Highly activatable and environment-insensitive optical highlighters for selective spatiotemporal imaging of target proteins

Article abstract of DOI:10.1021/ja212125w

Optical highlighters are photoactivatable fluorescent molecules that exhibit pronounced changes in their spectral properties in response to irradiation with light of a specific wavelength and intensity. Here, we present a novel design strategy for a new class of caged BODIPY (4,4-difluoro-4-bora-3a, 4a-diaza-s-indacene) fluorophores, based on the use of photoremovable protecting groups (PRPGs) with high reduction potentials that serve as both a photosensitive unit and a fluorescence quencher via photoinduced electron transfer (PeT). 2,6-Dinitrobenzyl (DNB)-caged BODIPY was efficiently photoactivated, with activation ratios exceeding 600-fold in aqueous solutions. We then combined this photoactivatable fluorophore with a SNAP (mutant of O ⁶-alkylguanine DNA alkyltransferase) ligand to obtain a small-molecule-based optical highlighter for visualization of protein dynamics, using the well-established SNAP tag technology. As proof of concept, we demonstrate spatiotemporal imaging of the fusion protein of epidermal growth factor receptor (EGFR) with SNAP tag in living cells. We also demonstrate highlighting of cells of interest in live zebrafish embryos, using the fusion protein of histone 2A with SNAP tag.

Full text of DOI:10.1021/ja212125w

Article
pubs.acs.org/JACS
Highly Activatable and Environment-Insensitive Optical Highlighters
for Selective Spatiotemporal Imaging of Target Proteins
Tomonori Kobayashi,†,∥ Toru Komatsu,† Mako Kamiya,‡ Claudia Campos,§,⊥ Marcos Gonzalez-Gaitan,§
́
́
́
Takuya Terai,† Kenjiro Hanaoka,† Tetsuo Nagano,† and Yasuteru Urano*,†,‡,∇
†Graduate School of Pharmaceutical Sciences and ‡Graduate School of Medicine, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku,
Tokyo 113-0033, Japan
§Departments of Biochemistry and Molecular Biology, Faculty of Sciences, Geneva University, 30 Quai Ernest-Ansermet, 1211
Geneva, Switzerland
S
* Supporting Information
ABSTRACT: Optical highlighters are photoactivatable fluo-
rescent molecules that exhibit pronounced changes in their
spectral properties in response to irradiation with light of a
specific wavelength and intensity. Here, we present a novel
design strategy for a new class of caged BODIPY (4,4-difluoro-
4-bora-3a,4a-diaza-s-indacene) fluorophores, based on the use
of photoremovable protecting groups (PRPGs) with high
reduction potentials that serve as both a photosensitive unit
and a fluorescence quencher via photoinduced electron transfer (PeT). 2,6-Dinitrobenzyl (DNB)-caged BODIPY was efficiently
photoactivated, with activation ratios exceeding 600-fold in aqueous solutions. We then combined this photoactivatable
fluorophore with a SNAP (mutant of O6-alkylguanine DNA alkyltransferase) ligand to obtain a small-molecule-based optical
highlighter for visualization of protein dynamics, using the well-established SNAP tag technology. As proof of concept, we
demonstrate spatiotemporal imaging of the fusion protein of epidermal growth factor receptor (EGFR) with SNAP tag in living
cells. We also demonstrate highlighting of cells of interest in live zebrafish embryos, using the fusion protein of histone 2A with
SNAP tag.
activation with ultraviolet (UV) light removes the PRPG
(uncaging) and abruptly switches on the fluorescence of the
parent fluorophore. The requirements for an effective caged
fluorophore are rapid and efficient photoactivation, good
brightness, and photostability of the uncaged fluorophore. We
have already developed a rapidly releasable and highly activatable
caged fluorophore based on a novel fluorescein derivative (caged
TokyoGreen).7 However, the xanthene fluorophore used in
TokyoGreen is pH-sensitive, like fluorescent proteins. BODIPY
(4,4-difluoro-4-bora-3a,4a-diaza-s-indacene)8,9 is a widely used
fluorescent dye, which has many features that make it suitable for
bioimaging, including high fluorescence quantum efficiency
(Φfl), photostability, excitation at the visible light wavelength
(>500 nm), high molecular extinction coefficient, and pH or
environmental insensitivity of fluorescence. However, caged
BODIPY has never been developed because BODIPY has no
functional group for direct fluorescence regulation, unlike other
fluorophores. However, we developed a novel molecular design
based on intramolecular photoinduced electron transfer
(PeT).10 We have shown that fluorophores bearing electron-
deficient moieties are almost nonfluorescent because singlet
excited state energy is lost as a consequence of PeT.11 ortho-
Nitrobenzyl groups are the most representative PRPGs, and are
INTRODUCTION
■
Optical highlighters1 have been developed for visualizing,
tracking, and quantifying molecular events in living cells with
high spatial and temporal resolution. They are generally
photoactivatable fluorescent proteins (PAFPs),2 which exhibit
pronounced changes in their spectral properties in response to
irradiation with light of a specific wavelength and intensity.
Although these genetically encoded tags are undoubtedly
powerful general tools for imaging proteins within living
systems,3,4 they are not without limitations. For example, these
relatively large molecules can significantly perturb the structure
of proteins to which they are attached. Further, most of them lose
their fluorescence under acidic conditions, so that they are not
applicable to imaging of endocytotic processes, for example. In
recent years, an alternative strategy for specifically tagging
proteins has emerged that blends the simplicity of genetically
encoded tags with the versatility of small-molecular probes.5 This
approach involves the incorporation of a unique chemical
functionality or mutated enzyme into the target protein. These
reporters are small, noninvasive, and nonperturbing handles that
can be modified in living systems through highly selective
reactions with exogenously delivered probes. In this strategy,
photoactivatable (caged) fluorophores6 are required. In general,
caged fluorophores are weakly fluorescent or nonfluorescent
when essential functional groups for fluorescence emission are
masked by photoremovable protecting groups (PRPGs). Photo-
Received: December 28, 2011
Published: June 13, 2012
© 2012 American Chemical Society
11153
dx.doi.org/10.1021/ja212125w | J. Am. Chem. Soc. 2012, 134, 11153−11160

Journal of the American Chemical Society
Article
Figure 1. Spectral properties of caged BODIPY. (a) Molecular design of ortho-nitrobenzyl caged BODIPY. (b) Absorption (left) and fluorescence
(right) spectral change of DNB-BDP (6, 1 μM) in chloroform upon UV irradiation at around 350 nm (1.25 mW/cm2 at 350 nm). Absorption spectra
were recorded before (red line) and after irradiation (60 min). Fluorescence spectra were recorded after the specified duration of UV irradiation; 0 (red
line), 1, 2, 5, 10, 20, 30, 40, 50, 60 min (black line). (c) Fluorescence ratio values at the emission maximum wavelength upon UV irradiation in
chloroform. Fluorescence ratio denotes the ratio of the fluorescence intensity after 120 s UV irradiation with respect to the fluorescence intensity before
irradiation at the emission maximum wavelength. (d) Photoactivation diagram of water-soluble DNB-caged BODIPY (DNB-BDP-PH, 7). (e)
Fluorescence spectral change of DNB-BDP-PH (1 μM) in sodium phosphate buffer (pH 7.4) upon UV irradiation at around 350 nm (1.25 mW/cm2 at
350 nm). Fluorescence spectra were recorded after the specified duration of UV irradiation; 0 (red line), 10, 20, 30, 40, 50, 60, 80, 100, 120, and 150 min
(black line).
also strongly electron-deficient.12 So, we expected that
conjugation of ortho-nitrobenzyl groups to a BODIPY
fluorophore would result in fluorescence quenching via a PeT
process, while photodeprotection would result in loss of PeT,
with recovery of the fluorescence. Here, we report rational design
of caged BODIPY and its application to optical highlighting of
epidermal growth factor receptor (EGFR) in living cells and cells
of interest in live zebrafish embryo by utilizing the well-
established SNAP (a mutant of O6-alkylguanine-DNA alkyl-
transferase) tag technology.
1,3,5,7-tetramethyl BODIPY skeleton as a platform for the
following two reasons: (1) it has a high value of Φfl due to its
sterically constrained structure,13 and (2) many varieties of
functional groups can be introduced into the 8-phenyl group for
fluorescence regulation or bioconjugation. The structure can be
considered as consisting of two orthogonal parts, the benzene
moiety and the fluorophore, since they are twisted and
conjugatively uncoupled. As PRPGs, we chose four ortho-
nitrobenzyl derivatives with different electrochemical and
photochemical properties, that is, 4,5-dimethoxy-2-nitrobenzyl
(DMNB), 2-nitrobenzyl (NB), 1-(2-nitrophenyl)ethyl (NPE),
and 2,6-dinitrobenzyl (DNB) groups. A benzyl (Bn)-protected
derivative was designed as a control compound. ortho-Nitro-
benzyl groups can protect a variety of functional groups. Among
them, we chose a phenolic group as a caging site, because the
RESULTS
■
Molecular Design of Caged BODIPY. The structures of
caged BODIPY are shown in Figure 1a. We chose the 8-phenyl-
11154
dx.doi.org/10.1021/ja212125w | J. Am. Chem. Soc. 2012, 134, 11153−11160

Journal of the American Chemical Society
Article
Figure 2. Photoactivation experiments with the DNB-BDP-PH-SNAP complex in vitro. (a) Covalent labeling of a SNAP-fusion protein using O6-
benzylguanine (BG) derivatives. (b) Structures of caged BODIPY-SNAP ligands. (c) Fluorescence image of an SDS−PAGE gel after irradiation. SNAP
(0.5 μM) was incubated for 2.5 h in the presence of BG-DNB-BDP-PH (9, 10 μM). After completion of labeling, the solutions were irradiated with UV
light (365 nm) for various periods of time (0, 1, 2, 4, 8, 12, 16 min). The irradiated solution was denatured by heating, and then subjected to SDS-PAGE.
(d) Fluorescence intensity of the SNAP band.
ether bond is stable under physiological conditions. Another
important point is the substitution position in the benzene
moiety. Generally, PeT is less effective with increasing distance
between the fluorophore and the quencher.14 Therefore, it
should be preferable to conjugate PRPGs at the nearest position
to the fluorophore. It should also be noted that there is only a
single ortho-nitrobenzyl group, and so the caged BODIPY is
expected to be rapidly activated, compared to traditional bis-
caged fluorescein or rhodamine.7 Also, the photogenerated
byproduct (ortho-nitrosobenzaldehyde) is expected to show low
cytotoxicity.
Synthesis and Spectroscopic Properties of Caged
BODIPY and Relationship between Fluorescence Quan-
tum Yield and Reduction Potential of the Electron
Acceptor. As shown in Supporting Information Figure S1, 2-
hydroxybenzaldehyde, used as a starting material, was protected
with a benzyl group, followed by condensation with 2,4-
dimethylpyrrole in the presence of trifluoroacetic acid, oxidation
with 2,3-dichloro-5,6-dicyano-p-benzoquinone, and coordina-
tion of boron trifluoride diethyl etherate to give Bn-BDP (2).
The benzyl group was removed by catalytic reduction to afford
the photoproduct, H-BDP (1). Three caged derivatives, DMNB-
BDP (3), NB-BDP (4), and NPE-BDP (5), were synthesized
from the corresponding benzaldehydes in the same manner as
described for Bn-BDP. With regard to the dinitrobenzyl
derivative (DNB-BDP, 6), the corresponding benzaldehyde
could not be obtained because of the formation of byproduct.
Therefore, we used the corresponding acetal to obtain DNB-
BDP. The Φfl values of caged BODIPY in various organic
solvents are summarized in Supporting Information Table S1.
Their absorption spectra were almost the same in every solvent,
except that DMNB-BDP had higher absorbance in the 350−400
nm region (Supporting Information Figure S2). The photo-
product H-BDP was highly fluorescent in every solvent. Bn-BDP,
the control compound with no nitro group, was as highly
fluorescent as H-BDP. In contrast, all nitrobenzyl-caged
BODIPYs were less fluorescent than Bn-BDP, although the Φfl
values differed to some extent from compound to compound.
The value of Φfl was the smallest for DNB-BDP, followed by
NPE-BDP, NB-BDP, and DMNB-BDP. The corresponding
benzene moiety of each caged BODIPY was independently
synthesized and their reduction potentials were examined by
means of cyclic voltammetry. Their reduction potentials in
acetonitrile were in the order DNB > NPE > NB > DMNB > Bn,
11155
dx.doi.org/10.1021/ja212125w | J. Am. Chem. Soc. 2012, 134, 11153−11160

Journal of the American Chemical Society
Article
Figure 3. Optical highlighting of EGFR-SNAP fusion protein. (a) Confocal fluorescence image of BG-Bn-BDP-PH (10, 2 μM)-loaded HeLa cells
transiently expressing EGFR-SNAP protein before washing (left panel) and after washing (right panel). (b) Confocal fluorescence image of BG-DNB-
BDP-PH (9, 2 μM)-loaded HeLa cells transiently expressing EGFR-SNAP protein before and after UV irradiation. UV irradiation was carried out at the
whole field for 5 min by a handy-held UV lamp (365 nm, 1.5 mW/cm2). After EGF addition to the irradiated cell culture (final concentration: 100 ng/
mL) and incubation at 37 °C for 3 h, a fluorescence image was captured. Scale bar: 15 μm. (c) Epifluorescence images of BG-DNB-BDP-PH (2 μM)-
loaded COS-7 cells transiently expressing EGFR-SNAP protein before and after UV irradiation. Localized UV irradiation was carried out for 30 s
through an objective lens with a pinhole. EGFR-SNAP expressing cells were marked with a cell-impermeable fluorescent marker, BGAlexa-546 (2 μM),
after labeling with BG-DNB-BDP-PH (2 μM).
in agreement with the values of Φfl. To provide further evidence
for fluorescence regulation based on PeT, we prepared methoxy-
and fluorine-substituted caged BODIPYs. It was reported that
the oxidation potential of methoxy-substituted BODIPY was
greatly decreased compared to those of fluorine-substituted
derivatives without any change in the spectral properties.15 For
almost all the caged derivatives, the Φfl was decreased for
methoxy derivatives compared to fluorine derivatives (Support-
ing Information Table S1).
Photoactivation Experiments in Cuvette. Solutions of
caged BODIPY were illuminated with UV light from a 500 W
xenon lamp equipped with a monochromator system, with
stirring. Experiments were initially carried out in five organic
solvents (acetonitrile, methanol, acetone, dichloromethane, and
chloroform). Representative changes of the absorbance and
fluorescence spectra of DNB-BDP upon UV irradiation in
chloroform are shown in Figure 1b. H-BDP was generated, with a
concomitant fluorescence increase, upon irradiation of DNB-
BDP, as confirmed by comparison of the 1H NMR spectra with
that of an authentic sample (Supporting Information Figure S3).
Fluorescence ratio values upon illumination in chloroform are
shown in Figure 1c. Among the compounds, DNB-BDP showed
the largest fluorescence ratio value. Similar results were obtained
in other solvents (Supporting Information Figure S4). Next, we
synthesized a water-soluble DNB-caged BODIPY in order to
examine photoactivatability in aqueous solution. To improve the
water solubility, we introduced a carboxylic acid into the benzene
moiety and two propionates into the fluorophore (DNB-BDP-
PH, 7, Figure 1d). The fluorescence of DNB-BDP-PH was
strongly quenched (Φfl = 0.001) in pH 7.4 buffered aqueous
solution, and a dramatic fluorescence increase (>600-fold) was
observed upon generation of the photoproduct (H-BDP-PH, 8,
Φfl = 0.658) by means of UV irradiation (Figure 1e). The
uncaging cross section (product of molar extinction coefficient
and uncaging quantum yield) of DNB-caged BODIPY was 50
M
−1 cm−1 (ε350 nm = 5.6 × 103 M−1 cm−1, Φuncage = 8.9 × 10−3),
which is comparable to those of other ortho-nitrobenzyl caged
compounds (Supporting Information Figure S5).16 Rhodamine
11156
dx.doi.org/10.1021/ja212125w | J. Am. Chem. Soc. 2012, 134, 11153−11160

Journal of the American Chemical Society
Article
Figure 4. Optical highlighting of cells of interest in a zebrafish embryo. (a and b) Confocal fluorescence images of embryonic spinal cord. The embryos
were injected with H2A-SNAP and double-stained with BG-DNB-BDP-PH and BG-Cy5. Before irradiation, one (a) or two (b) nuclei were selected
based on Cy5 fluorescence (regions surround by red line), and were clearly highlighted upon irradiation with a 405 nm laser.
is another candidate for a photobleaching-resistant fluorophore.
We have applied this strategy to rhodamine, and synthesized
NPE-caged and DNB-caged tetramethylrhodamine (TMR)
(Supporting Information Figure S6a). We found that the
fluorescence of these compounds was efficiently quenched in
aqueous solution, but serious photobleaching occurred with UV
illumination, as can be seen in the absorption spectra
(Supporting Information Figure S6b and S6c). The photo-
product 2-OH TMR did not show much photobleaching
(Supporting Information Figure S6d), so an unexpected reaction
involving the ortho-nitrobenzyl moiety might have occurred in
these cases.
Environment-Insensitivity of Caged BODIPY. We
examined the environment-insensitivity of the fluorescence of
DNB-caged BODIPY. Fluorescence of DNB-caged BODIPY was
efficiently quenched (Φfl < 0.01) in various solvents from polar to
nonpolar solvents (Supporting Information Table S1). Although
another PeT-regulated dye, DNB-caged TokyoGreen, has low
initial fluorescence under aqueous conditions, it emits moderate
fluorescence without photoactivation in less polar environments,
such as methanol or at a protein surface (Supporting Information
Figures S7 and S8). DNB-caged BODIPY was stable at
physiological pH and in the presence of some reducing enzymes
(Supporting Information Figures S9 and S10). Additionally, we
synthesized the photoproduct and examined its pH sensitivity. It
emitted stable fluorescence in acidic solutions from pH 2 to 7
(Supporting Information Figure S11).
Synthesis of Caged BODIPY-SNAP Ligand and Its
Reaction Rate with SNAP and Photoactivatability. To
examine the feasibility of using our fluorophores for small-
molecule-based optical highlighting, we used the SNAP tag
technology to specifically label target proteins.17 SNAP protein is
derived from O6-alkylguanine-DNA alkyltransferase (AGT),
which reacts covalently with benzylguanine (BG). This tag can
be specifically and covalently labeled with any fluorophore
carried by the benzyl group of BG (Figure 2a). Thus, we designed
and synthesized caged BODIPY-SNAP ligand (BG-DNB-BDP-
PH, 9), a fluorescent analogue (BG-Bn-BDP-PH, 10), and the
photoproduct (BG-H-BDP-PH, 11) (Figure 2b). They were
designed to be hydrophilic and membrane-impermeable so that
they could be easily washed out. BG was appended onto the
benzene moiety via an amide bond, and two propionates were
introduced into the fluorophore to increase the water solubility.
The reaction kinetics of these substrates with purified SNAP-
GST in vitro were comparable with that of a standard fluorogenic
substrate, BGFL (Supporting Information Figure S12). Non-
specific binding was not observed when BG-Bn-BDP-PH was
incubated with SNAP-GST in the presence of bovine serum
albumin (Supporting Information Figure S13). Next, we
examined the photoactivatability of DNB-BDP-PH-SNAP
complex in vitro. After completion of labeling in aqueous
solution buffered at pH 7.4, DNB-BDP-PH-SNAP complex was
UV-irradiated for various periods of time. The fluorescence
intensity of the SNAP band increased in an irradiation time-
dependent manner (Figure 2c,d).
Optical Highlighting of EGFR-SNAP Fusion Protein
with Caged BODIPY-SNAP Ligand. We selected epidermal
growth factor receptor (EGFR) as a target to test the validity of
our approach, since sufficient structural information is available
to design a fusion protein.18,19 We prepared a gene construct for a
fusion protein of EGFR with SNAP. After transient transfection
using standard liposome-mediated gene transfer techniques,
HeLa cells were incubated with BG-Bn-BDP-PH, a fluorescent
analogue, in order to identify optimal labeling conditions under a
fluorescence microscope. BG-Bn-BDP-PH was membrane-
impermeable, and nonspecific binding or accumulation of free
dye was not observed (Figure 3a). Then, we attempted optical
highlighting of EGFR-SNAP with BG-DNB-BDP-PH. After
labeling and washout of the unreacted probes, cells in the
whole field were irradiated with a handy UV lamp (365 nm).
Fluorescence increase was observed in transfected cells, and
following EGF stimulation resulted in internalization of EGFR
(Figure 3b). Further, we demonstrated localized photoactivation
experiments using pulse-chase labeling techniques (Figure 3c).
First, COS-7 cells transiently expressing EGFR-SNAP protein
were incubated with BG-DNB-BDP-PH and unreacted dye was
washed out. Then, the cells were pulse-labeled with a red-
fluorescent substrate, BGAlexa-546 (top panel, right). UV light
was focused to the small area of the cells showing red
fluorescence and irradiated through an objective lens under an
11157
dx.doi.org/10.1021/ja212125w | J. Am. Chem. Soc. 2012, 134, 11153−11160

Journal of the American Chemical Society
Article
epifluorescence microscope. After irradiation, we observed an
increase of fluorescence intensity at the irradiated site (bottom
panel), with subsequent fluorescence decay due to lateral
diffusion and vesicular trafficking of EGFR-SNAP ( Supporting
Information Figure S14 and Movie 1).20,21
PRPGs (Supporting Information Table S1). The above results
suggest that our PeT-regulated photoswitching design strategy
has worked efficiently.
Photoactivation experiments in cuvette showed that DNB-
caged BODIPY had the best properties as a caged fluorophore
(Figure 1b−e). DNB-BDP had the highest fluorescence
activation ratio among the four derivatives. NPE-BDP gave a
comparable fluorescence increase, but the fluorescence ratio
value was smaller because of its moderate initial fluorescence
(Supporting Information Figure S4). It was essential to use the
2,6-dinitrobenzyl group in our case from the viewpoint of
obtaining efficient PeT to minimize the initial fluorescence. In
practical applications, it is very important that the initial
fluorescence before photoactivation should be as low as possible
in order to obtain high-contrast fluorescence images. Here, we
should emphasize the point that in the design of caged
fluorophores based on the PeT mechanism, not only the
photochemical properties (such as Φuncage, molar extinction
coefficient, and reaction rate), but also the electrochemical
properties of PRPGs are critically important to achieve high
photoactivation fluorescence efficiency; this is different from the
case of usual caged bioactive compounds.
The DNB-caged BODIPY described here has several
advantages over traditional caged fluorophores. First, the initial
fluorescence of DNB-caged BODIPY is very low in solvents with
different polarity from water to chloroform, unlike that of
another PeT-regulated caged fluorophore, caged TokyoGreen
(Supporting Information Figures S7 and S8).7 Second, uncaged
BODIPY emits stable fluorescence insensitive to changes of
solvent polarity (Supporting Information Table S1) or pH
(Supporting Information Figure S11). In contrast, fluorescein24
or coumarin25 dyes have acidic groups in the fluorophore, and
their emissions are influenced by pH change around the
physiological range. Third, it can be activated by single-step
deprotection of PRPG, and therefore, rapid uncaging is possible
with minimal UV irradiation, whereas bis-caged fluorescein or
rhodamine require more prolonged illumination.
We also checked the pH-insensitivity of BODIPY dyes in living
cells. HeLa cells expressing EGFR-SNAP were labeled with BG-
H-BDP-PH, the photoproduct of our caged substrates, and with
BGFL, a BG derivative of caged fluorescein. The fluorescence of
BGFL is known to be quenched under slightly acidic conditions
owing to the formation of a protonated form, that is, it is pH-
sensitive. Before EGF stimulation, EGFR in plasma membrane
was labeled in the same way with each of the probes (Supporting
Information Figure S15a). After EGF stimulation, a part of EGFR
was internalized through endocytosis and could be observed as
fluorescent vesicles. Next, we added ammonium chloride to
neutralize these acidic vesicles (Supporting Information Figure
S15b). The fluorescence intensity did not change significantly in
the BODIPY-labeled cells, whereas that of the BGFL-labeled
cells increased by ∼40%. After washout of ammonium chloride,
the fluorescence intensity reverted to the original value.
Optical Highlighting of Target Nuclei in Zebrafish
Embryo with Caged BODIPY-SNAP Ligand. As another
application of our probe, we tried in vivo optical highlighting of
target nuclei in zebrafish embryo. We have previously
demonstrated that the combination of SNAP-tag with photo-
activatable substrates can be used for this purpose.22 To
investigate whether our newly developed BG-DNB-BDP-PH
can also be applied for highlighting cells of interest in living
zebrafish embryo, we injected embryos at the one-cell stage with
H2A-SNAP RNA and BG-DNB-BDP-PH/BG-Cy5. The em-
bryos were allowed to develop for 24 h, then a selected single
nucleus (Figure 4a, Supporting Information Movie 2) or two
nuclei (Figure 4b, Supporting Information Movie 3) were
illuminated with 405 nm laser light. We observed a marked
fluorescence increase only in the illuminated nuclei, confirming
that our probe can be used for highlighting cells of interest in the
developing embryo.
As a practical application to demonstrate the utility of our
approach, we investigated optical highlighting of EGFR. Small-
molecule-based optical highlighting has faced the critical
problem that selective labeling techniques for target molecules
in situ were not available, and consequently, nonspecific labeling
and dye accumulation within cells generated nontarget-derived
signals, so that quantitative analysis was impossible. Recently,
however, some promising techniques have emerged such as
SNAP tag,17 Halotag,26 and tetracysteine tag27 for protein
labeling in situ. Among them, SNAP tag technology is
commercially available and has been used for many biological
studies (Figure 2a). So, we designed a caged BODIPY-SNAP
ligand and its derivatives with suitable labeling kinetics for
biological application (Figure 2b). Caged BODIDPY-SNAP
complex worked efficiently in vitro (Figure 2c,d), and so we next
tried optical highlighting in living cells. After UV irradiation of
the whole field (Figure 3b) or a localized site (Figure 3c),
fluorescence increase was observed in transiently EGFR-SNAP-
expressing cells. Further, pH-insensitivity of BODIPY probes was
confirmed by the lack of fluorescence change in response to
neutralization of acidic vesicles containing the fusion protein
(Supporting Information Figure S15). These experiments
successfully demonstrated the optical highlighting of a target
protein in living cells. As another application of our probes, we
demonstrated optical highlighting of zebrafish nuclei (Figure
4a,b). After injection of BG-DNB-BDP-PH and 405 nm laser
DISCUSSION
■
Here, we have proposed a novel concept for molecular design of
caged fluorophores, as exemplified by caged BODIPYs. The basis
of this approach is that PRPGs with high reduction potentials can
be used as both a fluorescence quencher via a PeT process and a
photosensitive unit (Figure 1a). In general, the feasibility of
electron transfer between a fluorophore and a quencher can be
judged from the change in Gibbs free energy (ΔGeT). The ΔGeT
value can be calculated from measurable parameters by use of the
Rehm−Weller equation,23 as follows:
ΔGeT = Eox − Ered − E00 − C
where Eox and Ered are the oxidation potential of the electron
donor and the reduction potential of the electron acceptor,
respectively, E00 is the singlet excited energy of the fluorophore,
and C is the work term for the charge separation state. In the case
of caged BODIPY, all compounds have the same absorption
maxima, so that the E00 values can be determined easily. The
values of C may also be similar in view of the structural similarity
of the compounds. As a result, the efficiency of PeT can be
compared just in terms of Eox and Ered of the fluorophore and
benzene moieties, respectively. The Φfl of caged BODIPY
correlated well with Ered of the corresponding benzene moieties
in the same fluorophore or Eox of the fluorophores in the same
11158
dx.doi.org/10.1021/ja212125w | J. Am. Chem. Soc. 2012, 134, 11153−11160

Journal of the American Chemical Society
Article
illumination, a marked fluorescence increase was observed only
in the illuminated nuclei. This result suggests that caged
BODIPY could be a useful tool for cell lineage analysis in
developmental biology.
ACKNOWLEDGMENTS
■
We thank Dr. Haruhiko Bito, Dr. Hiroyuki Okuno, and Yayoi
Kondo in the University of Tokyo for their technical support in
DNA and cellular experiments. This work was financially
supported by the Ministry of Education, Culture, Sports, Science
and Technology of Japan (Grants for The Advanced and
Innovational Research Program in Life Sciences, 16370071 and
16659003, Grants 20117003 and 23249004 to Y.U., and
23113504 to M.K.). T.N. was also supported by the Hoh-
ansha Foundation. C.C. was supported by FCT (SFRH/BD/
15210/2004), Oncasym and Programa Gulbenkian de Doutor-
amento em Biomedicina. M.G.-G. and C.C. were supported by
the Swiss National Science Foundation, grants from the Swiss
SystemsX.ch initiative, and LipidX-2008/011, an ERC advanced
investigator grant, NCCR Chemical Biology and the Polish-
Swiss research program.
In this case, the SNAP tag is nearly as large as fluorescent
proteins. However, there are smaller and less perturbing tag
systems, such as tetracysteine27 or BODIPY/diacrylate tags,28
and their use should make it possible to optically highlight a
target protein with relatively little influence on its function.
Moreover, small-molecule-based optical highlighting techniques
should be applicable to biomolecules other than proteins, such as
glycans or lipids. Recently, bioorthogonal chemical reporter
selective labeling technologies,5 such as copper-free click
chemistry29 or Staudinger ligation,30 have been developed and
applied to many biological studies. It should be possible to
introduce these reactive moieties into caged BODIPYs using
simple methods of organic synthesis. Another important
application of caged fluorophores is superresolution microscopy,
such as PALM (photoactivation localized microscopy).31
Recently, Banala reported the application of a caged rhod-
amine110 derivative/SNAP tag for superresolution microscopy,
obtaining a high photon yield compared to that in the case of
fluorescent proteins.32 Likewise, our caged BODIPY/SNAP tag
is also expected to contribute to the color pallet of small-
molecule-based optical highlighters.
REFERENCES
■
(1) Shaner, N. C.; Patterson, G. H.; Davidson, M. W. J. Cell. Sci. 2007,
120, 4247−4260.
(2) Berkhusha, V. V.; Lukyanov, K. A. Nat. Biotechnol. 2004, 22, 289−
296.
(3) Patterson, G. H.; Lippincott-Schwartz, J. Science 2002, 297, 1873−
1877.
(4) Habuchi, S.; Ando, R.; Dedecker, P.; Verheijen, W.; Mizuno, H.;
Miyawaki, A.; Hofkens, J. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 9511−
9516.
(5) Prescher, J. A.; Bertozzi, C. R. Nat. Chem. Biol. 2005, 1, 13−21.
(6) Puliti, D.; Warther, D.; Orange, C.; Specht, A.; Goeldner, M. Bioorg.
Med. Chem 2011, 19, 1023−1029.
(7) Kobayashi, T.; Urano, Y.; Kamiya, M.; Ueno, T.; Kojima, H.;
Nagano, T. J. Am. Chem. Soc. 2007, 129, 6696−6697.
(8) Ulrich, G.; Ziessel, R.; Harriman, A. Angew. Chem., Int. Ed. 2008, 47,
1184−1201.
CONCLUSION
■
We have developed small-molecule-based optical highlighters
with high activation ratio and environmental insensitivity
(polarity or pH), by combining a new class of caged BODIPY
fluorophores with SNAP tag technology. We confirmed the
usefulness of this approach for two applications: optical
highlighting of (1) EGFR in cells expressing EGFR-SNAP
fusion protein, and (2) selected cells in zebrafish embryo
expressing H2A-SNAP fusion protein. We are currently working
to extend this approach for optical highlighting of small
biomolecules other than proteins.
(9) Loudet, A.; Burgess, K. Chem. Rev. 2007, 107, 4891−4932.
(10) De Silva, A. P.; Gunaratne, H. Q. N.; Gunnlaugsson, T.; Huxley,
A. J. M.; McCoy, C. P.; Rademacher, J. T.; Rice, T. E. Chem. Rev. 1997,
97, 1515−1566.
(11) Ueno, T.; Urano, Y.; Setsukinai, K.; Takakusa, H.; Kojima, H.;
Kikuchi, K.; Ohkubo, K.; Fukuzumi, S.; Nagano, T. J. Am. Chem. Soc.
2004, 126, 14079−14085.
ASSOCIATED CONTENT
(12) Pillai, V. N. R. Synthesis 1980, 1, 1−26.
■
(13) Cui, A.; Peng, X.; Fan, J.; Chen, X.; Wu, Y.; Guo, B. J. Photochem.
Photobiol., A 2007, 186, 85−92.
S
* Supporting Information
Experimental conditions; supplementary methods for chemical
synthesis and characterization of compounds; supplementary
experiments and data; Figures S1−S15; Table S1; supplementary
movies 1−3; complete ref 26. This material is available free of
charge via the Internet at http://pubs.acs.org.
(14) Matsumoto, T.; Urano, Y.; Shoda, T.; Kojima, H.; Nagano, T. Org.
Lett. 2007, 9, 3375−3377.
(15) Gabe, Y.; Ueno, T.; Urano, Y.; Kojima, H.; Nagano, T. Anal.
Bioanal. Chem. 2006, 386, 621−626.
(16) Aujard, I.; Benbrahim, C.; Gouget, M.; Ruel, O.; Baudin, J.;
Neveu, P.; Jullien, L. Chem.Eur. J. 2006, 12, 6865−6879.
(17) Keppler, A.; Gendreizig, S.; Gronemeyer, T.; Pick, H.; Vogel, H.;
Johnsson, K. Nat. Biotechnol. 2003, 21, 86−89.
(18) Carter, R. E.; Sorkin, A. J. Biol. Chem. 1998, 273, 35000−35007.
(19) Brock, R.; Hamelers, I. H.; Jovin, T. M. Cytometry 1999, 35, 353−
362.
AUTHOR INFORMATION
■
Corresponding Author
*To whom correspondence should be addressed. E-mail:
uranokun@m.u-tokyo.ac.jp
(20) Lajoie, P.; Partridge, E. A.; Guay, G.; Goetz, J. G.; Pawling, J.;
Lagana, A.; Joshi, B.; Dennis, J. W.; Nabi, I. R. J. Cell Biol. 2007, 179,
341−356.
Present Addresses
∥Chemistry Research Laboratories, Dainippon Sumitomo
Pharma Co. Ltd., 33-94 Enoki-cho, Suita Osaka 564-0053, Japan
(21) Sorkin, A.; von Zastrow, M. Nat. Rev. Mol. Cell Biol. 2002, 3, 600−
614.
⊥Instituto Gulbenkian de Cien
̂
cia, R. da Quinta Grande n°6,
(22) Campos, C.; Kamiya, M.; Banala, S.; Johnsson, K.; Gonzal
́
ez-
2780-156 Oeiras, Portugal
Gaitan, M. Dev. Dyn. 2011, 240, 820−827.
́
∇Basic Research Program, Japan Science and Technology
Agency, 7 Gobancho, Chiyoda-ku, Tokyo 102-0076, Japan.
(23) Rehm, D.; Weller, A. Isr. J. Chem. 1970, 8, 259−271.
(24) Zheng, G.; Guo, Y. M.; Li, W. J. Am. Chem. Soc. 2007, 129, 10616−
10617.
Notes
(25) Zhao, Y.; Zheng, Q.; Dakin, K.; Xu, K.; Martinez, M. L.; Li, W. J.
Am. Chem. Soc. 2004, 126, 4653−4663.
The authors declare no competing financial interest.
11159
dx.doi.org/10.1021/ja212125w | J. Am. Chem. Soc. 2012, 134, 11153−11160

Journal of the American Chemical Society
Article
(26) Los, G. V.; et al. ACS Chem. Biol. 2008, 3, 373−382.
(27) Griffin, B. A.; Adams, S. R.; Tsien, R. Y. Science 1998, 281, 269−
272.
(28) Lee, J.; Lee, S.; Zhai, D.; Ahn, Y.; Yeo, H.; Tan, Y.; Chang, Y.
Chem. Commun. 2011, 47, 4508−4510.
(29) Laughlin, S. T.; Baskin, J. M.; Amacher, S. L.; Bertozzi, C. R.
Science 2008, 320, 664−667.
(30) Hangauer, M. J.; Bertozzi, C. R. Angew. Chem., Int. Ed. 2008, 47,
2394−2397.
(31) Betzig, E.; Patterson, G H.; Sougrat, R.; Lindwasser, O. W.;
Olenych, S.; Bonifacino, J. S.; Davidson, M. W.; Lippincott-Schwartz, J.;
Hess, H. F. Science 2006, 313, 1642−1645.
(32) Banala, S.; Maurel, D.; Manley, S.; Johnsson, K. ACS Chem. Biol.
2012, 7, 289−293.
11160
dx.doi.org/10.1021/ja212125w | J. Am. Chem. Soc. 2012, 134, 11153−11160