Far-red fluorescence probe for monitoring singlet oxygen during photodynamic therapy

Article abstract of DOI:10.1021/ja504279r

Singlet oxygen (¹O₂), molecular oxygen in the lowest excited state, has a critical role in the cell-killing mechanism of photodynamic therapy (PDT). Although ¹O₂ phosphorescence measurement has been mainly used to monitor ¹O₂ formation during PDT, its intensity is far insufficient to obtain two-dimensional images of intracellular ¹O₂ with the subcellular spatial resolution using the currently available near-IR detector. Here, we propose a new far-red fluorescence probe of ¹O₂, namely, Si-DMA, composed of silicon-containing rhodamine and anthracene moieties as a chromophore and a ¹O₂ reactive site, respectively. In the presence of ¹O₂, fluorescence of Si-DMA increases 17 times due to endoperoxide formation at the anthracene moiety. With the advantage of negligible self-oxidation by photoirradiation (θ_δ < 0.02) and selective mitochondrial localization, Si-DMA is particularly suitable for imaging ¹O₂ during PDT. Among three different intracellular photosensitizers (Sens), Si-DMA could selectively detect the ¹O₂ that is generated by 5-aminolevulinic acid-derived protoporphyrin IX, colocalized with Si-DMA in mitochondria. On the other hand, mitochondria-targeted KillerRed and lysosomal porphyrins could not induce fluorescence change of Si-DMA. This surprising selectivity of Si-DMA response depending on the Sens localization and photosensitization mechanism is caused by a limited intracellular ¹O₂ diffusion distance (～300 nm) and negligible generation of ¹O₂ by type-I Sens, respectively. For the first time, we successfully visualized ¹O ₂ generated during PDT with a spatial resolution of a single mitochondrial tubule.

Full text of DOI:10.1021/ja504279r

Article
pubs.acs.org/JACS
Far-Red Fluorescence Probe for Monitoring Singlet Oxygen during
Photodynamic Therapy
Sooyeon Kim, Takashi Tachikawa, Mamoru Fujitsuka, and Tetsuro Majima*
The Institute of Scientific and Industrial Research (SANKEN), Osaka University, Mihogaoka 8-1, Osaka, Ibaraki 567-0047, Japan
*
S Supporting Information
1
ABSTRACT: Singlet oxygen ( O ), molecular oxygen in the
2
lowest excited state, has a critical role in the cell-killing mechanism
1
of photodynamic therapy (PDT). Although O phosphorescence
2
1
measurement has been mainly used to monitor O formation
2
during PDT, its intensity is far insufficient to obtain two-
1
dimensional images of intracellular O with the subcellular spatial
2
resolution using the currently available near-IR detector. Here, we
1
propose a new far-red fluorescence probe of O , namely, Si-DMA,
2
composed of silicon-containing rhodamine and anthracene
1
1
moieties as a chromophore and a O reactive site, respectively. In the presence of O , fluorescence of Si-DMA increases 17
2
2
times due to endoperoxide formation at the anthracene moiety. With the advantage of negligible self-oxidation by
1
photoirradiation (Φ < 0.02) and selective mitochondrial localization, Si-DMA is particularly suitable for imaging O during
Δ
2
1
PDT. Among three different intracellular photosensitizers (Sens), Si-DMA could selectively detect the O that is generated by 5-
2
aminolevulinic acid-derived protoporphyrin IX, colocalized with Si-DMA in mitochondria. On the other hand, mitochondria-
targeted KillerRed and lysosomal porphyrins could not induce fluorescence change of Si-DMA. This surprising selectivity of Si-
1
DMA response depending on the Sens localization and photosensitization mechanism is caused by a limited intracellular O
2
1
diffusion distance (∼300 nm) and negligible generation of O by type-I Sens, respectively. For the first time, we successfully
2
1
visualized O generated during PDT with a spatial resolution of a single mitochondrial tubule.
2
INTRODUCTION
superoxide anion can be generated by either type-I, type-II, or
secondary biological processes after photoirradiation, such as
mitochondrial damage. As summarized in previous reviews,
■
1
Singlet oxygen ( O ) is molecular oxygen in the lowest excited
2
2,4−6
1
1
state. In biological and medical studies, O is of a particular
2
the superoxide anion often acts as a “primary” reactive oxygen
importance because of its key role in photodynamic therapy
PDT), an emerging anticancer treatment using photo-
irradiation and photosensitizers (Sens). In PDT, O is the
first molecule immediately generated after photoirradiation in
the aerobic condition, resulting in direct oxidation of nearby
biomolecules, shutdown of the vascular system, and finally
inflammation or immune reaction to tumor tissue. Hence,
investigating the formation and diffusion dynamics of O is
species (ROS) because they can be further converted into other
(
1
2
1
ROS such as H O2 and O2 through various biological
2
2
reactions.
1
To detect O , phosphorescence measurement has been
generally used in PDT.
escence measurement, intracellular O lifetime (τ ) is found to
2
1
,7,8
From time-resolved phosphor-
3
1
2
Δ
1
be varied from a few tens of nanoseconds to 3 μs, depending on
2
2
,7,9
1
highly necessary to understand anticancer mechanisms of PDT
at the molecular level.
the experimental conditions.
Undoubtedly, monitoring O
2
with its luminescence provides the most direct information on
1
Cytotoxic mechanisms of Sens in PDT can be explained as
O formation and further behaviors. Unfortunately, however,
2
1,2
type-I and type-II. Type-II photosensitization includes all
the intensity of phosphorescence is typically much weaker than
kinds of reactions where the excited Sens is quenched by a
that of fluorescence. In addition, to obtain time trajectories of
3
1
collision with oxygen in the ground state ( O ). Meanwhile, in
O with good signal-to-noise ratio, the exchange of intra-
2
2
3
type-I photosensitization, adjacent molecules (except O2)
cellular liquid from H O to D O is mostly required. Thus, real-
2
2
quench the excited Sens often through electron transfer. Upon
1
time mapping (e.g., subsecond acquisition time) of O2
formation at the subcellular spatial resolution is extremely
1
photoirradiation of injected photosensitizers, a majority of O2
is produced via triplet−triplet intermolecular energy transfer
challenging using the currently available near-IR detector and
3
3
(
(
TTEnT) from Sens in the triplet excited state ( Sens*) to O
10,11
2
under physiological conditions.
Considering the current
eq 1).
1
problems, the fluorescence detection of O2 is ultimately
required to trace O formation and its dynamics at the
1
3
*
3
1
1
2
Sens + O → Sens + O
(1)
2
2
This is a typical reaction of 1O2 formation via type-II
photosensitization. On the other hand, it is notable that the
Received: April 29, 2014
©
XXXX American Chemical Society
A
dx.doi.org/10.1021/ja504279r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
subcellular level under physiological conditions (i.e., H O-based
culture media).
time and to distinguish different localizations of Sens and
photosensitization mechanisms.
2
Despite continuous efforts on the development of a
1
fluorescent probe for O2 detection (see examples in
EXPERIMENTAL SECTION
12−15
■
Supporting Information Figure S1),
enabled imaging O in living cells.
Oxygen Sensor Green (SOSG), a widely used commercial O2
probe, exhibits several critical weak points in live cell imaging:
self-oxidation upon photoirradiation, cell impermeability (or
nonspecific staining at the relatively high incubating concen-
tration; see Figure S2 and ref 21), and a necessity of green
excitation (λ = 508 nm) where a cell autofluorescence signal
To the best of our knowledge, only Song et al.
and Dai et al. have succeeded in monitoring intracellular O2
produced during photoirradiation of Sens using the fluorescent
probe based on europium complexes. Surprisingly, however,
there has been no adequate O fluorescence probe to monitor
intracellular O during PDT at the subcellular level. The
drawbacks of SOSG and lack of a suitable and successful
organic probe for O mapping have inspired us to develop a
new far-red fluorescent probe for the O detection.
In this study, we have synthesized a far-red fluorescent probe
merely a few cases
1
16−20
Materials. All chemical reagents and solvents used for the synthesis
were purchased from Sigma-Aldrich Chemical Co., Tokyo Chemical
Industries, Nacalai Tesque, and Wako Pure Chemical, and used as
received. 2-Methylbenzene-substituted Si-rhodamine (Si-Me, Figure 1)
was synthesized according to the previous report by Koide et al., and
Si-DMA was synthesized as described in Supporting Information. Si-
DMA and Si-Me were stored in dimethyl sulfoxide (DMSO) at −20
°C in the dark. We chose the following Sens for living cell
experiments: 5-aminolevulinic acid (5-ALA, Sigma-Aldrich), a
precursor of heme, and tetra-(N-methyl-4-pyridyl)porphyrin
TMPyP4, Tokyo Chemical Industries). SOSG, CellROX Green,
MitoTracker Green FM, and Dextran 10000 MW labeled by Alexa
Fluor 488 or Alexa Fluor 647 (A488- and A647-dextran, respectively)
were purchased from Molecular Probes and used for O and ROS
detection, mitochondrial and lysosomal markers, respectively.
Ex Vivo Steady-State Measurements. Without a notation, all
bulk spectroscopic measurements were performed in methanol at the
spectroscopic grade or pH 7.5 phosphate buffered saline (PBS)
solution without Ca and Mg (both purchased from Nacalai Tesque).
Ground-state absorption and fluorescence emission spectra were
measured using a Shimadzu UV-3100 and Horiba FluoroMax-4,
respectively. Bulk irradiation was performed using a xenon source
Moreover, Singlet
2
1
23
abs
2
1,22
16
is induced.
1
7
1
(
1
1
2
2
1
2
1
2
1
2
1
of O , composed of 9,10-dimethylanthracene (DMA) and
silicon-containing rhodamine (Si-rhodamine) moieties, namely,
Si-DMA (Figure 1). Recently, Si-rhodamine has been actively
2
(
LAX-C100, Asahi Spectra) and band-pass filter (BA510-550,
Olympus).
Time-Resolved Phosphorescence Measurement. The samples
of Rose Bengal and Si-Me were prepared in 1:9 DMSO/MeOH
3
solution in a 1 × 1 × 4 cm quartz cell. The second-harmonic
−2
−1
oscillation (532 nm, 4 ns fwhm, 5.0 mJ cm pulse ) from a Q-
switched Nd:YAG laser (Continuum, Surelite II-10) was used for the
excitation light. The photoinduced luminescence from the sample cell
was collected with quartz lenses, passed through a monochromator,
and then introduced into a near-IR photomultiplier tube module
(
Hamamatsu Photonics, H10330A-75). After being amplified by a 350
MHz amplifier unit (Stanford Research, SR445A), the output of the
photomultiplier was sent to a gated photon counter (Stanford
Research, SR400) under direct control from a PC via the GPIB
1
interface. To measure the lifetime of O , the signals were accumulated
2
(
five repetitions) by changing the delay time from 0 to 50 μs with a
gate width of 1.0 μs.
Cells and Cell Culture. HeLa cells and RAW 264.7 macrophages
were kindly provided by the RIKEN BRC through the National Bio-
Resource Project of the MEXT, Japan, and Prof. Tsuyoshi Nishi
(SANKEN, Osaka University), respectively. Without a notation, cell
experiments carried out in this study were performed using HeLa cells.
HeLa cells and RAW 264.7 macrophages were cultivated in Dulbecco’s
modified Eagle medium (D6429, Sigma) supplemented with 10% fetal
bovine serum (10099-141, Gibco) at 37 °C in a humidified incubator
under 5% CO2.
Figure 1. Chemical structures of Si-DMA, its peroxidized product (Si-
DMEP), and Si-Me.
Live Cell Imaging. To monitor fluorescence increase of Si-DMA
during photoirradiation of Sens, an Olympus IX81 inverted
fluorescence microscope and a 640 nm CW laser (Coherent) were
used to irradiate Sens and monitor the probe simultaneously. A 35 mm
μ-dish with a glass bottom (ibidi) with HeLa cells was excited through
an oil objective (Olympus, PlanApo 100×/1.40 oil). The emission
image was collected with the same objective and recorded by an
EMCCD camera (Roper Scientific, Evolve 512) through a dichroic
beamsplitter (Semrock, Di02-R635) and a band-pass filter (Chroma,
HQ690/70). During data acquisition, the same environment as an
2
3−26
applied for the detection of various biological substances
27
and super-resolution imaging both in vitro and in vivo. In
light of the previous works, we affirmed that Si-rhodamine is a
promising far-red chromophore for the intracellular 1O2
mapping. Here, Si-rhodamine is found to exhibit 3 times
1
incubator (37 °C and 5% CO
Live Cell Instrument). Pseudocolor fluorescence images were
prepared by reprocessing the obtained movie file using OriginPro
2
) was maintained using a Chamlide TC
lower quantum yield of O generation (Φ ) than that of 2,7-
dichlorofluorescein, a chromophore of SOSG. Furthermore, Si-
2
Δ
(
DMA can selectively reside in mitochondria and provide an
9
.1 (OriginLab) and ImageJ.
To confirm the localization of dyes, in particular, cellular organelles,
1
instantaneous response to intracelluar O . Compared to the
2
1
16−19
previously reported intracellular O2 probes,
Si-DMA
2
an objective scanning confocal microscope (PicoQuant, MicroTime
200) coupled with an Olympus IX71 inverted fluorescence microscope
1
shows outstanding results to visualize intracellular O in real-
B
dx.doi.org/10.1021/ja504279r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 2. Absorbance and fluorescence changes of Si-DMA in methanol (a,b) and in PBS solution (c) upon photoirradiation of codissolved
TMPyP4 for 50 min (black to red). Black arrows in (a,c) indicate absorbance changes due to peroxidation of DMA (bold lines) and dissociation of
H-aggregates (dashed lines), while a red arrow in (b) shows fluorescence increase during 50 min irradiation. (d) Si-DMA fluorescence increase
−2
during 50 min photoirradiation. [Si-DMA] = 50 μM, [TMPyP4] = 5 μM, and 510−550 nm irradiation at 0.07 W cm with magnetic stirring was
1
used for O formation.
2
1
was used. An 8-well μ-slide (ibidi) with HeLa cells was excited at two
wavelengths, 405/640 nm or 485/640 nm, using a pulsed laser
in methanol) by reacting with O and forming endoperoxide
2
on the center ring of DMA because PET from DMA to Si-
rhodamine is prohibited (Figure 1). Approximately, 18-fold
increase in fluorescence was observed in MeOH (Figure 2b,d,
red), accompanying a disappearance of anthracene absorbance
black arrows, Figure 2a). Similar spectral changes were
observed when O was induced by a chemical catalyst (Figure
S4). In addition, it has been reported that some anthracene
derivatives, such as arylanthracene and alkynylanthracene,
undergo a reversible reaction with O2 upon increasing
(
PicoQuant) controlled by a PDL-800B driver (PicoQuant) through
an oil objective (Olympus, UPlanSAPo 100x/1.40 oil/0.17/FN26.5).
Subsequently, the emission was collected with the same objective and
detected by a single photon avalanche photodiode (Micro Photon
Devices, PDM 50CT and 100CT) through a beam splitter (90%
transmission, 10% reflection), suitable band-pass filters, and 75 μm
pinhole for spatial filtering to reject out-of-focus signals. Two images
for green and red channels were further processed to obtain a merged
image using OriginPro 9.1 (OriginLab) and ImageJ.
(
1
2
1
temperature, that is, a formation of endoperoxide and release
29
RESULTS AND DISCUSSION
of O2. Nonetheless, the DMA moiety in Si-DMA does not
release O after endoperoxidation but is probably prone to be
decomposed at temperatures above 70 °C, which is much
■
Synthesis and Optical Properties of Si-DMA. As
2
23
27
proposed by Nagano et al. and Johnsson et al., we first
synthesized Si-containing xanthone, and then connected it to 2-
bromo-9,10-dimethylanthracene, which is a reactive site for
21
greater than that at physiological conditions.
Interestingly, Si-DMA is found to form H-aggregates in the
aqueous solution (Figure 2c). This is because substituting an
anthracene moiety reduces the hydrophilicity of Si-DMA as
compared to Si-Me, which is a monomer in PBS solution at the
same concentration (5 μM). A blue-shifted absorption peak at
625 nm is clear evidence to indicate the H-aggregates of
1
1
O , resulting in Si-DMA (19% yield). O -induced conversion
2
2
of Si-DMA to Si-DMEP is illustrated in Figure 1. Detailed
synthesis description and characterization can be found in
Supporting Information.
Figure 2 shows time-dependent changes in absorbance and
30−32
fluorescence of Si-DMA upon photoirradiation of coincubated
rhodamine chromophores.
This blue-shifted shoulder
1
TMPyP4. In the absence of O , the excited Si-rhodamine is
gradually decreases as Si-DMEP forms (Figure 2c, red
arrow), implying that endoperoxide formation at the
anthracene ring disrupts a stacking of rhodamine fluorophores.
In Figure 2d, the slower rate of Si-DMEP formation in the case
of H-aggregates compared to that of monomeric Si-DMA is
presumably caused by (1) a decrease in the absolute
chromophore concentration due to the aggregate formation
2
quenched by photoinduced electron transfer (PET) from
DMA, resulting in dim fluorescence (Φ = 0.01 in methanol).
Polarity-dependent fluorescence quenching of Si-DMA sup-
ports that Si-DMA is indeed deactivated by PET (Figure S3).
This intramolecular PET and subsequent charge recombination
process in Si-DMA are considered to be similar to the
fl
2
2,28
previously reported xanthene−anthracene dyads.
more, Si-DMA converts into bright form (Si-DMEP, Φ = 0.17
Further-
fl
and (2) necessity of aggregate dissociation by (or for) reacting
1
with O . A similar result is also obtained in the presence of
2
C
dx.doi.org/10.1021/ja504279r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
1
chemically prepared O (Figure S5). Furthermore, a larger
2
fluorescence increment observed in PBS solution (22-fold)
than that in MeOH (18-fold) is probably due to additional
fluorescence quenching upon H-aggregate formation before
photoirradiation. Overall, both monomeric and H-aggregates of
1
Si-DMA can react to O , while Si-DMEP does not favor the
2
aggregate formation due to a loss of anthracene planarity.
1
Evaluations of Si-DMA as a O Probe. Si-DMA is tested
2
1
to confirm whether it can respond to O quantitatively and
2
selectively without self-oxidation of the dye. First, both Si-DMA
monomer and H-aggregates show linear fluorescence increase
1
versus [ O ] generated by NaClO/H O (Figure S5).
2
2
2
1
Furthermore, Si-DMA exhibits a good selectivity toward O2
out of other ROS, which agrees with the previous anthracene-
1
12,13,16
1
based O probes (Figure 3).
In addition, no significant
Figure 4. Time profiles of O
2
phosphorescence intensity (P.I.) of
2
Rose Bengal (black) and Si-Me (red) observed at 1280 nm (λ = 532
nm at 5.0 mJ cm pulse ) in 1:9 DMSO/MeOH solution. Both
samples were prepared to have the same absorbance of 0.5 at 532 nm
fluorescence decrease was observed in various ROS solutions,
indicating chromophore stability of Si-DMA against ROS.
ex
−2
−1
(path length = 1 cm). [Rose Bengal] = 21.5 μM and [Si-Me] = 250
μM.
Mitochondrial Localization of Si-DMA. Positive net
3
5,36
charge of a dye, such as that of rhodamine 123 (R123)
lipophilic cyanine cation, JC-1,
and
3
7,38
facilitates its mitochondrial
accumulation due to mitochondrial membrane potential.
39
Here, we found that Si-DMA can reside in mitochondria with
high selectivity (Figure 5). The long, thin, and branched fibrillar
Figure 3. Selectivity of Si-DMA toward 1O2 among other ROS.
Maximum fluorescence intensity divided by the absorbance at 640 nm
was compared. [Si-DMA] = 5 μM and [ROS] = 10 mM in 1:1
MeOH/PBS solution, stirred for 10 min.
As briefly mentioned earlier, one critical weak point of SOSG
1
is its non-negligible O generation due to the excitation of the
2
2
2
chromophore moiety. Even though most of the SOSG
molecules are deactivated via ultrafast intramolecular PET (k
PET
11
−1
=
9.7 × 10 s ), trivial impurities of initially bright SOSG,
33
such as nonsubstituted fluorescein (∼0.3%) or peroxidized
SOSG molecules, cannot be deactivated via PET and are in
charge of O2 generation upon continuous photoirradia-
1
Figure 5. HeLa cells stained with [Si-DMA] = (a) 20 and (b) 100 nM
for 1 h. (c−e) Colocalization test of Si-DMA in mitochondria. (c)
MitoTracker Green and (d) [Si-DMA] = 100 nM. Clear
colocalizations of two dyes are observed in the merged image (e).
Intensities of each image have been adjusted to obtain a clear picture.
Scale bar = 10 μm.
2
1,22,33
1
tion.
To observe O produced during PDT that often
2
requires prolonged photoirradiation from a few minutes to
hours, self-oxidation of the dye by light excitation should be
avoided. In order to confirm that Si-rhodamine is an
1
appropriate chromophore as a O probe, Φ of Si-Me was
2
Δ
determined using time-resolved phosphorescence measurement
(
Figure 4). Using Rose Bengal as a reference (Φ = 0.76 in
Δ
3
4
MeOH), Φ of Si-Me is approximately 0.02. This value is
fluorescence image of Si-DMA clearly proves its selective
Δ
40
considered to reflect the possible maximum Φ of Si-DMA and
localization in mitochondria. In addition, it does not have a
strong preference to reside in mitochondria around the nucleus
or edge of the cell, different from 5-ALA-derived PpIX
(discussed later). On the other hand, concentrations higher
than 1 μM seem to result in nonspecific localization and even
higher than 5 μM shows cytotoxicity determined by vacuole
formation and cell detachment (Figure S7). The latter effect is
due to the loss of mitochondrial membrane potential upon
Δ
Si-DMEP when deactivation pathways via PET is completely
blocked. Furthermore, Φ of Si-Me is 3 times smaller than that
Δ
of 2,7-dichlorofluorescein, a fluorophore of SOSG (Φ
=
Δ
2
2
0
.06).
Finally, Si-DMA is considered to be fairly photostable, as
shown in Figure 2d (black). During 50 min irradiation, no
photobleaching is observed in the absence of photosensitizer.
In addition, this photostability is maintained when Si-DMA is
introduced to HeLa cells (Figure S6). This result is in
39
excess loading of Si-DMA. Hence, in this study, we chose [Si-
DMA] less than 100 nM to ensure the homogeneous, selective,
and nontoxic staining of mitochondria (Figure 5a,b).
24
accordance with the previous study on Si-rhodamine.
D
dx.doi.org/10.1021/ja504279r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
It is notable that connecting DMA to Si-rhodamine does not
used type-II Sens in a model study of PDT because it is water-
affect its intrinsic localization, whereas azo-benzene-substituted
soluble and can bind to a DNA aptamer, finally delivered to the
26
49−51
Si-rhodamine is mainly localized in the lysosome. Meanwhile,
SOSG, a trianionic molecule, stains HeLa cells in a nonspecific
and inefficient manner (Figure S2). More importantly,
mitochondria are the main target organelles in PDT because
mitochondrial actions and dysfunctions are closely related to
nucleus of cancer cells.
Prior to coincubating with Si-
DMA, we confirm that TMPyP4 first resides in lysosomes after
24 h incubation (Figure 6b). Next, 640 nm irradiation triggers
lysosomal rupture, followed by relocalizations of TMPyP4 to
the cytoplasm and nucleus (Figure S10). Relocalization of Sens
in lysosomes is a well-known phenomenon in PDT, which
41−43
the programmed cell death.
Thus, a selective mitochon-
1
drial staining of Si-DMA indeed takes advantage of visualizing
maximizes a cytotoxic effect of O2 from dispersed Sens
1
43,52
intracellular O particularly during PDT.
molecules.
2
1
Fluorescence Detection of Intracellular O : Depend-
Since fluorescence of TMPyP4 is not intense enough (Φ =
2
fl
53
ent on the Localization of Sens. To deliver Sens selectively
0.047 in H O) and Q-band absorbance at 640 nm is relatively
2
to mitochondria, we chose 5-ALA, which is a commercialized
PDT drug as a precursor for the biosynthesis of PpIX.
small, lysosomal rupture events cannot be monitored by
TMPyP4 fluorescence with the same photoirradiation and
detection conditions used to image Si-DMA. Thus, we
additionally stained lysosomes using A647-dextran to observe
intracellular events and fluorescence changes of Si-DMA
simultaneously with the excitation of TMPyP4. As depicted
by the white arrows in Figure 7b, lysosomes containing A647-
dextran and TMPyP4 are distinguishable due to their granular
structures and strong fluorescence. Upon 640 nm irradiation for
2
,42,44−46
The excess amount of biosynthesized PpIX, which could not be
bound to Fe to form heme, is accumulated in mitochondria,
and this phenomenon readily happens especially in cancer cells.
When the localization of PpIX derived by [5-ALA] = 150 μg
−
1
mL was traced, the synthesized PpIX was found to be initially
localized in mitochondria and soon diffused to the cytoplasm
(Figure 6a). In particular, PpIX is predominantly localized in
1
0 s, lysosomal rupture and TMPyP4 diffusion substantially
occur. However, fluorescence of Si-DMA does not increase
even after TMPyP4 relocalization (Figures 7b and S9).
In short, the heterogeneous fluorescence increase shown in
Figure 7a is considered to reflect the localization of PpIX
because Si-DMA is localized in mitochondria almost homoge-
neously over the cell body (Figure 5). In contrast, Si-DMA
1
does not react with O generated by TMPyP4 (Figure 7b),
2
which changes its location from lysosome to cytoplasm and
Figure 6. Intracellular localizations of Sens. (a) Fluorescence of
MitoTracker Green, 5-ALA-derived PpIX, and merged points are
depicted in green, red, and yellow, respectively. (b) Fluorescence of
A488-dextran, TMPyP4, and merged points are depicted in green, red,
and yellow, respectively. (c) Fluorescence of KillerRed monitored after
nucleus upon photoirradiation (Figure S10). Taken together,
1
Si-DMA can respond to only mitochondrial-originating O .
2
This is because of the short diffusion distance of intracellular
1
9
O (at longest, approximately 300 nm) in aqueous solution.
2
2
=
4 h transfection of mitochondria-targeted KillerRed vector. Scale bar
10 μm.
1
Fluorescence Detection of Intracellular O : Depend-
2
ent on the Type of Sens. We investigated Si-DMA responses
depending on the type of Sens. In the previous section, 5-ALA-
derived PpIX and TMPyP4 were used, and both are standard
4
2,47,48
the perinuclear region as reported previously.
In light of
34
the previous reports, 5-ALA was incubated for 4 h in order to
type-II Sens with high Φ (0.54 and 0.74, respectively). As a
Δ
4
4,45
maximize the amount of synthesized PpIX (Figure S8).
comparison, KillerRed was chosen as an exemplary case of
photosensitizing protein, and it could be expressed selectively
Figure 7a shows reprocessed pseudocolor images of Si-DMA
fluorescence in HeLa cells during 640 nm irradiation of PpIX. It
is clearly shown that photoirradiation of PpIX induces an
increase of Si-DMA fluorescence rapidly (within 10 s) and
more significantly in the perinuclear region as compared to the
cell edge. The same tendency is observed when the incubation
concentration of Si-DMA is increased to 100 nM (Figure S9).
54
in mitochondria (Figure 6c). Upon photoirradiation, the
chromophore of KillerRed, the N-acylimine group composed of
Gln-Tyr-Gly, is excited and subsequently quenched by one-
electron reduction from the adjacent amino acid residue,
resulting in the radical anion of the chromophore (type-I
photosensitization). Its characteristic phototoxicity originates
Here, an evident fluorescence increase along the mitochondria
1
55,56
structure provides a much clearer picture of intracellular O
2
1
generation as compared to the previous O probes with 5-
2
1
5,17
ALA-derived PpIX.
It should be confirmed that the fluorescence increase of Si-
Pseudocolor images of Si-DMA and KillerRed fluorescence
(λem = 610 nm), collected by the same detection channel
1
54
DMA shown in Figures 7a and S7 is caused by O generation
2
and not by environmental changes of mitochondria upon PDT.
We have monitored approximately 4−6-fold deceleration in the
initial rates of fluorescence increase by the addition of a cell-
(655−725 nm), are shown in Figure 7c. Since λ of KillerRed
abs
54
is 585 nm, two-color excitation at 532 and 640 nm is used to
excite KillerRed and Si-DMA, respectively. In fact, KillerRed
fluorescence is not negligible compared to that of Si-DMA (red
spots in the 0 s image of Figure 7c). During 10 min
photoirradiation, however, only photobleaching of both species
was observed without fluorescence increase of Si-DMA
1
permeable O quencher, sodium azide (NaN ) (Figure 8).
2
3
Furthermore, mitochondrial environmental changes, such as
viscosity, did not influence the fluorescence of Si-DMA (Figure
S12a).
Subsequently, another type-II Sens that exhibits non-
mitochondrial localization is introduced to monitor fluores-
cence response of Si-DMA. TMPyP4 is one of the frequently
(Figures 7c and S9). As shown in Figure 3, Si-DMA can only
1
respond to O but not superoxide, hydrogen peroxide, and
2
1
other ROS. Thus, this result implies that the amount of O
2
E
dx.doi.org/10.1021/ja504279r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 7. Pseudocolor fluorescence images of HeLa cells with Si-DMA and (a) 5-ALA-induced PpIX, (b) 24 h incubated TMPyP4 and lysosome
marker (A647-dextran), and (c) mitochondria-targeted KillerRed. Granular structures indicated by white arrows in (b) are lysosomes that are rapidly
ruptured by TMPyP4 irradiation. [Si-DMA] = 25 nM was incubated for 1 h in the HeLa cells containing corresponding Sens. (a,b) One-color
−
2
1
irradiation at 640 nm (0.6 W cm ) was used to generate O and monitor Si-DMA fluorescence simultaneously. Meanwhile, two-color irradiation
2
−2
was required to excite Si-DMA and KillerRed (c); 0.5 and 0.6 W cm were used for 532 and 640 nm irradiation, respectively. Scale bar = 10 μm.
1
O Diffusion Distance and Its Concentration Affect
2
Fluorescence Increase of Si-DMA. To understand different
responses of Si-DMA to three Sens (Figure 7), we first
1
considered the total cumulative [ O ] during photoirradiation
2
of Sens. The extent of PDT cytotoxic effects, so-called PDT7
dose, is influenced by three factors: light, oxygen, and drug.
Here, we used the same light intensity of 640 nm irradiation
(
except KillerRed), and there is only a slight difference in Q-
band absorbance of PpIX and TMPyP4. Furthermore, the
previous study found merely a marginal difference of ΔpO
2
between mitochondria or endosomes and air-saturated medium
4
4,58
in HeLa cells.
Overall, nearly identical light doses and
homogeneous pO over a whole cell infer that light dose and
2
3
[
O ] are not conclusive factors to derive the different results in
Figure 8. Fluorescence changes of Si-DMA during 15 s of 640 nm
irradiation at 0.6 W cm with PpIX induced by 5-ALA (black), with
2
−2
Si-DMA fluorescence, as shown in Figure 7.
1
the addition of [NaN ] = 2 mM (red), 1 mM (blue), and Si-DMA
Next, we consider the effective area of O within its lifetime.
3
2
without Sens and NaN (green). All data are obtained by averaging
Time-resolved phosphorescence measurement is the most
direct method to study formation and decay of intracellular
3
mean count changes monitored in 5−8 cells.
1
1,10
1
O . Although the reported lifetime of O varies depending
2
2
on the experimental conditions (a few tens of nanoseconds to 3
2
,7,9
1
generated by KillerRed is far insufficient to induce a conversion
of Si-DMA to Si-DMEP even though KillerRed is colocalized in
mitochondria. On the other hand, photoirradiation of KillerRed
causes relocalization and DNA binding of CellROX Green,
which infers the generation of oxygen radical species (Figure
S13). Generation of a superoxide anion in a proportional
manner upon photoirradiation of KillerRed is in good accord
μs),
the longest travel distance of intracellular O has been
2
7
reported to be ∼268 nm over a period t of twice its lifetime.
Furthermore, a diameter of mitochondria fibril is ∼200−400
40
nm, and both Si-DMA and accumulated PpIX are considered
39,46
to be located inside the mitochondrial inner membrane.
Such a complete colocalization within a diffusible area of O is
regarded as the main reason for a highly localization-selective
response of Si-DMA. On the other hand, O diffusion across
1
2
57
1
with the previous report.
2
F
dx.doi.org/10.1021/ja504279r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
different intracellular organelles is theoretically impossible as
dimmer fluorescence of Si-DMA (Figure S14). Simultaneously,
vacuole formation and granular appearance of cell structures
were monitored (data are not shown here but are identical to
Figure 1 of ref 48). Moreover, fluorescence of Si-DMA upon
photoirradiation rarely increased in dying or dead cells. This is
considered to be due to (1) defocusing caused by cell
morphology change, (2) leak of Si-DMA upon a loss of
proven by imaging of Si-DMA fluorescence (Figure 7b).
1
In the case of KillerRed, the inefficient O generation of
2
KillerRed is proven by its photocytotoxicity that is rather higher
5
9
in H O than in D O and silence at 1270 nm upon
2
2
photoirradiaiton where 1O2 phosphorescence can be ob-
served. Instead, hydrogen peroxide and a superoxide anion
5
7
1
48
can form O as a byproduct of the Haber−Weiss reaction (i.e.,
mitochondrial membrane potential, and (3) increase in
2
•
−
1
•
−
1
H O + O
→ O + OH + OH). Vegh et al., however,
overall intracellular viscosity, making a collision between O
2
2
2
2
2
60
reported that KillerRed exhibits the quantum yield of hydrogen
peroxide generation less than 1% because the generated
superoxide anion is quenched inside of the protein by amino
and Si-DMA difficult. In this sense, slight fluorescence
decrease of Si-DMA shown in Figure 7b (Sens: TMPyP4) is
1
not caused by
O -derived photo-oxidation but rather by
2
5
7
acid residues. In fact, even if the superoxide anion is largely
intracellular environmental change as described above. With
further systematic and quantitative studies on the correlation
among the fluorescence change of Si-DMA, PDT dose, and cell
death pathways (e.g., apoptosis and necrosis), Si-DMA will
possibly work as a barometer to report acute cytotoxicity and
cell viability during PDT.
generated in a living cell, various enzymatic systems such as
3
5,6
superoxide dismutase will quench superoxide to O . In
2
accordance with this result, we could not monitor the
fluorescence increase of Si-DMA under excessive ROS
generation in mitochondria induced by an oxidative burst in
RAW 264.7 macrophage (Figure S12c; the detailed discussion
CONCLUSION
In this study, we have proposed a new and promising far-red
fluorescent probe, Si-DMA, for the intracellular O detection.
can be found in Supporting Information). This finding further
■
1
supports that Si-DMA maintains its high selectivity to O in a
2
1
living cell, while it does not respond to the trace amount of
2
1
intracellular O . Obtained results in this study are summarized
As a dyad of Si-rhodamine and DMA, Si-DMA is present as a
monomer and H-aggregates in organic solvents and neutral
2
in Table 1.
buffered solutions, respectively, while both of the states can
1
1
Table 1. Summary of Si-DMA Responses to Intracellular O
in the Four Experimental Designs
react with O . A cationic charge and adequate lipophilicity of
2
2
Si-DMA facilitate the selective mitochondrial localization. In
the cell experiment, fluorescence of Si-DMA increases only by
photoirradiation of colocalized and type-II Sens, PpIX. On the
other hand, type-II Sens with non-mitochondrial localizations,
mitochondria-targeted KillerRed, and an oxidative burst of the
1
mitochondrial origin
O generation
Si-DMA
fluorescence
2
1
entry
PpIX
of O ?
mechanism
2
a
yes
no
TTEnT
increase
a
TMPyP4
KillerRed
TTEnT
no change
no change
no change
macrophage fails to change Si-DMA fluorescence (Table 1).
yes
yes
electron transfer
catalytic reaction
1
4,8
This result confirms a short diffusion distance of O and
2
oxidative
burst
1
substantial generation of O by type-II Sens (Φ of PpIX =
2
Δ
1
a
0.54). This is the first example of a O fluorescence probe to
suggest a clear image of O formation during PDT at the
2
TTEnT: triplet−triplet intermolecular energy transfer.
1
2
subcellular level. To conclude, we would like to emphasize the
potential of Si-rhodamine derivatives to be exploited in the
single-molecule imaging of intracellular O2 during PDT.
Potential Usages of Si-DMA. To conclude, other potential
1
usages of Si-DMA are proposed in addition to visualization of
1
mitochondrial O during PDT. Using a newly designed far-red
Unfortunately Si-DMA exhibits a limitation in this application
2
1
fluorescence probe, Si-DMA, we succeeded in increasing the
because of its dim fluorescence before reacting to O2.
1
1
spatial resolution of O detection up to a single mitochondrial
Prospective candidates of the new O probe are required to
2
2
tubule (Figure 7a). This fluorescence increase was never
monitored in the case of TMPyP4 that resides in lysosomes,
nucleus, and cytoplasm. Considering this highly localization-
selective response, we first suggest that Si-DMA can be used as
mitochondrial markers of Sens to confirm its accurate
internalization. As mentioned above, mitochondria is the
main target organelle in the PDT study because dysfunction
exhibit a completely dark initial state, enough brightness
(extinction coefficient >50 000 M− cm ; Φ > 0.1), good
1
−1
61
fl
cell permeability, and water solubility.
ASSOCIATED CONTENT
Supporting Information
■
*
S
Experimental and synthesis details including dye staining,
41−43
of mitochondria exerts the strongest cytotoxicity.
Practi-
1
KillerRed expression, H NMR and mass spectrum of Si-DMA,
cally, an infinitesimal addition of Si-DMA to the culture media
can quickly stain mitochondria (cf. [Si-DMA] = 20 nM and 30
min incubation was enough to stain). So far, R123 and other
cationic dyes have been used to visualize changes in the
mitochondrial structure and to confirm localization of a
particular substance in mitochondria. In addition to both
usages, Si-DMA can further provide information about whether
and supporting figures. This material is available free of charge
via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
majima@sanken.osaka-u.ac.jp
■
Notes
the mitochondria-targeted type-II Sens works properly in terms
1
The authors declare no competing financial interest.
of O generation.
2
Second, Si-DMA can be presumably used to indicate PDT
dose and cell viability. Upon prolonged photoirradiation at 640
nm (2−4 min), we could observe fragmentation of the
mitochondrial structure and cell shrinkage accompanied by
ACKNOWLEDGMENTS
We are grateful to Prof. A. Sugimoto (SANKEN, Osaka
University) for helpful advice on chemical synthesis, and Prof.
■
G
dx.doi.org/10.1021/ja504279r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
T. Kinoshita and Prof. Y. Maeda (iFReC, Osaka University) for
help with establishing the biological experimental setup. Also,
we acknowledge Prof. W.E. Moerner (Stanford University),
Prof. T. Nishi (SANKEN, Osaka University), Dr. M. Morita
Uchiyama, M.; Morokuma, K.; Nagano, T.; Hanaoka, K. Angew. Chem.,
Int. Ed. 2013, 52, 13028.
(
27) Lukinavici
̌
us, G.; Umezawa, K.; Olivier, N.; Honigmann, A.;
Yang, G.; Plass, T.; Mueller, V.; Reymond, L.; Correa
Schultz, C.; Lemke, E. A.; Heppenstall, P.; Eggeling, C.; Manley, S.;
̂
, I. R.; Luo, Z. G.;
(
Kyoto University), and Dr. J. Seong (iFReC, Osaka
Johnsson, K. Nat. Chem. 2013, 5, 132.
University) for fruitful discussions. This work has been partly
supported by the Innovative Project for Advanced Instruments,
Renovation Center of Instruments for Science Education and
Technology, Osaka University, and a Grant-in-Aid for Scientific
Research (Projects 25220806, 25288035, and others) from the
Ministry of Education, Culture, Sports, Science and Technol-
ogy (MEXT) of the Japanese Government.
(
28) Miura, T.; Urano, Y.; Tanaka, K.; Nagano, T.; Ohkubo, K.;
Fukuzumi, S. J. Am. Chem. Soc. 2003, 125, 8666.
29) Fudickar, W.; Linker, T. J. Am. Chem. Soc. 2012, 134, 15071.
(
(30) Martinez, V.; Arbeloa, F.; Prieto, J.; Lopez, T.; Arbeloa, I. J. Phys.
Chem. B 2004, 108, 20030.
(31) Nasr, C.; Liu, D.; Hotchandani, S.; Kamat, P. V. J. Phys. Chem.
1
(
996, 100, 11054.
32) McCann, T. E.; Kosaka, N.; Koide, Y.; Mitsunaga, M.; Choyke,
P. L.; Nagano, T.; Urano, Y.; Kobayashi, H. Bioconjugate Chem. 2011,
22, 2531.
(33) Ragas, X.; Jimenez-Banzo, A.; Sanchez-Garcia, D.; Batllori, X.;
Nonell, S. Chem. Commun. 2009, 2920.
(34) Redmond, R. W.; Gamlin, J. N. Photochem. Photobiol. 1999, 70,
391.
REFERENCES
■
(
(
1) Ogilby, P. R. Chem. Soc. Rev. 2010, 39, 3181.
2) Advances in Photodynamic Therapy: Basic, Translational, and
Clinical; Hamblin, M. R., Mroz
Medicine & Biology Series; Artech House: Boston, 2008.
3) Castano, A. P.; Mroz, P.; Hamblin, M. R. Nat. Rev. Cancer 2006,
, 535.
́
, P., Eds.; Artech House Engineering in
(
6
(35) Johnson, L. V.; Walsh, M. L.; Chen, L. B. Proc. Natl. Acad. Sci.
U.S.A. 1980, 77, 990.
(
4) Kanofsky, J. R. Chem.−Biol. Interact. 1989, 70, 1.
(36) Galluzzi, L.; Zamzami, N.; de la Motte Rouge, T.; Lemaire, C.;
Brenner, C.; Kroemer, G. Apoptosis 2007, 12, 803.
(37) Smiley, S. T.; Reers, M.; Mottola-Hartshorn, C.; Lin, M.; Chen,
A.; Smith, T. W.; Steele, G. D.; Chen, L. B. Proc. Natl. Acad. Sci. U.S.A.
1991, 88, 3671.
(38) Salvioli, S.; Ardizzoni, A.; Franceschi, C.; Cossarizza, A. FEBS
Lett. 1997, 411, 77.
(39) Skulachev, V. P. Trends Biochem. Sci. 2001, 26, 23.
(40) Schmidt, R.; Wurm, C. A.; Punge, A.; Egner, A.; Jakobs, S.; Hell,
S. W. Nano Lett. 2009, 9, 2508.
(41) Kroemer, G.; Dallaporta, B.; Resche-Rigon, M. Annu. Rev.
Physiol. 1998, 60, 619.
(42) Steinbach, P.; Wedmgandt, H.; Baumgartner, R.; Kriegmair, M.;
̈
̈
Hofstadter, F.; Knuchel, R. Photochem. Photobiol. 1995, 62, 887.
(43) Castano, A. P.; Demidova, T. N.; Hamblin, M. R. Photodiagn.
Photodyn. Ther. 2004, 1, 279.
(44) Mik, E.; Stap, J.; Sinaasappel, M.; Beek, J.; Aten, J.; van
Leeuwen, T.; Ince, C. Nat. Methods 2006, 3, 939.
(45) Atif, M.; Firdous, S.; Khurshid, A.; Noreen, L.; Zaidi, S. S. Z.;
Ikram, M. Laser Phys. Lett. 2009, 6, 886.
(
(
(
5) Winterbourn, C. C. Nat. Chem. Biol. 2008, 4, 278.
6) Dixon, S. J.; Stockwell, B. R. Nat. Chem. Biol. 2014, 10, 9.
7) Jarvi, M. T.; Niedre, M. J.; Patterson, M. S.; Wilson, B. C.
Photochem. Photobiol. 2006, 82, 1198.
8) Kanofsky, J. R.; Hoogland, H.; Wever, R.; Weiss, S. J. J. Biol.
Chem. 1988, 263, 9692.
9) Skovsen, E.; Snyder, J. W.; Lambert, J. D. C.; Ogilby, P. R. J. Phys.
Chem. B 2005, 109, 8570.
10) Celli, J. P.; Spring, B. Q.; Rizvi, I.; Evans, C. L.; Samkoe, K. S.;
Verma, S.; Pogue, B. W.; Hasan, T. Chem. Rev. 2010, 110, 2795.
11) Snyder, J. W.; Skovsen, E.; Lambert, J. D. C.; Poulsen, L.;
Ogilby, P. R. Phys. Chem. Chem. Phys. 2006, 8, 4280.
12) Tanaka, K.; Miura, T.; Umezawa, N.; Urano, Y.; Kikuchi, K.;
Higuchi, T.; Nagano, T. J. Am. Chem. Soc. 2001, 123, 2530.
(
(
(
(
(
(
(
13) Liu, Y.-J.; Wang, K.-Z. Eur. J. Inorg. Chem. 2008, 2008, 5214.
14) Ruiz-Gonzalez, R.; Zanocco, R.; Gidi, Y.; Zanocco, A. L.; Nonell,
́
S.; Lemp, E. Photochem. Photobiol. 2013, 89, 1427.
15) Pedersen, S. K.; Holmehave, J.; Blaikie, F. H.; Gollmer, A.;
Breitenbach, T.; Jensen, H. H.; Ogilby, P. R. J. Org. Chem. 2014, 79,
079.
16) Song, B.; Wang, G.; Tan, M.; Yuan, J. J. Am. Chem. Soc. 2006,
28, 13442.
17) Dai, Z.; Tian, L.; Xiao, Y.; Ye, Z.; Zhang, R.; Yuan, J. J. Mater.
Chem. B 2013, 1, 924.
18) Xu, K.; Wang, L.; Qiang, M.; Wang, L.; Li, P.; Tang, B. Chem.
Commun. 2011, 47, 7386.
19) Song, D.; Cho, S.; Han, Y.; You, Y.; Nam, W. Org. Lett. 2013,
5, 3582.
20) Oliveira, M. S.; Severino, D.; Prado, F. M.; Angeli, J. P. F.;
(
3
(
1
(
(46) Peng, Q.; Berg, K.; Moan, J.; Kongshaug, M.; Nesland, J. M.
Photochem. Photobiol. 1997, 65, 235.
(47) Gaullier, J.-M.; Gez
C.; Bazin, M.; Morliere, P.; Dubertret, L. Photochem. Photobiol. 1995,
62, 114.
̀ ̀
e, M.; Santus, R.; Melo, T. S. E.; Maziere, J.-
̀
(
(48) Gollmer, A.; Besostri, F.; Breitenbach, T.; Ogilby, P. R. Free
Radical Res. 2013, 47, 718.
(
1
(
(49) Shieh, Y.-A.; Yang, S.-J.; Wei, M.-F.; Shieh, M.-J. ACS Nano
2010, 4, 1433.
(50) Yuan, Q.; Wu, Y.; Wang, J.; Lu, D.; Zhao, Z.; Liu, T.; Zhang, X.;
Tan, W. Angew. Chem., Int. Ed. 2013, 52, 13965.
(51) Yin, M.; Li, Z.; Liu, Z.; Ren, J.; Yang, X.; Qu, X. Chem. Commun.
2012, 48, 6556.
(52) da Silva, E. F. F.; Pedersen, B. W.; Breitenbach, T.; Toftegaard,
R.; Kuimova, M. K.; Arnaut, L. G.; Ogilby, P. R. J. Phys. Chem. B 2011,
116, 445.
(53) Kalyanasundaram, K. Inorg. Chem. 1984, 23, 2453.
(54) Bulina, M. E.; Lukyanov, K. A.; Britanova, O. V.; Onichtchouk,
D.; Lukyanov, S.; Chudakov, D. M. Nat. Protoc. 2006, 1, 947.
(55) Carpentier, P.; Violot, S.; Blanchoin, L.; Bourgeois, D. FEBS
Lett. 2009, 583, 2839.
(56) Pletnev, S.; Gurskaya, N. G.; Pletneva, N. V.; Lukyanov, K. A.;
Chudakov, D. M.; Martynov, V. I.; Popov, V. O.; Kovalchuk, M. V.;
Wlodawer, A.; Dauter, Z.; Pletnev, V. J. Biol. Chem. 2009, 284, 32028.
(57) Vegh, R. B.; Solntsev, K. M.; Kuimova, M. K.; Cho, S.; Liang, Y.;
Loo, B. L. W.; Tolbert, L. M.; Bommarius, A. S. Chem. Commun. 2011,
47, 4887.
Motta, F. D.; Baptista, M. S.; Medeiros, M. H. G.; Di Mascio, P.
Photochem. Photobiol. Sci. 2011, 10, 1546.
(
21) Gollmer, A.; Arnbjerg, J.; Blaikie, F.; Pedersen, B.; Breitenbach,
T.; Daasbjerg, K.; Glasius, M.; Ogilby, P. Photochem. Photobiol. 2011,
7, 671.
22) Kim, S.; Fujitsuka, M.; Majima, T. J. Phys. Chem. B 2013, 117,
3985.
23) Koide, Y.; Urano, Y.; Hanaoka, K.; Terai, T.; Nagano, T. ACS
Chem. Biol. 2011, 6, 600.
24) Koide, Y.; Urano, Y.; Hanaoka, K.; Piao, W.; Kusakabe, M.;
Saito, N.; Terai, T.; Okabe, T.; Nagano, T. J. Am. Chem. Soc. 2012,
34, 5029.
25) Egawa, T.; Hanaoka, K.; Koide, Y.; Ujita, S.; Takahashi, N.;
8
(
1
(
(
1
(
Ikegaya, Y.; Matsuki, N.; Terai, T.; Ueno, T.; Komatsu, T.; Nagano, T.
J. Am. Chem. Soc. 2011, 133, 14157.
(
26) Piao, W.; Tsuda, S.; Tanaka, Y.; Maeda, S.; Liu, F.; Takahashi,
S.; Kushida, Y.; Komatsu, T.; Ueno, T.; Terai, T.; Nakazawa, T.;
H
dx.doi.org/10.1021/ja504279r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(
58) Finikova, O. S.; Lebedev, A. Y.; Aprelev, A.; Troxler, T.; Gao, F.;
Garnacho, C.; Muro, S.; Hochstrasser, R. M.; Vinogradov, S. A.
ChemPhysChem 2008, 9, 1673.
(
59) Serebrovskaya, E. O.; Edelweiss, E. F.; Stremovskiy, O. A.;
Lukyanov, K. A.; Chudakov, D. M.; Deyev, S. M. Proc. Natl. Acad. Sci.
U.S.A. 2009, 106, 9221.
(
60) Kuimova, M.; Botchway, S.; Parker, A.; Balaz, M.; Collins, H.;
Anderson, H.; Suhling, K.; Ogilby, P. Nat. Chem. 2009, 1, 69.
61) Roy, R.; Hohng, S.; Ha, T. Nat. Methods 2008, 5, 507.
(
I
dx.doi.org/10.1021/ja504279r | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX