Selective and ratiometric fluorescent trapping and quantification of protein vicinal dithiols and in situ dynamic tracing in living cells

Article abstract of DOI:10.1021/ja5079656

Protein vicinal dithiols play fundamental roles in intracellular redox homeostasis due to their involvement in protein synthesis and function through the reversible vicinal dithiol oxidation to disulfide. To provide quantitative information about the glob

Full text of DOI:10.1021/ja5079656

Article
pubs.acs.org/JACS
Selective and Ratiometric Fluorescent Trapping and Quantification
of Protein Vicinal Dithiols and in Situ Dynamic Tracing in Living Cells
Chusen Huang,†,‡ Ti Jia,‡ Mengfang Tang,† Qin Yin,† Weiping Zhu,*,† Chao Zhang,† Yi Yang,†
Nengqin Jia,*,‡ Yufang Xu,*,† and Xuhong Qian†
†State Key Laboratory of Bioreactor Engineering, Shanghai Key Laboratory of Chemical Biology, School of Pharmacy, East China
University of Science and Technology, 130 Meilong Road, Shanghai 200237, China
‡The Education Ministry Key Laboratory of Resource Chemistry and Shanghai Key Laboratory of Rare Earth Functional Materials,
Department of Chemistry, College of Life and Environmental Sciences, Shanghai Normal University, 100 Guilin Road, Shanghai
200234, China
S
* Supporting Information
ABSTRACT: Protein vicinal dithiols play fundamental roles in intracellular redox
homeostasis due to their involvement in protein synthesis and function through the
reversible vicinal dithiol oxidation to disulfide. To provide quantitative information
about the global distribution and dynamic changes of protein vicinal dithiols in living
cells, we have designed and synthesized a ratiometric fluorescent probe (VTAF) for
trapping of vicinal dithiol-containing proteins (VDPs) in living cells. VTAF exhibits a
ratiometric fluorescence signal upon single excitation, which enables self-calibration
of the fluorescence signal and quantification of endogenous vicinal dithiols of VDPs.
Its potential for in situ dynamic tracing of changes of protein vicinal dithiols under
different cellular redox conditions was exemplified. VTAF facilitated the direct
observation of subcellular distribution of endogenous VDPs via ratiometric
fluorescence imaging and colocalization assay. And the results suggested that there
are abundant VDPs in mitochondria. Moreover, some redox-sensitive VDPs are also
present on cell surface which can respond to redox stimulus. This ratiometric fluorescence technique presents an important
extension to previous fluorescence intensity-based probes for trapping and quantifying protein vicinal dithiols in living cells, as
well as its visible dynamic tracing of VDPs.
approach has identified some new vicinal dithiol-containing
proteins (VDPs) in cell lysate. Additionally, arsenite (III)-
affinity chromatography,19−21 biotinylated conjugates of
PAO,22,23 and dimaleimide fluorogens have also been
introduced to make the proteomic study of VDPs with
different possible functions (Supporting Information Scheme
S1). However, these methods focused on developing strategies
for unveiling VDPs in cell lysate and cannot collect more
information about the global distribution and dynamic
changes of endogenous VDPs in situ. Thus, in order to
investigate the essential roles of protein vicinal dithiols in
cellular redox homeostasis and protein function in living cells,
the vicinal dithols of VDPs should be specifically labeled and
trapped ideally in living cells prior to cell lysate24
Previously, we developed a highly selective fluorescent
probe (NPE) based on 2-p-aminophenyl-1,3,2-dithiarsenolane
(VTA2) and naphthalimide, which could enable an in situ
imaging of VDPs through the direct fluorescence readout.25
Most recently, a monoarsenical fluorescent probe
(TRAP_Cy3) based on cyanine chromophore has also been
exploited to identify that protein vicinal dithiols can decrease
INTRODUCTION
■
Redox-based post-translational modification of protein thiols
plays fundamental roles in cellular homeostasis and cell
signaling,1−4 and reversible protein vicinal dithiol oxidation to
disulfide holds a particularly prominent position due to its
involvements in protein synthesis and function, and
implications in cancer,5,6 diabetes,7 platelet function,8 human
immunodeficiency virus type 1 (HIV-1),9−11 neurodegenera-
tion,12 and so forth. Apart from its special role as the redox
center, the protein vicinal dithiols are also a key linking center
for formation and stabilization of protein structures during the
protein post-translational modification, which is considered to
be closely related to protein functions.13−16 Therefore,
trapping and identifying cellular protein vicinal dithiols levels
are valuable not only for further investigation of the protein
post-translational modifications, but also for the redox related
disease diagnosis and pathophysiology elucidation.17
Since trivalent arsenicals can selectively interact with
protein vicinal dithiols, p-aminophenylarsenoxide (PAO) and
thiol alkylation agents have been widely used to estimate
different concentrations of monothiol and total thiol groups
through selective labeling of vicinal dithiols with PAO, then
the protein vicinal dithiols can be quantified by another thiol
alkylation agent after removal of PAO.18 This common
Received: August 4, 2014
Published: September 15, 2014
© 2014 American Chemical Society
14237
dx.doi.org/10.1021/ja5079656 | J. Am. Chem. Soc. 2014, 136, 14237−14244

Journal of the American Chemical Society
Article
oxidative protein damage in adaptive microbial responses.26
TRAP_Cy3 and NPE can trace the redox response of protein
vicinal dithiols within living cells. However, a potential
drawback of these chemical probes is the background
fluorescence from unreacted or nonspecifically bound probes.
A long washing procedure (at least 15 min) is needed to
remove the unbound probes in order to reduce the
fluorescence background, which might make measurements
prone to artifacts. Moreover, although the endogenous VDPs
have been visualized by labeled fluorescent probes, the
fluorescence-intensity based signal may be obstructed by the
probe’s concentration, inhomogeneous cellular distribution,
variations in excitation intensity, sample thickness, and other
environment factors such as changes in pH and temperature,
thus diminishing its capacity to provide more accurate and
quantitative information about dynamic changes of vicinal
dithiols on VDPs in living cells. Comparatively, ratiometric
fluorescent probes are of considerable practical value for
reducing the influence of above factors since the detectable
signal can be reflected through the ratio of emission intensity
at two wavelengths.27−31
Herein, we report a ratiometric fluorescent probe (VTAF)
based on fluorescence resonance energy transfer (FRET)
mechanism for endogenous VDPs and exemplify its potentials
for imaging protein vicinal dithiols in in vitro studies. VTAF
exhibits a ratiometric fluorescence signal upon single
excitation, which enables self-calibration of the fluorescence
signal and quantification of endogenous vicinal dithiols of
VDPs within live cells. As a result, the VDPs are specifically
labeled by a coumarin fluorophore after their interaction with
VTAF. Because the emissions of VTAF and fluorescently
labeled VDPs are different, unreacted VTAF does not need to
be washed out (Scheme 1). To our knowledge, VTAF is the
first ratiometric fluorescent probe for trapping endogenous
VDPs in live cells.
RESULTS AND DISCUSSION
■
Design and Synthesis of VTAF. Our investigation begins
with design and synthesis of a selective and ratiometric VDPs-
responsive fluorescent probe based on FRET mechanism. We
have previously synthesized a highly selective receptor (2-p-
aminophenyl-1,3,2-dithiarsenolane (VTA2)) for VDPs based
on the specific interaction of trivalent arsenic with protein
vicinal dithiols, while monothiols of protein have much lower
affinity.32 VTA2 displayed high stability and biocompatibility
(higher cell permeability and lower toxicity) compared to
PAO through introduction of a 1,2-ethanedithiol on the
structure of PAO.25,33 We envisioned that a FRET platform
was constructed by combining two fluorophores to the
trivalent arsenic based receptor with similar structure to
VTA2. These two fluorophores were the energy donor and
acceptor of a FRET pair. Thus, we designed the target probe
VTAF (Scheme 1), which bears a coumarin and a
naphthalimide for signal transduction, and the aforementioned
arsenic moiety (VTAF1, Supporting Information Scheme S2)
for selective interaction with VDPs. When excited at 405 nm,
the FRET will move from the coumarin to the naphthalimide.
After vicinal dithiols on VDPs selectively interchange with 2,3-
dimercaptopropanol moiety of VTAF, two fluorophores are
no longer in proximity, which is essential for efficiency, and
hence, FRET is prohibited (Scheme 1). Thus, the ratiometric
signal will be related to VDPs. Meanwhile, VDPs will also be
specifically labeled by coumarin moiety, which will make the
in situ tracing and further proteomic study of VDPs possible.
As the fluorescence signal of coumarin moiety was released
after the VDPs were specifically labeled by VTAF, a wash-out
of the unreacted probe is no longer necessary.
The synthetic route for preparing VTAF was depicted in
Supporting Information Scheme S2. First, protecting group
2,3-dimercaptopropanol was introduced into PAO to furnish
VTAF1, which was further modified by coumarin (FRET
donor) and 1,8-naphthalimide (FRET acceptor). The
integration of two diglycol amine groups at 4,5-position of
1,8-naphthalimide can tune the hydrophilicity and lipophilicity
of VTAF, which enables both the water solubility and cell-
permeability of the probe. VTAF was synthesized in nine
steps under mild conditions. Similarly, a control probe
without trivalent arsenic group (VTAFC, Scheme 1a) was
also synthesized (Supporting Information Scheme S3).
Scheme 1. (a) Chemical Structures of VTAF and VTAFC;
(b) Design Strategy of VTAF for Selective and Ratiometric
Measurements of VDPs
Spectroscopic Properties and Optical Responses to
VDPs. A pH titration was conducted. As displayed by the
emission at 540 and 470 nm, respectively, negligible
fluorescence changes were observed during the pH range of
6−10 (Supporting Information Figure S1). Thus, a PBS buffer
(10 mM PBS, 5 mM EDTA, pH 7.4, 0.5% DMSO) was
chosen as the test medium in the molecular level assay.
Meanwhile, a stability test for VTAF in PBS buffer was also
conducted to check whether the urethane group of VTAF will
be hydrolyzed by water at neutral pH. As shown in
Supporting Information Figure S2, no fluorescence changes
were observed, which indicated that VTAF was not
hydrolyzed even after 3 h. Next, the fluorescence response
of VTAF toward VDPs was tested. Reduced bovine serum
albumin (rBSA) was taken as the model VDPs. There were
about 17 pairs of vicinal thiols in rBSA after the BSA was
treated with reducing agent dithiothreitol (DTT). As shown
in Figure 1a, upon addition of 1 equiv amount of rBSA,
fluorescence emission centered at 470 nm increased
14238
dx.doi.org/10.1021/ja5079656 | J. Am. Chem. Soc. 2014, 136, 14237−14244

Journal of the American Chemical Society
Article
To test whether other biological analytes could interfere
with this detection, fluorescence responses of VTAF for some
small amino acids (or peptide) were also studied. As shown in
Figure 2a, no significant fluorescence changes of VTAF were
Figure 1. (A) Time-dependent fluorescence response of VTAF (2
μM) to 1 equiv rBSA (2 μM) in PBS buffer. (B) Ratiometric
fluorescence response of VTAF (2 μM) to different concentrations
of rBSA in PBS buffer. (C) Ratiometric determination of rBSA by
calculation of ratio of emission intensity at 470 and 540 nm. (D)
Linear relationship between ratio and concentration of rBSA (from
0.14 to 1.6 μM). λex =405 nm. Error bars (SD) represent three
independent experiments; log(ratio) = log(ratio I470/I540), rBSA =
reduced BSA (with vicinal dithiols).
Figure 2. Selectivity of VTAF to rBSA. (A) Changes in fluorescence
emission spectra of VTAF (1 μM) with addition of rBSA (1 μM)
and other small amino acids (100 μM) after 60 min in PBS buffer.
(B). Changes in fluorescence emission of VTAF (2 μM) with
addition of rBSA (2 μM) and oBSA (2 μM). (C) Ratio changes in
fluorescence intensity (based on the peak heights at the maxima, 470
and 540 nm, respectively) with addition of rBSA and other
interfering reagents after 60 min. λex = 405 nm. Error bars (SD)
represent three independent experiments. (D) Selectivity of VTAF to
VDPs was identified by SDS-PAGE. rBSA = reduced BSA (with
vicinal dithiols), oBSA = oxidized BSA (without vicinal dithiols).
remarkably with a concomitant minor increase of fluorescence
emission at 540 nm under the excitation wavelength at 405
nm. It was observed that the fluorescence intensity changed
immediately and then reached a plateau within 60 min. This
result demonstrates that VTAF interacts with rBSA very fast
in the beginning, and then reached equilibrium. Control probe
VTAFC was also applied for the interaction with rBSA. As
suggested in Supporting Information Figure S3, negligible
changes in the fluorescence signal were observed in the
presence of 1 equiv of rBSA, even the reaction time extended
to 60 min. These results indicated that the trivalent arsenic
moiety of VTAF was needed for its selective interaction with
VDPs. In the following molecular-based assays, the detection
was allowed to react for 60 min. To test the sensitivity of
VTAF for rBSA, the changes in fluorescence emission were
recorded upon gradual addition of rBSA (from 0 to 2 μM).
Excitation at 405 nm produces a fluorescence spectrum
featuring one dominant emission band centered at 470 nm
(Figure 1b). When the amount of rBSA increased to 2 μM,
fluorescence changes at 470 nm reached a plateau, which can
be directly reflected by the fluorescence ratio of emission at
470 and 540 nm (ratio I470/I540, Figure 1c). From the
concentration-dependent ratio changes, the detection limit of
VTAF for rBSA was determined to be 0.14 μM. Then, we
take the logarithm of ratio I470/I540 and concentration of
rBSA, which was defined as log (ratio I470/I540) and
log(rBSA), respectively. Through a linear regression by Origin
observed in the presence of 100-fold amounts of different
common amino acids (or peptide) such as cysteine (Cys),
homocysteine (Hcy), glutathione (GSH), etc. In contrast,
upon addition of 1 equiv of rBSA, a blue shift of about 70 nm
can be observed. Similarly, 1 equiv of oxidized BSA (oBSA,
which was treated with hydrogen peroxide as an oxidant to
remove vicinal dithiols on BSA) also induced negligible
fluorescence changes of VTAF (Figure 2b). These results can
be recorded by the ratio changes of fluorescence intensity
(Ratio I470/I540, Figure 2c), from which the high selectivity of
VTAF to rBSA can be clearly observed. The ratio of
fluorescence intensities at 470 and 540 nm varies from 0.5
(in the absence of rBSA) to 3.2 (in the presence of rBSA)
with a 6-fold increase. Other common biological analytes
(even the concentration was 100-fold excess over VTAF)
exhibited no discernible ratio changes. Specifically, the
ratiometric sensing of rBSA is unperturbed by the varied
concentrations of VTAF (Supporting Information Figure S4).
These data demonstrated that VTAF represents a more
accurate probe for determination of VDPs through the
ratiometric signals. Meanwhile, electrophoresis studies were
also carried out to further identify the high selectivity of
VTAF to VDPs. A fluorescent band was observed only in the
lane loaded with rBSA and VTAF, whereas the lane loaded
with oBSA and VTAF exhibited no fluorescence signal.
Coomassie Brilliant Blue (CBB) staining demonstrated that
the fluorescent band was related to the formations of rBSA−
probe complex (Figure 2d). There was no fluorescent band
observed when VTAFC, which lacks of the cyclic dithiaarsane
moiety, was used for labeling. These results indicated that the
8.0 software, a linear relationship between log(ratio I470/I540
)
and log(rBSA) was observed (Figure 1d). Hence, a linear
relationship between changes of ratio I470/I540 and the
concentrations of rBSA (changes from 0.14 to 1.6 μM,
Figure 1c,d) can be deduced. The obtained linear curve makes
quantitative detection of rBSA very convenient over this
concentration range. This also presents a platform for
quantitative determination of VDPs by using VTAF.
14239
dx.doi.org/10.1021/ja5079656 | J. Am. Chem. Soc. 2014, 136, 14237−14244

Journal of the American Chemical Society
Article
cyclic dithiaarsane moiety in VTAF was cleaved by vicinal
dithiols of rBSA, and the coumarin moiety was attached to
rBSA through the interaction of vicinal dithiols of rBSA with
trivalent arsenic of coumarin moiety (Scheme 1). All these
results demonstrated that VTAF was a highly selective
fluorescent probe for VDPs.
MCF-7 cells contain a high level of VDPs. An approximately
5-fold decrease in ratio for cells treated with VTAF and PAO
was observed, which reveals that the labeling of cellular VDPs
was inhibited by PAO through the competitive binding of
PAO with vicinal dithiols on VDPs. The ratio of cells treated
with VTAFC was about 0.21, which is approximately 12-fold
decrease compared to the VTAF-loaded cells. These results
demonstrated that VTAF was a highly selective and effective
tool for ratiometric trapping of endogenous VDPs in live cells.
In the case of protein labeling in live cells, many developed
labeling technologies based on synthetic small fluorescent
probes need a “washout” procedure to remove intracellular
free probes. Herein, VTAF can specifically label endogenous
VDPs without long-time washing procedure. In the specific
protocol, cells were incubated with VTAF in a fresh medium
(serum-free DMEM) for 30 min. After the culture medium
was replaced with fresh PBS to remove extracellular free
VTAF, cell images were conducted under confocal microscopy
immediately. A clear strong fluorescence signal at coumarin
channel and a relatively weak fluorescence signal at
naphthalimide channel were observed (Supporting Informa-
tion Figure S6, Bottom). In addition, the procedure for
replacing culture medium with fresh PBS is applied to remove
extracellular free VTAF, which will make the semiquantitative
determination of averaged fluorescence intensity at naph-
thalimide channel more accurate. As indicated in Supporting
Information Figure S6 (top), before the culture medium was
replaced with fresh PBS, there were no changes in the
fluorescence signal from cells at both coumarin and
naphthalimide channel except that some extracellular yellow
spots were removed, which will be beneficial for ratiometric
quantitative determination of endogenous VDPs in live cells.
Accordingly, VTAF can specifically label endogenous VDPs in
live cells without long-time washing procedure. Additionally,
live cells were incubated with VTAF in a fresh medium
(serum-free DMEM) without wash procedure, then the cell
imaging of endogenous VDPs was conducted directly under
confocal microscopy for continuous observation of fluores-
cence signal at both coumarin channel and naphthalimide
channel at the time point of 30, 45, and 60 min, respectively.
As shown in Supporting Information Figure S7, the
fluorescence signal at both coumarin channel and naphthali-
mide channel did not change at the time of 30, 45, and 60
min, respectively. Especially for the fluorescence signal at
naphthalimide channel, there is no increase in the extracellular
fluorescence intensity. Similarly, no loss of intracellular
fluorescence intensity was observed at the naphthalimde
channel. Thus, this result demonstrated that the naphthali-
mide unit might not be excluded by cells after the
naphthalimide unit leaves from the coumarin units in
VTAF. And the cellular concentration of naphthalimide unit
did not change after the interaction of VTAF with VDPs,
which could enable an accurate semiquantitative analysis in
live cells. Through the semiquantitative calculation, there was
no significant difference in the ratio at the time point of 30
and 60 min, respectively (Supporting Information Figure S8).
This result confirmed that the interaction between VTAF and
endogenous VDPs in live cells is completed in 30 min.
Compared to the relatively slow response time of VTAF to
rBSA (about 1 h), we deduced that the interaction between
VTAF and VDPs is a dynamic equilibrium (please see the
detailed discussion in the Supporting Information).
In Situ Imaging of Endogenous VDPs with VTAF in
Live Cells. We then explored whether VTAF could be used
to detect the endogenous VDPs in live cells. Previously, some
different VDPs have been unveiled in MCF-7 cells, and these
VDPs are found to be related with redox regulation.34 Thus,
we performed the ratiometric imaging of endogenous VDPs in
living MCF-7 cells by confocal microscopy. Initially, the
cytotoxicity of VTAF was tested by MTT assay. As
demonstrated in Supporting Information Figure S5, above
90% of MCF-7 cells survived after the cells were incubated
with 5 μM of VTAF for 12 h, and after 24 h, the cell viability
remained at 85%, which indicated the very low cytotoxicity of
VTAF when it was used in 5 μM. Thus, concentration of
VTAF was taken as 5 μM for the next cell imaging
experiments. Then incubation of MCF-7 with VTAF at 5
μM was conducted for 30 min at 37 °C. The fluorescence
signal was collected before removal of extracellular free VTAF.
Upon excitation by 405 nm laser line, a strong fluorescence
signal at coumarin channel (emission signal was collected at
450−490 nm) and a relatively weak fluorescence signal at
naphthalimide channel (emission signal was collected at 530−
590 nm) were observed (Figure 3a, top). When VTAF was
Figure 3. (A) In situ fluorescence imaging of cellular VDPs in MCF-
7 cells. VTAF and VTAFC: Cells were treated with 5 μM of VTAF
and VTAFC, respectively. VTAF + PAO: VTAF was coincubated
with 25 μM PAO in cells. Fluorescence was collected at coumarin
channel, 450−490 nm (left, green); naphthalimide channel, 530−590
nm (middle, yellow) upon excitation by 405 nm laser line. Scale bar,
75 μm. (B) Semiquantitative determination of endogenous VDPs in
MCF-7 cells according to the ratio of averaged fluorescence intensity
of coumarin channel (450−490 nm) to naphthalimide channel
(530−590 nm). The semiquantitative calculation was conducted by
ImageJ software. Error bars represent SD.
coincubated with PAO at 25 μM in live cells, the fluorescence
signal at coumarin channel decreased with the increase of
fluorescence signal at naphthalimide channel (Figure 3a,
middle). Treatment of cells with VTAFC (5 μM) leads to
negligible fluorescence signal at coumarin channel and strong
fluorescence signal at naphthalimide channel (Figure 3a,
bottom). The semiquantitative calculation of ratio of
coumarin channel to naphthalimide channel at averaged
emission intensity was further conducted. As suggested in
Figure 3b, the high ratio for VTAF-loaded cells indicates that
14240
dx.doi.org/10.1021/ja5079656 | J. Am. Chem. Soc. 2014, 136, 14237−14244

Journal of the American Chemical Society
Article
Dynamic Tracing of Endogenous VDPs with VTAF in
Live Cells. The successful use of VTAF for trapping of
endogenous VDPs prompted us to apply this probe for the
monitoring of dynamic changes of VDPs under a redox
stimulus in living cells. Diamide is a commonly used thiol
oxidant that can oxidize protein vicinal thiols to disulfide.35
After VTAF and diamide were coincubated in MCF-7 cells,
the fluorescence intensity at coumarin channel decreased and
an increase in fluorescence intensity at naphthalimide channel
was observed (Figure 4a, VTAF+Diamide). Semiquantitative
Through the plot analysis of regions of interest (ROI)
across MCF-7 cells, it is noteworthy that the MCF-7 cell
membrane exhibited brighter fluorescence at coumarin
channel when the cells were loaded with 5 μM VTAF,
followed by exposure to 10 mM DTT 10 min later (Figure
5b), which may indicate an increase in levels of protein vicinal
Figure 4. Dynamic tracing of VDPs under redox reagents stimulus in
live cells according to ratiometric fluorescence change (coumarin
channel (450−490 nm)/naphthalimide channel (530−590 nm)). (A)
Ratiometric fluorescence imaging of VDPs in live cells. Excitation
wavelength: 405 nm. Emission collection: coumarin channel, 450−
490 nm (left, green); naphthalimide channel, 530−590 nm (middle,
yellow). Scale bar, 75 μm. (B) Semiquantitative determination of
endogenous VDPs in MCF-7 cells according to ratio of averaged
fluorescence intensity at coumarin channel (450−490 nm) and
naphthalimide channel (530−590 nm). VTAF, VTAF + Diamide:
cells loaded with VTAF, VTAF and Diamide, respectively. VTAF +
DTT (P): cells pretreated with DTT (10 mM) for 10 min, then
loaded with VTAF. VTAF + DTT (L): cells loaded with VTAF first
and incubation for 10 min, then DTT was loaded. VTAF (5 μM ),
DTT (10 mM), and Diamide (50 μM) were used in these
experiments. Error bars represent SD.
Figure 5. Top: Fluorescence imaging of live cells with VTAF. Scale
bar 75 μm. Middle: Plot analysis of regions of interest (ROI) across
MCF-7 cells. Excitation wavelength: 405 nm. Emission collection:
450−490 nm (coumarin channel). Bottom: Intensity profile of linear
ROI across MCF-7 cells. ROI was conducted by ImageJ software.
Pretreated with DTT: cells pretreated with DTT (10 mM) for 10
min, then loaded with VTAF. Later loading of DTT: cells loaded
with VTAF first and incubation for 10 min, then DTT was loaded.
VTAF (5 μM) and DTT (10 mM) were used in these experiments.
dithiols in cell membrane. To further investigate whether
there is an increase in levels of protein vicinal thiols in living
cell membrane under reductant stimulus, the cells were
preincubated with 10 mM DTT for 10 min, then 5 μM VTAF
was loaded and incubated for another 20 min. As suggested in
Figure 4a (DTT pretreated + VTAF) and Figure 5a
(pretreated with DTT), fluorescence signal at coumarin
channel increased along the cell membranes while the
cytoplasm exhibited very weak fluorescence signal compared
to DTT-untreated cells (Figure 5c). A comparison of enlarged
images based on plot analysis of ROI and intensity profile of
linear ROI across MCF-7 cells (Figure 5, middle and bottom)
further demonstrated that fluorescence signal on MCF-7 cell
membranes increased under DTT stimulus. Similarly, the ratio
value of coumarin channel/naphthalimide channel also
increased to about 4.0 (Figure 4b, VTAF + DTT (P)),
which was 1.6-fold compared to DTT untreated cells. This
result displayed an interesting phenomenon and we speculated
that DTT pretreatment makes some more redox-sensitive
protein disulfides on cell surface become protein vicinal
dithiols that can be specifically labeled and visualized by
VTAF. This hypothesis can be partially supported by previous
study that reveals the existence of some redox active protein
vicinal dithiols on cell surface.36 This interaction between
redox-sensitive cell-surface protein vicinal dithiols and VTAF
determination clearly demonstrated that about 1.8-fold
decrease in ratios of fluorescence intensity was observed,
compared to no diamide treated cells (Figure 4b, VTAF +
Diamide), which indicated that levels of protein vicinal thiols
decreased owning to the formation of disulfide under the
oxidant stimulus. Similarly, cells were loaded with 5 μM
VTAF, followed by exposure to 10 mM DTT 10 min later.
After incubation for another 20 min, the fluorescence signal at
coumarin channel was a little brighter and fluorescence
intensity at naphthalimide channel decreased to baseline levels
(Figure 4a, VTAF + DTT (later loading)). The observed ratio
increase was about 1.3-fold compared to DTT untreated cells
(Figure 4b, VTAF + DTT (L)). These data demonstrated
that VTAF can trace the changes of VDPs under different
cellular redox states. Thus, VTAF can be applied for
identifying and studying the levels of cellular protein vicinal
dithiols that play key roles in cellular redox homeostasis,
which is of considerable significance for both VDPs related
disease diagnosis and exploration of its diverse pathophysiol-
ogy.
14241
dx.doi.org/10.1021/ja5079656 | J. Am. Chem. Soc. 2014, 136, 14237−14244

Journal of the American Chemical Society
Article
hinders the probe’s penetration into cells. Then only a little
VTAF enters the cytoplasm and intracellular VDPs are
visualized. Through the plot analysis of ROI across MCF-7
cells (Figure 5), we can conclude that reductant stimulation
leads to an increase in levels of endogenous protein vicinal
dithiols in MCF-7 cells. Moreover, some redox-sensitive VDPs
are present on cell surface which may regulate the function of
live cells through the reversible protein vicinal dithiols
oxidation to disulfides.
Subcellular Distribution Study of Intracellular VDPs
and Ratiometric Images with VTAF. We have previously
studied the subcellular distribution of intracellular VDPs in
Chang liver cells.25 Herein, the localization of intracellular
VDPs was also investigated in live MCF-7 cells. With the use
of confocal microscopy, the fluorescence signal at coumarin
channel displayed a punctuated pattern mainly concentrated
in a small, eccentric perinuclear zone (Figure 6, VTAF;
(detailed information on Pearson’s correlation coefficient can
be found in Supporting Information Figure S10).37,38 This
results demonstrated that there are abundant VDPs in the
mitochondria of live MCF-7 cells. As the key organelle for
respiration, mitochondria are particularly susceptible to
oxidative damage,39,40 and the mitochondrial VDPs may
play a key role in both oxidative damage and redox signaling.
These results were also consistent with previous report that
abundant arsenic-binding proteins in MCF-7 cell lysate were
mitochondrial related proteins.34
Next, ratiometric fluorescence imaging of endogenous
VDPs in live cells was also investigated. The ratio images
were constructed from fluorescence signal of coumarin
channel and naphthalimide channel by using ImageJ software.
After the cells were treated with VTAF, the subcellular
distribution of endogenous VDPs was observed through the
emission ratio values (Figure 7, ratiometric image). The
Figure 7. Ratiometric fluorescence imaging of endogenous VDPs
with VTAF (5 μM) within live cells. Top: confocal ratiometric
images of MCF-7 cells treated with VTAF. Scale bar 25 μm. Bottom:
enlarged ratiometric image of single cells treated with VTAF. Scale
bar 25 μm. (A) Fluorescence was collected at 450−490 nm
(coumarin channel, left, green), (B) 530−590 nm (naphthalimide
channel, middle, yellow) upon excitation by 405 nm laser line.
Pseudocolor illustrates the ratio image generated from fluorescence
intensities of coumarin channel to naphthalimide channel. Pseudo-
color: yellow = low ratio, blue = high ratio.
Figure 6. Top: Colocalization of VTAF (5 μM) with Mito-Tracker
Deep Red (50 nM) in live MCF-7 cells. Scale bar 25 μm. Bottom:
enlarged images of single cell stained with VTAF and Mito-Tracker
Deep Red. Scale bar 10 μm. Excitation wavelength for VTAF: 405
nm. Emission collection: coumarin channel, 450−490 nm (VTAF,
green). The excitation of Mito-Tracker Deep red is 633 nm, and
>650 nm emission light was collected (mitochondria, red). Overlay:
the merged images of VTAF and mitochondria (overlay, orange).
Pearson’s correlation coefficient of 0.86
0.04 was obtained by
using the JACoP plugin on ImageJ and Image-Pro Plus (IPP 6.0)
software.
cytoplasm matrix showed the emission ratio value of about
0.9. The punctuated patterns mainly concentrated in a small,
eccentric perinuclear zone displayed emission ratio value of
2.7, which is 3-fold higher than ratio values in cytoplasm
matrix. These results further confirmed that VTAF-labled
VDPs mainly localized in mitochondria. Meanwhile, ratio
value of about 0.9 in cytoplasm matrix suggested low levels of
VDPs were also present in cytoplasm matrix. Thus, the
ratiometric images may enable a better understanding of
subcellualr distribution of endogenous VDPs through the
direct readout of ratio values.
Flow Cytometric Analysis of VDPs in Suspension
Cells with VTAF. Finally, we determined whether VTAF can
be applied in flow cytometric analysis in tracing changes in
levels of intracellular protein vicinal dithiols upon redox
regulation in suspension cells. HL-60 cells (human promye-
locytic leukemia cells) were loaded with VTAF (or VTAFC)
and redox stimulants, respectively. Histograms showed the
intracellular levels of vicinal dithiols on VDPs and
Supporting Information Figures S6A and S9A). Meanwhile,
the fluorescence signal in cytoplasm matrix and nucleus was
scarcely observed. By contrast, the weak fluorescence signal at
naphthalimide channel showed a wide pattern in cells
(Supporting Information Figures S6B and S9B). These results
confirmed that the fluorescence signal in punctuated pattern
was mainly from labeled VDPs. And the punctuated pattern
indicated a mitochondrial localization. Then a colocalization
assay was conducted through the coincubation of Mito-
Tracker Deep Red (a commercial probe for specific staining
of mitochondria) with VTAF in live MCF-7 cells. As depicted
in Figure 6, mitochondria were stained by Mito-Tracker Deep
Red and exhibited clear fluorescence signal at red channel
(>650 nm) under excitation at 633 nm. The merged image of
VTAF and mitochondria mainly showed orange fluorescence
signal (Figure 6), which indicates the fluorescence signal at
two channels overlapped well. Quantitative image analysis
showed a high Pearson’s correlation coefficient of 0.86 0.04
14242
dx.doi.org/10.1021/ja5079656 | J. Am. Chem. Soc. 2014, 136, 14237−14244

Journal of the American Chemical Society
Article
fluorescence signal from VTAF (or VTAFC) labeled VDPs
was collected through a 450/50 band-pass filter with a total of
104 cells assessed (Figure 8 and Supporting Information
new endogenous redox-sensitive VDPs with diverse functions
will also be unveiled, which can enable the discovery of
promising new drug target for redox related diseases such as
inflammation and cancers.
ASSOCIATED CONTENT
■
S
* Supporting Information
Detailed experimental procedures and characterization of
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Authors
wpzhu@ecust.edu.cn
nqjia@shnu.edu.cn
Figure 8. (a) Flow cytometric analysis of HL-60 cells loaded with
VTAF and redox stimulants (10 mM DTT and 50 μM Diamide).
Violet 1 channel: 425−475 nm (450/50 band-pass filter). (b)
Normalized mean fluorescence intensity per cell. Blank: HL-60 cells
loaded without VTAF. VTAF, VTAFC, and VTAF + Diamide: cells
loaded with VTAF, VTAFC, VTAF and Diamide, respectively.
VTAF + DTT: cells loaded with VTAF first and incubation for 10
min, then DTT was loaded and incubation for another 20 min.
yfxu@ecust.edu.cn
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank the financial support from National Natural Science
Foundation of China (Grants 21476077, 21236002,
21373138, 21302125), National Basic Research Program of
China (973 Program, 2013CB733700, 2010CB126100),
National High Technology Research and Development
Program of China (863 Program, 2011AA10A207), Shanghai
Pujiang Program, Doctoral Fund of Ministry of Education of
China (Grant No. 20133127120005), Program for Shanghai
Sci. & Tech. Committee (Grants 13ZR1458800), Fundamen-
tal Research Funds for the Central Universities, Program for
Cultivation of Young Teacher of Shanghai University.
Figure S11). Compared to cells treated with VTAFC, a
significant increase in fluorescence intensity was observed
from the histogram of the VTAF-stained HL-60 cells (Figure
8a). In the case of cells loaded with VTAF and diamide, a
decrease in the fluorescence intensity was observed. Similarly,
the higher fluorescence intensity was observed in the VTAF
and DTT treated cells (Figure 8a). Quantification of mean
fluorescence intensity per cell clearly establishes a dynamic
tracing in intracellular levels of proteins’ vicinal dithiols under
redox stimulus (Figure 8b). Notably, these results demon-
strated that VTAF can be used in flow cytometry for
simultaneous quantification of endogenous protein vicinal
dithiols in HL60 cells. Hence, the VTAF-based flow
cytometric analysis may present a new tool for the accurate
evaluation of the roles of vicinal dithiols on VDPs against
oxidative stress, which would be helpful for a better
understanding of essential roles of protein vicinal dithiols in
inflammation and cancer.
REFERENCES
■
(1) Maron, B. A.; Tang, S. S.; Loscalzo, J. Antioxid. Redox Signaling
2013, 18, 270−287.
(2) Reddie, K. G.; Carroll, K. S. Curr. Opin. Chem. Biol. 2008, 12,
746−754.
(3) Poole, L. B.; Nelson, K. J. Curr. Opin. Chem. Biol. 2008, 12,
18−24.
(4) Ying, J.; Clavreul, N.; Sethuraman, M.; Adachi, T.; Cohen, R. A.
Free Radical Biol. Med. 2007, 43, 1099−1108.
(5) Valko, M.; Rhodes, C. J.; Moncol, J.; Izakovic, M.; Mazur, M.
Chem−Biol. Interact. 2006, 160, 1−40.
(6) Zhang, X.; Yan, X.; Zhou, Z.; Yang, F.; Wu, Z.; Sun, H.; Liang,
W.; Song, A.; Lallemand-Breitenbach, V.; Jeanne, M.; Zhang, Q.;
Yang, H.-Y.; Huang, Q.; Zhou, G.; Tong, J.; Zhang, Y.; Wu, J.; Hu,
H.; de The, H.; Chen, S.; Chen, Z. Science 2010, 328, 240−243.
(7) Houstis, N.; Rosen, E. D.; Lander, E. S. Nature 2006, 440,
944−948.
(8) Sugatani, J.; Steinhelper, M. E.; Saito, K.; Olson, M. S.;
Hanahan, D. J. J. Biol. Chem. 1987, 262, 16995−17001.
(9) Matthias, L. J.; Hogg, P. J. Antioxid. Redox Sign. 2003, 5, 133−
138.
(10) Sahaf, B.; Heydari, K.; Herzenberg, L. A.; Herzenberg, L. A.
Arch. Biochem. Biophys. 2005, 434, 26−32.
CONCLUSION
■
In summary, we have presented a ratiometric fluorescent
probe (VTAF) for highly selective detection of endogenous
protein vicinal dithiols in live cells for the first time.
Compared to previous fluorescent probes for VDPs,25,26
VTAF embodies the quantitative determination of vicinal
dithiols of VDPs through the readout of ratiometric
fluorescence signal. Furthermore, VTAF was applied for
specifically labeling cellular VDPs without washing procedures,
which was beneficial for the in situ trapping of endogenous
VDPs. In particular, the dynamic changes of cellular vicinal
dithiols of VDPs can be traced by VTAF when the cells were
stimulated by different redox reagents. The subcellular
distribution of endogenous VDPs was clearly visualized in
live cells through colocalization experiments and ratiometric
images. This ratiometric fluorescent probe offers a noninvasive
tools for collecting the global distribution and dynamic
changes of endogenous VDPs in live cells, which may greatly
contribute to further investigation on the essential roles of
protein vicinal dithiols in cellular redox homeostasis and
protein function in living cells. Through this approach, some
(11) Ou, W.; Silver, J. Virology 2006, 350, 406−417.
(12) Jain, S.; McGinnes, L. W.; Morrison, T. G. J. Virol. 2008, 82,
12039−12048.
(13) Cook, K. M.; Hogg, P. J. Antioxid. Redox Signaling 2013, 18,
1987−2015.
(14) Borges, C. R.; Lake, D. F. Antioxid. Redox Signaling 2014, 21,
392−395.
(15) Walsh, C. T.; Garneau-Tsodikova, S.; Gatto, G. J. Angew.
Chem., Int. Ed. 2005, 44, 7342−7372.
14243
dx.doi.org/10.1021/ja5079656 | J. Am. Chem. Soc. 2014, 136, 14237−14244

Journal of the American Chemical Society
Article
(16) Gilbert, H. F. In Methods in Enzymology; Lester, P., Ed.;
Academic Press: New York, 1995; Vol. 251, pp 8−28.
(17) Bachi, A.; Dalle-Donne, I.; Scaloni, A. Chem. Rev. 2013, 113,
596−698.
(18) Requejo, R.; Chouchani, E. T.; James, A. M.; Prime, T. A.;
Lilley, K. S.; Fearnley, I. M.; Murphy, M. P. Arch. Biochem. Biophys.
2010, 504, 228−235.
(19) Hoffman, R. D.; Lane, M. D. J. Biol. Chem. 1992, 267, 14005−
14011.
(20) Gitler, C.; Zarmi, B.; Kalef, E. Anal. Biochem. 1997, 252, 48−
55.
(21) McStay, G. P.; Clarke, S. J.; Halestrap, A. P. Biochem. J. 2002,
367, 541−548.
(22) Moaddel, R.; Sharma, A.; Huseni, T.; Jones, G. S.; Hanson, R.
N.; Loring, R. H. Bioconjugate Chem. 1999, 10, 629−637.
(23) Heredia-Moya, J.; Kirk, K. L. Biorg. Med. Chem. 2008, 16,
5743−5746.
(24) Turnbull, L.; Campbell, E. M.; Thomas, N. Science 2013, 340,
766.
(25) Huang, C.; Yin, Q.; Zhu, W.; Yang, Y.; Wang, X.; Qian, X.;
Xu, Y. Angew. Chem., Int. Ed. 2011, 50, 7551−7556.
(26) Fu, N.; Su, D.; Cort, J. R.; Chen, B.; Xiong, Y.; Qian, W.-J.;
Konopka, A. E.; Bigelow, D. J.; Squier, T. C. J. Am. Chem. Soc. 2013,
135, 3567−3575.
(27) Whitney, M.; Savariar, E. N.; Friedman, B.; Levin, R. A.; Crisp,
J. L.; Glasgow, H. L.; Lefkowitz, R.; Adams, S. R.; Steinbach, P.;
Nashi, N.; Nguyen, Q. T.; Tsien, R. Y. Angew. Chem., Int. Ed. 2013,
52, 325−330.
(28) Chang, C. J.; Jaworski, J.; Nolan, E. M.; Sheng, M.; Lippard, S.
J. Proc. Natl. Acad. Sci. U S A 2004, 101, 1129−1134.
(29) Zhou, X.; Su, F.; Lu, H.; Senechal-Willis, P.; Tian, Y.; Johnson,
R. H.; Meldrum, D. R. Biomaterials 2012, 33, 171−180.
(30) Fan, J.; Hu, M.; Zhan, P.; Peng, X. Chem. Soc. Rev. 2013, 42,
29−43.
(31) Takakusa, H.; Kikuchi, K.; Urano, Y.; Sakamoto, S.;
Yamaguchi, K.; Nagano, T. J. Am. Chem. Soc. 2002, 124, 1653−1657.
(32) Zahler, W. L.; Cleland, W. W. J. Biol. Chem. 1968, 243, 716−
719.
(33) Huang, C.; Yin, Q.; Meng, J.; Zhu, W.; Yang, Y.; Qian, X.; Xu,
Y. Chem.Eur. J. 2013, 19, 7739−7747.
(34) Zhang, X.; Yang, F.; Shim, J.-Y.; Kirk, K. L.; Anderson, D. E.;
Chen, X. Cancer Lett. 2007, 255, 95−106.
(35) Kosower, N. S.; Kosower, E. M.; Wertheim, B.; Correa, W. S.
Biochem. Biophys. Res. Commun. 1969, 37, 593−596.
(36) Donoghue, N.; Yam, P. T. W.; Jiang, X.-M.; Hogg, P. J. Protein
Sci. 2000, 9, 2436−2445.
(37) Adler, J.; Parmryd, I. Cytometry, Part A 2010, 77, 733−742.
(38) French, A. P.; Mills, S.; Swarup, R.; Bennett, M. J.; Pridmore,
T. P. Nat. Protocols 2008, 3, 619−628.
(39) Murphy, M. P. Biochem. J. 2009, 417, 1−13.
(40) Finkel, T. Nat. Rev. Mol. Cell Biol. 2005, 6, 971−976.
14244
dx.doi.org/10.1021/ja5079656 | J. Am. Chem. Soc. 2014, 136, 14237−14244