric fluorescent responses. Two-photon fluorescence
microscope experiments demonstrated the biological utility of the probes for the visualization of endogenous 1O2 in macrophage cells.
Singlet dioxygen (1O2) is a highly reactive oxygen species
(hROS) that oxidizes a broad range of biomolecules, such
aqueous milieu.12 Furthermore, direct monitoring of the
phosphorescence of 1O2 lacks practical utility because the
emission features very low photoluminescence quantum
yields (PLQYs ≈ 10ꢀ6).13 Thus there is a strong need for
1O2 probes with high sensitivity and large dynamic ranges.
Among various probes, photoluminescent probes are
the most attractive because they provide many advantages
in bioimaging applications.14 Fluorescent 1O2 probes have
been developed by conjugating 1O2 traps and fluorophores,
1
as nucleic acids,1ꢀ3 lipids,4 and proteins.4ꢀ6 These O2
oxidation chemistries provide key mechanisms in the
regulation of intracellular signaling pathways7ꢀ11 and thus
are critically linked to human pathophysiology. However,
studies toward understanding molecular mechanisms of
1O2 in vivo have been significantly retarded by difficulties
due to its short half-life (τ1/2 = 10ꢀ6 to 10ꢀ5 s) in the
such as fluorescein,15ꢀ17 poly(fluorene),18ꢀ20 cyanine,21
a
† Ewha Womans University.
Eu(III) complex,22 and a Re(I) complex.23 Reactive dienes,
including 9,10-disubstituted anthracene15ꢀ18,20,22,23 and
‡ Kyung Hee University.
(1) Kang, P.; Foote, C. S. J. Am. Chem. Soc. 2002, 124, 4865–4873.
(2) Cadet, J.; Bellon, S.; Berger, M.; Bourdat, A.-G.; Douki, T.;
Duarte, V.; Frelon, S.; Gasparutto, D.; Muller, E.; Ravanat, J.-L.;
Sauvaigo, S. Biol. Chem. 2002, 383, 933–943.
(3) Ravanat, J.-L.; Cadet, J. Chem. Res. Toxicol. 1995, 8, 379–388.
(4) Wilkinson, F.; Helman, W. P.; Ross, A. B. J. Phys. Chem. Ref.
Data 1995, 24, 663–1021.
(5) Davies, M. J. Biochem. Biophys. Res. Commun. 2003, 305, 761–
770.
(6) Matheson, I. B. C.; Etheridge, R. D.; Kratowich, N. R.; Lee, J.
Photochem. Photobiol. 1975, 21, 165–171.
(13) Schweitzer, C.; Schmidt, R. Chem. Rev. 2003, 103, 1685–1757.
(14) Gomes, A.; Fernandes, E.; Lima, J. L. F. C. J. Biochem. Biophys.
Methods 2005, 65, 45–80.
(15) Umezawa, N.; Tanaka, K.; Urano, Y.; Kikuchi, K.; Higuchi, T.;
Nagano, T. Angew. Chem., Int. Ed. 1999, 38, 2899–2901.
(16) Tanaka, K.; Miura, T.; Umezawa, N.; Urano, Y.; Kikuchi, K.;
Higuchi, T.; Nagano, T. J. Am. Chem. Soc. 2001, 123, 2530–2536.
(17) Gollmer, A.; Arnbjerg, J.; Blaikie, F. H.; Pedersen, B. W.;
Breitenbach, T.; Daasbjerg, K.; Glasius, M.; Ogilby, P. R. Photochem.
Photobiol. 2011, 87, 671–679.
€
(7) Klotz, L.-O.; Kroncke, K.-D.; Sies, H. Photochem. Photobiol. Sci.
2003, 2, 88–94.
(18) Koylu, D.; Sarrafpour, S.; Zhang, J.; Ramjattan, S.; Panzer,
M. J.; Thomas, S. W., III Chem. Commun. 2012, 48, 9489–9491.
(19) Altınok, E.; Friedle, S.; Thomas, S. W., III Macromolecules
2013, 46, 756–762.
(20) Zhang, J.; Sarrafpour, S.; Pawle, R. H.; Thomas, S. W., III
Chem. Commun. 2011, 47, 3445–3447.
(8) Ryter, S. W.; Tyrrell, R. M. Free Radical Biol. Med. 1998, 24,
1520–1534.
(9) Wang, X.; Martindale, J. L.; Liu, Y.; Holbrook, N. J. Biochem. J.
1998, 333, 291–300.
€
(10) Kroncke, K.-D.; Klotz, L.-O.; Suschek, C. V.; Sies, H. J. Biol.
Chem. 2002, 277, 13294–13301.
(11) Klotz, L.-O.; Briviba, K.; Sies, H. Methods Enzymol. 2000, 319,
(21) Xu, K.; Wang, L.; Qiang, M.; Wang, L.; Li, P.; Tang, B. Chem.
Commun. 2011, 47, 7386–7388.
130–143.
(22) Song, B.; Wang, G.; Tan, M.; Yuan, J. J. Am. Chem. Soc. 2006,
128, 13442–13450.
(12) Lindig, B. A.; Rodgers, M. A. J. J. Phys. Chem. 1979, 83, 1683–
1688.
(23) Liu, Y.-J.; Wang, K.-Z. Eur. J. Inorg. Chem. 2008, 5214–5219.
r
10.1021/ol401421r
XXXX American Chemical Society

1
histidine,21 served as the O2 traps, and fluorescence re-
sponses were obtained after the [2 þ 4] cycloaddition reac-
tions between 1O2 and the traps. The utility of these probes
has been limited because their monotonic fluorescence turn-
on or turn-off responses suffer from artifacts associated with
sensor localization and instrumental drifts.15ꢀ17,19,21ꢀ23 In
addition, some fluorophores, such as chlorinated fluorescein,
convert ground-state dioxygen (3O2) into 1O2 under
photoexcitation,17,24 which would severely alter the cellular
1O2 levels. Furthermore, probes examined for their utility for
the detection of biological 1O2 are extremely rare.21,22 These
drawbacks present a strong rationale for development of new
probes capable of ratiometric fluorescent detection and
minimized 1O2 photosensitization.
Scheme 2. Synthesis of StPBF, a Ratiometric Fluorescent Probe
a
for 1O2
1,3-Diphenylisobenzofuran (DPBF) is an attractive
1
platform for the construction of O2 probes because it
rapidly reacts with 1O2 at a rate constant of 9.6 ꢁ 108 Mꢀ1 sꢀ1
in water25,26 and because it does not photosensitize 1O2 due
to effective suppression of the triplet state population.27
However, the 1O2 products, an endoperoxide or 1,2-diben-
zoylbenzene, are nonfluorescent. Thus DPBF itself offers
little practical utility for 1O2 detection. We envisioned that the
a The synthetic routes to other probes are shown in the Supporting
Information.
Synthetic routes to StPBF are depicted in Scheme 2. The
syntheses of PPBF and PyPBF are described in the
Supporting Information (Scheme S1). The fluorophore, a
4-(diphenylamino)stilbene unit, was obtained through the
two-step synthesis, comprising a HornerꢀWadsworthꢀ
Emmons olefination followed by a Pd(II)-catalyzed Heck
reaction. The Re(I) dimer catalysis protocol established by
the group of Kuninobu and Takai was employed for the
construction of the isobenzofuran unit.28 The 1O2 probes and
their precursors were characterized by standard spectro-
scopic methods, and the data fully agreed with the proposed
structures (Supporting Information, Figures S1ꢀS12).
Scheme 1. Ratiometric Fluorescent Probes for Detection of 1O2
Table 1. Photophysical Data for the 1O2 Probes
λabs
(nm, log ε)
λems
(nm)b
PLQY
(%)c
τobs
(ns)d
FI
ratioe
DPBF
415 (4.26)
298 (3.78)
411 (4.35)
324 (3.96)
438 (4.40)
358 (4.22)
450 (4.50)
382 (4.16)
455
ndf
72 ( 7.4
ndf
1.46
ndf
turn-off
80
displacement of the phenyl moiety of DPBF with fluorescent
1
chromophores will produce ratiometric fluorescent O2 re-
1
a
þ O2
PPBF
476
370
512
398
505
435
6.4 ( 1.0
0.78 ( 0.12
28 ( 2.7
6.1 ( 0.5
39 ( 6.0
41 ( 8.2
1.38
0.425
1.48
1.17
2.29
1.58
sponses: the π-conjugation over 1-phenylisobenzofuran and
the chromophore would be shortened by the 1O2 reaction to
give hypsochromically shifted fluorescence emission instead
of fluorescence turn-off (see Scheme 1). To test this idea, we
1
a
þ O2
PyPBF
352
1
a
þ O2
StPBF
14
1
1
a
designed a series of O2 probes comprising phenanthrene
þ O2
(PPBF), pyrene (PyPBF), and 4-(diphenylamino)stilbene
(StPBF) as fluorophores. Herein, we report the design,
syntheses, sensing properties, and biological applications
of the novel ratiometric fluorescent 1O2 probes. Fluorescent
a O2-saturated THF solutions containing 10 μM probe and 100 nM
photosensitizer were photoirradiated under 365 nm for 100 s. b λex =415nm
(DPBF), 331 nm (PPBF), 358 nm (PyPBF), and 350 nm (StPBF). c Photo-
luminescence quantum yield relative to 9,10-diphenylanthracene (PLQY =
0.90, cyclohexane) as a standard. d Fluorescence lifetime determined through
picosecond (λex = 377 nm) time-correlated single photon counting techni-
1
responses of the probes were compared to correlate O2
sensitivity and the extent of π-conjugation.
ques. e Fluorescence intensity (FI) ratio: PPBF, FI370 nm/FI476
;
nm
PyPBF, FI398 nm/FI512 nm; StPBF, FI435 nm/FI505 nm. f Not detected.
(24) Gandin, E.; Lion, Y.; Vorst, A. V. d. Photochem. Photobiol.
1983, 37, 271–278.
(25) Onitsuka, S.; Nishino, H.; Kurosawa, K. Tetrahedron 2001, 57,
6003–6009.
(26) Aubry, J.-M.; Pierlot, C.; Rigaudy, J.; Schmidt, R. Acc. Chem.
Res. 2003, 36, 668–675.
(27) Zhang, X.-F.; Li, X. J. Lumin. 2011, 131, 2263–2266.
1
As expected, the O2 probes exhibited significant red
shifts in their fluorescence peak wavelengths relative to DPBF
(10 μM in THF, Δν = νems(DPBF) ꢀ νems(1O2 probe):
PPBF, 970 cmꢀ1; StPBF, 2180 cmꢀ1; PyPBF, 2450 cmꢀ1
)
B
Org. Lett., Vol. XX, No. XX, XXXX

due to the effective conjugation to the fluorophores. The
probes are highly fluorescent with photoluminescence quan-
tum yields (PLQYs) of 6.4ꢀ39%. To evaluate 1O2 sensing
properties, we employed a biscyclometalated Ir(III) com-
plex, IrF (Scheme 1), as a 1O2 photosensitizer because it has
425 mu (Supporting Information, Figure S16). These values
correspond to [1,2-diketone þ Na]þ and [endoperoxide þ
Na]þ, respectively (see Scheme 1). It appeared that the
1
dominant O2 reaction product was 1,2-diketone because
13C NMR spectra (100 MHz) for the product displayed two
carbonyl peaks at δ 197.3 and 197.4 ppm (acetone-d6;
Supporting Information, Figure S17). HPLC chromato-
grams acquired during the 1O2 reaction displayed the com-
plete conversion of PPBF to a single product (Supporting
Information, Figure S18). Indeed, the experimental UVꢀvis
absorption spectrum of the product is in good agreement
with a simulated spectrum of the 1,2-diketone of PPBF (TD-
DFT, B3LYP/6-311þG(d,p)//uB3LYP/6-311þG(d,p):C-
PCM (THF); Supporting Information, Figure S19 and Table
S1). Taken together, the studies provide strong evidence that
1
the probes react with O2 to yield the 1,2-diketone having
hypsochromically shifted fluorescence emission.
The observed rates (kobs) for the reaction between 1O2
and the probes were determined by monitoring the tem-
poral changes in the fluorescence intensity ratios (StPBF,
PyPBF, and PPBF) or the total fluorescence intensity
(DPBF) during the continuous photoirradiation. The
probe concentration was varied in the range of 2ꢀ20 μM
to keep the pseudo first-order kinetics of the probe. As
shown in Figure 1b, the plot of kobs as a function of the
probe concentration follows a linear line. The largest slope
Figure 1. (a) Fluorescence response of 10 μM StPBF toward 1O2.
The arrows indicate the direction of the spectral changes. The
inset photos show fluorescence emission of the StPBF solution
before (top) and after (bottom) the 1O2 generation. (b) Plots of
observed reaction rates (kobs) as a function of the concentration
of the 1O2 probes.
been proved to be highly efficient for 1O2 generation.29 As
shown in Figure 1, 100 s photoirradiation (λex = 365 nm) of
an O2-bubbled (30 min) THF solution containing 10 μM
StPBF and 100 nM IrF led to fluorescence color changes
from green to blue. The fluorescence peak wavelength shift-
ed from 505 to 435 nm, and the corresponding fluorescence
is observed for StPBF (6.25 Mꢀ1 ꢀ1), and the slope
s
decreases in the order of PyPBF (3.71 Mꢀ1 sꢀ1) > PPBF
(0.557 Mꢀ1 sꢀ1) > DPBF (0.252 Mꢀ1 sꢀ1). Although the
slope cannot be equated with the reaction rate constant
under the experimental conditions, the results reveal that
the 1O2 reactivity increases in proportion with the extent of
π-conjugation. By contrast, the oxidation potentials of the
probes displayed a weak correlation with the 1O2 reactivity
(Supporting Information, Figure S20).
intensity (FI) ratio at 435 nm versus 505 nm (i.e., FI435
/
nm
FI505 nm) increased by a factor of 14. The clear isosbestic
point at 490 nm indicates the absence of any fluorescent
byproduct. By contrast, the ratiometric change was not ob-
served without IrF or photoirradiation (Supporting Infor-
1
mation, Figure S13a). In addition, the addition of an O2
1
scavenger, 1 mM NaN3,30 abrogated the fluorescent O2
response (Supporting Information, Figure S13b). These
results unambiguously indicate that the fluorescence change
1
was attributed to O2. PPBF and PyPBF displayed similar
1
fluorescence ratiometric responses to O2 with changes in
the fluorescence intensity ratios of factors of 80 (PPBF,
FI370 nm/FI476 nm) and 352 (PyPBF, FI398 nm/FI512 nm),
respectively (Table 1 and Supporting Information, Figure S14).
The UVꢀvis absorption spectra of 10 μM probes dis-
played hypsochromic shifts (Δν = νabs(before reaction) ꢀ
ν
abs(after reaction): StPBF, ꢀ3960 cmꢀ1;PyPBF,ꢀ5100 cmꢀ1;
PPBF, ꢀ6530 cmꢀ1) upon the reaction with 1O2, indicating
products with wide band gap energies (Supporting Informa-
tion, Figure S15). Positive mode ESIꢀMS spectra for the 1O2
reaction product of PPBF revealed peaks at m/z = 409 and
Figure 2. Fluorescence 1O2 selectivity of StPBF over other
reactive oxygen species (O2•ꢀ, 1 mM KO2; ClOꢀ, 1 mM NaOCl;
1 mM t-BuOOH; 1 mM H2O2; t-BuO•, 1 mM FeSO4 þ 100 μM t-
•
BuOOH; OH, 1 mM FeSO4 þ 100 μM H2O2) and chemical
oxidants (CAN, 1 mM (NH4)2CeIV(NO3)6; DDQ, 1 mM 2,3-
dichloro-5,6-dicyano-benzoquinone).
(28) Kuninobu, Y.; Nishina, Y.; Nakagawa, C.; Takai, K. J. Am.
Chem. Soc. 2006, 128, 12376–12377.
(29) You, Y.; Lee, S.; Kim, T.; Ohkubo, K.; Chae, W.-S.; Fukuzumi,
S.; Jhon, G.-J.; Nam, W.; Lippard, S. J. J. Am. Chem. Soc. 2011, 133,
18328–18342.
1
1
The O2 probes exhibited great selectivity toward O2
over other reactive oxygen species, including O2•ꢀ, ClOꢀ,
(30) Harbour, J. R.; Issler, S. L. J. Am. Chem. Soc. 1982, 104, 903–
905.
Org. Lett., Vol. XX, No. XX, XXXX
C

fluorescence increase in channel 1, indicating endogenous
1
production of O2. The corresponding increases in the
fluorescence intensity ratios of the channel 1 images over
the channel 2 images were greater than the standard devia-
tions (Supporting Information, Figure S24). Additional
micrographs are included in Supporting Information,
1
Figure S25. As expected for the O2 probe, pretreatment
1
with 2 mM histidine, an intracellular O2 scavenger,32
suppressed the 1,2-diketone fluorescence. To the best of
our knowledge, this is the first visualization of intracellular
1O2 by ratiometric fluorescent signaling.
To summarize, we developed a series of ratiometric
1
fluorescent probes for the detection of O2. The probes
1
underwent an irreversible reaction with O2, producing
hypsochromically shifted fluorescence. The ratiometric
responses were not affected by the probe concentrations
(Supporting Information, Figure S26), demonstrating the
benefit of the ratiometric fluorescent signaling. The probes
were capable of detecting 1O2 among various reactive oxy-
Figure 3. Fluorescent detection of intracellular 1O2 in RAW
264.7 macrophage cells. Cells treated with 10 μM PyPBF for 30 min
were visualized through dual channels: channel 1, λex = 750 nm
(two photon) and λems = 377ꢀ481 nm; channel 2, λex = 750 nm
(two photon) and λems = 490ꢀ568 nm; ratio image, channel
1/channel 2. The cells were subsequently incubated with 2 μM
phorbol myristate acetate (PMA, 30 min) to stimulate endogenous
1O2 generation (middle panels). Control experiments were per-
formed for the cells preincubated with 1O2 scavenger, 2 mM
histidine (30 min; bottom panels).
1
gen species. We further established that greater O2 sensi-
tivity could be accomplished using the isobenzofuran
platforms with elongated π-conjugation. Finally, biological
utility was demonstrated by dual channel two-photon fluor-
escent visualization of endogenous 1O2 in macrophage cells.
•
t-BuO•, OH, and H2O2, and chemical oxidants, such as
Acknowledgment. This work was supported by the CRI
(W.N.), GRL (2010-00353) (W.N.), and WCU (R31-2008-
000-10010-0) (W.N.) programs from the National Re-
search Foundation (NRF) of Korea.
(NH4)2CeIV(NO3)6 and DDQ (see Figure 2 for StPBF and
Supporting Information, Figure S21 for other probes).
Although the 1O2 probes formed nano aggregates in
aqueous solutions (pH 7.4, 50 mM HEPES þ 0.1 M KCl þ
10 vol % THF or CH3CN) as observed in the FEꢀSEM
image (Supporting Information, Figure S22), their fluores-
cent 1O2 responses were retained (Supporting Information,
Figure S23).
Supporting Information Available. Experimental details;
scheme depicting synthetic routes to PPBF and PyPBF;
1
figures displaying copies of H and 13C NMR spectra of
1
the O2 probes and their precursor, fluorescence spectra
1
Having established the O2 sensing capabilities of the
obtained without IrF nor photoirradiation, or in the presence
of NaN3, fluorescence 1O2 responses of PPBF and PyPBF,
UVꢀvis absorption changes of the 1O2 probes, the ESIꢀMS
spectra, 13C NMR spectrum of the reaction product of PPBF,
HPLC chromatograms, TDꢀDFT calculation results, cyclic
probes, we performed two-photon fluorescence micro-
scopic experiments to visualize endogenous 1O2 in RAW
264.7 macrophage cells. We chose PyPBF because it
exhibited the best cell permeability. PyPBF was nontoxic to
the cells as judged from the bright-field images (Figure 3).
RAW 264.7 cells treated with 10 μM PyPBF (30 min) pro-
duced intensefluorescenceemission atλems = 490ꢀ568 nm
(channel 2) under two-photon excitation (λex = 750 nm),
whereas the fluorescence emission due to the 1,2-diketone
of PyPBF at λems = 377ꢀ481 nm (channel 1) was barely
detected (Figure 3). Subsequent stimulation of the cells by
2 μM phorbol myristate acetate (PMA)31 provoked a
1
voltammograms, O2 selectivity of PPBF and PyPBF, the
FEꢀSEM image of the aggregates of PyPBF, fluorescent 1O2
responses and selectivity of PyPBF in buffer solution, calcu-
lated fluorescence intensity ratios for the fluorescence micro-
graphs, more cell images, and the fluorescent 1O2 responses of
StPBF at different probe concentrations; tables listing the
summary of TDꢀDFT calculation results and the Cartesian
coordinates of the optimized geometries. This material is
available free of charge via the Internet at http://pubs.acs.org.
(31) Kanofsky, J. R.; Hoogland, H.; Wever, R.; Weiss, S. J. J. Biol.
Chem. 1988, 263, 9692–9696.
(32) Tomita, M.; Irie, M.; Ukita, T. Biochemistry 1969, 8, 5149–5160.
The authors declare no competing financial interest.
D
Org. Lett., Vol. XX, No. XX, XXXX