Design and synthesis of a FlAsH-type Mg2+ fluorescent probe for specific protein labeling

Article abstract of DOI:10.1021/ja410031n

Although the magnesium ion (Mg²⁺) is one of the most abundant divalent cations in cells and is known to play critical roles in many physiological processes, its mobilization and underlying mechanisms are still unknown. Here, we describe a novel

Full text of DOI:10.1021/ja410031n

Article
pubs.acs.org/JACS
Design and Synthesis of a FlAsH-Type Mg2+ Fluorescent Probe for
Specific Protein Labeling
Tomohiko Fujii,†,‡,⊥ Yutaka Shindo,†,⊥ Kohji Hotta,† Daniel Citterio,§ Shigeru Nishiyama,∥ Koji Suzuki,§
and Kotaro Oka*,†
†Department of Biosciences and Informatics, §Department of Applied Chemistry, and ∥Department of Chemistry, Faculty of Science
and TechnologyKeio University, 3-14-1 Hiyoshi, Kohoku-ku, Yokohama 223-8522, Japan
‡Graduate School of Medicine, University of Tokyo, 7-3-1 Hong, Bunkyo-ku, Tokyo 113-0033, Japan
S
* Supporting Information
ABSTRACT: Although the magnesium ion (Mg2+) is one of the
most abundant divalent cations in cells and is known to play
critical roles in many physiological processes, its mobilization and
underlying mechanisms are still unknown. Here, we describe a
novel fluorescent Mg2+ probe, “KMG-104-AsH”, composed of a
highly selective fluorescent Mg2+ probe, “KMG-104”, and a
fluorescence-recoverable probe, “FlAsH”, bound specifically to a
tetracysteine peptide tag (TCtag), which can be genetically
incorporated into any protein. This probe was developed for
molecular imaging of local changes in intracellular Mg2+
concentration. KMG-104-AsH was synthesized, and its optical properties were investigated in solution. The fluorescence
intensity of KMG-104-AsH (at λem/max = 540 nm) increases by more than 10-fold by binding to both the TCtag peptide and
Mg2+, and the probe is highly selective for Mg2+ (Kd/Mg = 1.7 mM, Kd/Ca ≫ 100 mM). Application of the probe for imaging of
Mg2+ in HeLa cells showed that this FlAsH-type Mg2+ sensing probe is membrane-permeable and binds specifically to tagged
proteins, such as TCtag-actin and mKeima-TCtag targeted to the cytoplasm and the mitochondrial intermembrane space. KMG-
104-AsH bound to TCtag responded to an increase in intracellular Mg2+ concentration caused by the release of Mg2+ from
mitochondria induced by FCCP, a protonophore that eliminates the inner membrane potential of mitochondria. This probe is
expected to be a strong tool for elucidating the dynamics and mechanisms of intracellular localization of Mg2+.
nisms, however, measuring the changes in Mg2+ concentration
in specific areas of cells is required. KMG-104 diffuses over the
INTRODUCTION
The magnesium ion (Mg2+) is an abundant divalent cation in
■
cells. Intracellular Mg2+ plays critical roles in many physio-
entire cytoplasm and is not appropriate for analyzing the
dynamics of intracellular Mg2+ around particular proteins, target
logical processes including the regulation of hundreds of
enzymes, activities of membrane proteins including ion
organelles, and Mg2+ transporters or channels.
In the present study, therefore, a novel Mg2+ selective
fluorescent probe, KMG-104-AsH, was developed by combin-
ing KMG-104 and the fluorescent probe for protein-labeling,
“FlAsH”. FlAsH-type probes are nonfluorescent in the free state
and have a high specific affinity in their biarsenical structure for
tetracysteine-tagged proteins.16,17 When the probe binds to the
tetracysteine-tag (TCtag) to form the probe−TCtag complex,
its fluorescence turns on. However, KMG-104-AsH has a
double quenching mechanism: one from the biarsenical−TCtag
binding system in the FlAsH substructure, and a second from
the Mg2+ binding system in KMG-104 substructure. Accord-
ingly, KMG-104-AsH only becomes fluorescent when binding
to both the TCtag and Mg2+. Therefore, this probe was shown
to be useful for localized molecular imaging of changes in Mg2+
concentration.
channels,1−4 and the activity of enzymes involved in ATP
synthesis in mitochondria.5 Mg2+ is also involved in several
disorders. A decrease in Mg2+ concentration and abnormal
metabolic processing of intracellular Mg2+ have been implicated
in the development of migraines, diabetes, hypertension, and
Parkinson’s disease.6−10 Furthermore, Mg2+ has been reported
to exert preventive and ameliorating effects in a rat model of
Parkinson’s disease based on 1-methyl-4-phenylpyridinium
(MPP+) toxicity in dopaminergic neurons.11 Although Mg2+
has been shown to modulate many phenomena in cells, the
underlying molecular mechanisms have not been elucidated in
detail despite the development of several Mg2+ probes.12 This
can partly be attributed to the fact that, in comparison to
indicators for Ca2+, many Mg2+ indicators designed for use in
cells are not sufficiently selective. Therefore, our group has
developed the highly selective Mg2+ indicator KMG-104.13
KMG-104 has been used widely and revealed Mg2+
mobilization in cytoplasm in various types of cells.14,15 For
detailed understanding of Mg2+ mobilization and its mecha-
Received: September 30, 2013
Published: January 21, 2014
© 2014 American Chemical Society
2374
dx.doi.org/10.1021/ja410031n | J. Am. Chem. Soc. 2014, 136, 2374−2381

Journal of the American Chemical Society
Article
evaporated in vacuo to give crude 7 (138 mg, quant) as a dark red
solid, which was used without further purification.
EXPERIMENTAL SECTION
■
Synthesis and General Procedures. NMR spectra were
obtained on JEOL JNM-EXC400, JNM-AL400, and JNM-
ALPHA400 spectrometers [1H (400 MHz), 13C (100 MHz)] in a
deuteromethanol (CD3OD) solution using CD2HOD as an internal
standard and a DMSO-d6 solution using DMSO5d as an internal
standard. High-resolution ESI-mass spectra were obtained on a Waters
LCT Premier XE (Waters, Milford MA). HPLC purification was
performed on a system consisting of a JASCO PU-2080 Plus pump
equipped with a JASCO UVIDEC-100-V detector using a Senshu Pak
PEGASIL ODS column (10ϕ × 150 mm) (Senshu Scientific, Tokyo,
Japan). The solvents used for HPLC were obtained from Nakalai
Tesque (Kyoto, Japan) and filtered through a MILLIPORE
OMNIPORE membrane filter (0.45 μm pore size, 47 mm diameter,
Millipore, Billerica, MA) before use. Silica gel column chromatography
was performed on silica gel 60N (spherical, neutral, 63-210 μm, Kanto
Chemical, Tokyo, Japan). Reversed-phase open-column chromatog-
raphy was performed on PEGASIL PREP ODS-7515-12A (Senshu
Scientific). TLC was performed on silica gel plates F254 [0.25 mm
(analytical) and 0.50 mm (preparative)] (Merck, Tokyo, Japan). 4-
Fluororesorcinol was purchased from Sigma-Aldrich (St. Louis, MO).
Synthesis of KMG-104-AsH. 1-(Bis(2,4-dihydroxy-5-
fluorophenyl)methyl)-4-oxo-4H-quinolizine-3-carboxylate Ethyl
Ester (5). A stirred solution of ethyl 1-formyl-4-oxo-4H-quinolizine-
3-carboxylate (compound 4) (284 mg, 1.16 mmol) and 4-
fluororesorcinol (330 mg, 2.58 mmol) in CH2Cl2/Et2O (= 1:1, 6.0
mL) was mixed with methanesulfonic acid (0.48 mL, 8% (v/v) vs
solvent) at room temperature under Ar.18 After the solution was
stirred for 17 h, ice-cold water (30 mL) was added into the mixture in
an ice bath to give a yellow precipitate, and the supernatant was
removed by decantation. The yellow residue was dissolved in MeOH,
and the solution was neutralized by Amberlite IRA-400 (OH−) and a
small amount of saturated NaHCO3 aq to pH 6−7 and filtered. The
filtrate was evaporated in vacuo and purified by silica gel column
chromatography (chloroform−methanol = 7:1) to give 531 mg (95%)
of compound 5 as a yellow powder.
1-(2,7-Difluoro-6-hydroxy-4,5-bis(1,2,3-dithioarsolan-2-yl)-3-oxo-
3H-xanthen-9-yl)-4-oxo-4H-quinolizine-3-carboxylic Acid (1, KMG-
104-AsH). A stirred solution of crude 7 (138 mg) in dry NMP (2.0
mL) was carefully mixed with anhydrous diisopropylethylamine (0.40
mL), arsenic trichloride (0.15 mL, 1.76 mmol), and palladium(II)
acetate (50 mg, 0.23 mmol) under Ar. After being stirred at room
temperature for 18 h, the mixture was poured into 20 mL of 0.5 M
potassium phosphate buffer (pH 6.2)/acetone (3:2) to give a dark red
solution. After the addition of 1,2-ethanedithiol (0.50 mL) and
chloroform (16 mL), the mixture was stirred at room temperature for
30 min. The mixture was diluted with water (30 mL) and extracted
with chloroform (15 mL × 3). The organic layers were combined and
dried with Na2SO4. After evaporation, the crude product was purified
by silica gel chromatography (chloroform−EtOAc = 1:1 + 0.1% TAF
to chloroform−EtOAc = 1:3 + 0.1% TAF). The eluted orange fraction
was concentrated, triturated with 50% EtOH, and filtered. The
obtained pellet was dried overnight on the filter without vacuum19 to
give 15.3 mg (17% in three steps) of the final product 1 as a bright
orange powder.
1: 1H NMR (400 MHz, CDCl3−CD3OD (1:1), rt) 9.18 (1H, d, J =
7.3 Hz), 8.03 (1H, d, J = 1.9 Hz), 7.51 (1H, t, J = 7.3 Hz), 7.07 (1H, d,
J = 9.8 Hz), 6.86 (1H, t, J = 7.3 Hz), 6.50 (2H, d, J = 10.8 Hz), 3.04
(8H, m) ppm; 13C NMR (100 MHz, DMSO-d6, rt) 174.8, 166.4,
160.9, 159.3, 159.1, 158.9, 158.6, 157.4, 151.0, 148.6, 144.5, 140.2,
137.5, 130.0, 124.8, 120.1, 119.6, 117.9, 117.3, 114.3, 112.4, 109.4,
105.3, 104.2, 42.7, 42.6 ppm; HR ESI MS (positive) [M]+ found m/z
767.8420, C27H18NO6S4As2F2 requires 767.8417.
Fluorescent Properties of KMG-104-AsH. A 1 mM solution of
KMG-104-AsH in DMSO was prepared, and 1 μL was added into 1
mL of 10 mM MOPS/KOH (pH 7.3) buffer containing 100 mM KCl,
10 mM MESNa, 3 mM TCEP, 10 μM EDT, MgCl2, CaCl2, or other
ions at varying concentrations, and TCtag peptide (Ac-
FLNCCPGCCMEP) (0 or 1 μM in 50% acetonitrile containing
0.1% TFA). Custom TCtag peptide and its analogues were purchased
from Toray Research Center. Absorbance spectra were recorded on a
U-2001 spectrophotometer (HITACH) using quartz cuvettes.
Fluorescence spectra were recorded in a F-4500 fluorophotometer
(HITACHI) using quartz cuvettes. The excitation wavelength was set
to 505 nm.
5: 1H NMR (CD3OD) 9.48 (1H, d, J = 6.9 Hz), 7.93 (1H, s), 7.90
(1H, d, J = 8.8 Hz), 7.81 (1H, dt, J = 1.5, 6.9 Hz), 7.41 (1H, dt, J = 1.5,
8.3 Hz), 6.47 (2H, dd, J = 2.5, 7.8 Hz), 6.40 (2H, d, J = 12.2 Hz), 6.20
(1H, s), 4.27 (2H, q, J = 6.8, Hz), 1.29 (3H, t, J = 6.8 Hz) ppm; 13C
NMR (CD3OD) 167.6, 157.3, 152.1, 147.8, 146.1, 145.5, 145.2, 145.1,
140.7, 136.0, 130.9, 123.9, 120.7, 120.6, 118.8, 118.3, 118.3, 117.3,
117.1, 105.9, 105.1, 61.7, 38.5, 14.6 ppm; HR ESI MS [M]+ found m/
Determination of the Apparent Dissociation Constant. The
apparent dissociation constant (Kd) of the KMG-104-AsH-TCtag
complex for Mg2+ was calculated as 1.7 mM. To determine the Kd, the
Benesi−Hildebrand plot method was used. The increase in
fluorescence intensity (I − I0) was measured and fitted to the
following equation
z 484.1207, C25H19NO7F requires 484.1208.
2
1-(2,7-Difluoro-6-hydroxy-3-oxo-3H-xanthen-9-yl)-4-oxo-4H-qui-
nolizine-3-carboxylate Ethyl Ester (6, KMG-104 Ethyl Ester). A
solution of 5 (141 mg, 0.29 mmol) in methanesulfonic acid (3 mL)
was heated to 110 °C for 2 h. After the solution was cooled to 0 °C,
cold water (30 mL) was added to give a dark orange solid. The
precipitate was filtered and washed twice with water. The filtrate was
extracted with EtOAc (30 mL) three times, and the combined organic
layer was washed with brine and evaporated in vacuo. The residue was
combined with the dark orange precipitate and purified by silica gel
chromatography (chloroform−methanol = 5:1 to chloroform−
methanol = 5:1 + 0.1% TFA). The dark orange fraction was further
purified with a silica gel preparative TLC plate F254 (0.50 mm,
MERCK) (chloroform−methanol = 4:1 + 0.1% TFA) to give 80.7 mg
of 6 (60%) as a bright orange powder.
1-(2,7-Difluoro-6-hydroxy-3-oxo-3H-xanthen-9-yl)-4-oxo-4H-qui-
nolizine-3-carboxylic acid (2, KMG-104). A stirred solution of 6 (29
mg, 0.062 mmol) in H2O/EtOH = 1:1 (2.5 mL) was mixed with 1.0 M
LiOH aq (0.25 mL) at 0 °C. The mixture was stirred at room
temperature for 1 h, acidified by TFA to pH 5−6, and extracted four
times with EtOAc. The organic layer was evaporated in vacuo to give 2
(15.7 mg, 58%), which was used for the next step without further
purification.
1/(I − I0) = 1/(A[probe − tag complex])
+ 1/(AKd[probe − tag complex]) × 1/([Mg2+])
where I is the measured fluorescence intensity, I0 is the fluorescence
intensity at 0 mM of Mg2+, and A is a constant. Plotting 1/(I − I0) and
1/([Mg2+]), Kd was obtained from the value of (y-intercept)/(slope)
using an approximate line. Fluorescence intensities for the calculation
of Kd are shown in Figure 3.
Gene Construction. The DNA sequence encoding the TCtag
peptide (FLNCCPGCCMEP) was fused to the N′-terminus of actin
by polymerase chain reaction (PCR). The PCR product was
sequenced to ensure fidelity and subcloned into a pcDNA3.1(+)
vector (Invitrogen, Carlsbad, CA) at the KpnI and EcoR I restriction
enzyme sites. mKeima was purchased from Amalgaam (Tokyo, Japan),
and its C′-terminus was fused to the TCtag sequence. To localize
mKeima−TCtag to the MIS, a MIS localization signal was also
conjugated to the N′-terminus of mKeima−TCtag. MIS targeted
mKeima−TCtag was inserted into the pEGFP-N1 vector (Clontech,
Mountain View, CA), exchanging for EGFP, between the EcoR I and
Not I restriction enzyme sites, and mKeima−TCtag was inserted
between the BamH I and Not I sites. The recombinant plasmids were
1-(2,7-Difluoro-6-hydroxy-4,5-bis(trifluoroacetoxymercuri)-3-
oxo-3H-xanthen-9-yl)-4-oxo-4H-quinolizine-3-carboxylic Acid (7). A
mixture of 2 (51 mg, 0.12 mmol) and HgO (50 mg, 0.23 mmol) in
TFA (3.0 mL) was stirred at room temperature overnight and
2375
dx.doi.org/10.1021/ja410031n | J. Am. Chem. Soc. 2014, 136, 2374−2381

Journal of the American Chemical Society
Article
Figure 1. Molecular design of KMG-104-AsH: (a) KMG-104-AsH (1) was derived from KMG-104 (2), a highly selective Mg2+ probe, and FlAsH
(3), a fluorescent probe that specifically binds to a tetra-cysteine peptide tag (TCtag). (b) KMG-104-AsH only emits fluorescence when bound to
both the TCtag and Mg2+. KMG-104-AsH allows for the detection of local changes in Mg2+ concentration around any protein genetically modified to
express the TCtag.
to the charged β-diketone site20,21 in the quinolidine derivative
amplified in Escherichia coli DH5α (Toyobo, Tokyo, Japan) and
purified using a Qiagen plasmid maxi kit (Qiagen, Tokyo, Japan).
moiety. In its unbound state, the fluorescence emission is
quenched by photoinduced electron transfer (PET).22−25 The
Cell Culture and Transient Transfection. HeLa cells were
cultured in DMEM supplemented with 10% FBS and 1% penicillin/
FlAsH probe is also quenched in its unbound state, presumably
streptomycin (Invitrogen) in an incubator maintained at 37 °C and a
humidified atmosphere of 5% CO2. For experimental use, cells were
by vibrational deactivation or PET mechanisms,17 and becomes
fluorescent when specifically binding to the tetracysteine
peptide sequence (TCtag; amino acid sequence:
FLNCCPGCCMEP) due to the formation of a rigid and
constrained peptide complex.16,17,26 We focused our attention
on the xanthene fluorophore system, which is the molecular
subunit common to both of these fluorescent probes. It was
hypothesized that by attaching the Mg2+-binding moiety of
KMG-104, 4-oxo-4H-quinolizine-3-carboxylic acid, to the 9-
position of 2,7-difluoro-4,5-bis(1,2,3-dithioarsolan-2-yl)-
xanthene in FlAsH, a new Mg2+ probe possessing the functions
of both KMG-104 and FlAsH would be obtained (Figure 1a).
This novel Mg2+ probe “KMG-104-AsH” was designed to emit
bright fluorescence only when bound to both Mg2+ and the
TCtag. This unique property is due to the independent
fluorescence quenching mechanisms at the Mg2+-binding site
(PET-type response) and at the arsenic atoms (free rotation).16
While the FlAsH-type probe is activated upon binding to the
TCtag peptide sequence, KMG-104-AsH remains quenched by
a PET mechanism in the KMG-104 substructure in the absence
of Mg2+. Conversely, without binding to the TCtag peptide,
KMG-104-AsH remains in its turned-off state even in the
presence of sufficient Mg2+. The TCtag is a small peptide tag
that can be genetically fused to any protein and expressed
transiently in cells. KMG-104-AsH is expected to act as
selectively as KMG-104 for the detailed detection of Mg2+
dynamics, and as specifically as FlAsH for the detection of
proteins containing the TCtag (Figure 1b). Because these
combined functionalities could be of value for elucidating the
seeded onto glass-based dishes (Iwaki, Tokyo, Japan). Each amplified
plasmid (2 μg for each dish) was transfected into the HeLa cells with
the Lipofectamine LTX and Plus reagent (Invitrogen). Experiments
were performed 24 h after transfection.
Fluorescence Measurements Using Confocal Scanning
Microscopy. KMG-104-AsH was stored at −20 °C as a 1 mM
stock solution in DMSO. Transfected HeLa cells were incubated at 37
°C for 1 h with 100 μL of HBSS (buffered to pH7.3 by 10 mM
HEPES) containing 1 μM KMG-104-AsH and 10 μM EDT. The cells
were incubated twice for 10 min in HBSS containing 250 μM EDT to
wash out the excess probe binding to peptide sequences other than
TCtag. After the washing solution was removed, the dish was filled
with HBSS, and cells were visualized with a confocal laser scanning
microscope (FV-1000, Olympus, Tokyo, Japan) equipped with a 60×
oil immersion objective lens (Olympus) and a DM405-440/515
dichroic mirror (Olympus). Fluorescence was measured with a
photomultiplier (Olympus). KMG-104-AsH was excited by an Ar
laser (515 nm). For simultaneous measurements of KMG-104-AsH
and mKeima, these probes were excited at 515 nm (with the Ar laser)
and at 440 nm (with the laser diode), respectively. The signals were
separated with a 600 nm dichroic mirror and detected 530−550 nm
(for KMG-104-AsH) and 630−730 nm (for mKeima) using
monochrometers.
Time-lapse imaging of the responses to the FCCP stimulus was
carried out for the above stained cells. FCCP solution was added to
bring the final concentration to 5 μM, and the change in fluorescence
intensity was recorded.
RESULTS
■
Molecular Design. KMG-104 is a Mg2+-selective indicator
showing a fluorescence increase behavior upon binding of Mg2+
2376
dx.doi.org/10.1021/ja410031n | J. Am. Chem. Soc. 2014, 136, 2374−2381

Journal of the American Chemical Society
Article
molecular mechanisms of Mg2+ mobilization, KMG-104-AsH
was synthesized and its properties were analyzed.
emission intensity were observed upon the formation of the
KMG-104-AsH−TCtag−Mg2+ complex (Figure 2 and Table
Synthesis of the KMG-104-AsH. The novel Mg2+
fluorescent probe “KMG-104-AsH” (1) was synthesized as
shown in Scheme 1. The quinolizine derivative 4, which was
Scheme 1. Organic Synthesis of KMG-104-AsH
Figure 2. Absorption and fluorescence emission spectra of KMG-104-
AsH (1 μM) in the absence or presence of Mg2+ ([Mg2+] = 0 or 100
mM) and of the TC tag peptide (Ac-FLNCCPGCCMEP, 0 μM or 4
μM). All spectra were measured at pH 7.3 (10 mM MOPS buffer, 100
mM KCl, 10 mM MESNa, 3 mM TCEP, 10 μM EDT). For the
measurement of emission spectra, KMG-104-AsH was excited at 505
nm.
S1, Supporting Information). In the absence of the TCtag
peptide, there was only a slight increase in the maximum
fluorescence intensity of KMG-104-AsH when the Mg2+ ion
concentration was increased from 0 to 100 mM. In contrast, the
addition of 4 μM TCtag peptide induced a more than 10-fold
increase of the maximum fluorescence intensity dependent on
the Mg2+ concentration. These results suggest that the
fluorescence intensity of KMG-104-AsH increased only upon
binding of the probe molecule to both the TCtag peptide and
Mg2+, as was expected from the design of the molecule. The
Kd/Mg value of KMG-104-AsH was calculated to be 1.7 mM,
which is similar to KMG-104 (2.1 mM).13
prepared as described previously,20 was used as the starting
material. Compound 4 was treated with commercially available
4-fluororesorcinol in the presence of 8% (v/v) catalytic
methanesulfonic acid18 to trigger the Friedel−Crafts reaction
leading to compound 5. Initial attempts to synthesize KMG-
104 ethyl ester (6) by oxidative cyclization of 5 with DDQ in
AcOH/benzene18 were unsuccessful. Therefore, purified 5 was
treated with neat methanesulfonic acid at 110 °C yielding
KMG-104 ethyl ester (6) by the desired dehydrative
cyclization.13,27,28 During purification by silica gel column
chromatography and preparative TLC to obtain pure 6, no
cyclic intermediates were detected, indicating that oxidation
was prevented. In the next step, KMG-104 (2) was obtained by
hydrolysis of 6 with 0.25 M LiOH at room temperature for 1 h.
This synthetic process resulted in an overall yield of KMG-104
of 33%, which is superior to the 5.8% overall yield reported
previously.13 In addition, the mercuration of KMG-104 was
accomplished with mercuric trifluoroacetate, prepared by HgO
addition into TFA.19,29,30 After excess TFA was removed in
vacuo, crude mercuric KMG-104 (7) was treated in one pot
with arsenic trichloride and then 1,2-ethanedithiol (EDT) to
yield the biarsenic compound KMG-104-AsH (1). The
purification of 1 required the trituration of the product after
column chromatography, instead of drying in a vacuum line, to
prevent the potential decrease of purity caused by the removal
of EDT from the arsenic groups and polymerization.19 Finally,
KMG-104-AsH (1) was obtained as an orange powder.
The selective binding affinity of KMG-104-AsH for Mg2+
compared to Ca2+, which is another abundant divalent cation in
cells, was investigated and compared (Figure 3a). No changes
in fluorescence intensity were observed in the presence of Ca2+
concentrations in the range of 0.1−100 mM. The dissociation
constant of KMG-104-AsH for Ca2+ (Kd/Ca) was estimated to
be above 100 mM, which is significantly higher than Kd/Mg of
1.7 mM. Other cations, such as Na+, K+, Ni2+, Mn2+, Co2+, Zn2+,
Fe2+, Cu2+ and Cu+, also had no effect on the fluorescence of
KMG-104-AsH and on the Mg2+ binding ability of KMG-104-
AsH in the concentration range based on physiological
environment (Figure 3b and Figure S1,Supporting Informa-
tion) as well as pH (5.5−8) (Figure 4). The pKa value of this
probe was calculated to be 4.2 in 0.1 M acetic acid buffer
containing TCtag and 100 mM Mg2+ (Figure S2, Supporting
Information). Moreover, in the presence of TCtag analogues
with one or more cysteine residues in the peptide tag replaced
by alanine, KMG-104-AsH showed only slight Mg2+ concen-
tration-dependent fluorescence intensity changes in the
presence of 250 μM EDT (Figure 5). Although KMG-104-
AsH bound to TCtag analogues in the buffer containing 10 μM
EDT, 250 μM EDT, which was used for washing out KMG-
104-AsH in the cell-staining procedure, was sufficient to
dissociate KMG-104-AsH from peptide sequences other than
TCtag sequence (Figure S3, Supporting Information). The
same EDT concentration (250 μM) was also suitable for the
staining of cells withKMG-104-AsH. At lower EDT concen-
tration, the background fluorescence was higher. On the other
hand, at higher EDT concentration, KMG-104-AsH was
Properties of KMG-104-AsH. Although the absorption
spectra of KMG-104-AsH did not show significant spectral
shifts or changes in absorbance, drastic changes in fluorescence
2377
dx.doi.org/10.1021/ja410031n | J. Am. Chem. Soc. 2014, 136, 2374−2381

Journal of the American Chemical Society
Article
Figure 3. Ion selectivity of KMG-104-AsH: selectivity of KMG-104-AsH for Mg2+ compared to Ca2+. (a) Fluorescence intensities of KMG-104-AsH
(1 μM) with TCtag (4 μM) at 540 nm (excitation at 490 nm) were measured in the presence of Ca2+ or Mg2+ at concentrations ranging from 0.1 to
100 mM. (b) Fluorescence intensities of KMG-104-AsH binding to TCtag with (black bar) or without (gray bar) 100 mM Mg2+ were measured in
the presence of Ca2+ (1 mM), Na+ (10 mM), K+ (100 mM), Ni2+, Mn2+, Co2+, Zn2+, Fe2+, Cu2+, or Cu+ (1 μM). Cu+ was reduced from Cu2+ by
addition of 100 equiv of ascorbic acid relative to the Cu2+ concentration. The concentrations of those ions were chosen according to the
physiological concentrations. All measurements were performed in 10 mM MOPS buffer at pH 7.3.
It was also demonstrated that the concentration of KMG-
104-AsH applied for cellular staining is not detrimental to cell
viability (Figure S5, Supporting Information). These results
clearly indicate that KMG-104-AsH is a suitable probe for live
cell imaging.
Application of KMG-104-AsH to Cultured Cells. KMG-
104-AsH was applied to fluorescence imaging in cultured HeLa
cells expressing TCtag-labeled proteins. As shown in Figure 6a,
the TC-tagged actins were clearly stained with KMG-104-AsH,
indicating the formation of a KMG-104-AsH−TCtag−-Mg2+
complex in the intracellular environment. These results also
demonstrated the permeability of the cell membrane to KMG-
104-AsH without acetoxymethylation of the carboxyl group.
The intracellular localization of KMG-104-AsH was comparable
Figure 4. Effect of pH on the fluorescence intensity of KMG-104-AsH
bound to the TCtag: Probe and tag peptide were prepared in buffer at
pH 5.5−6.5 (10 mM MES) and pH 7.0−8.0 (10 mM MOPS) and
then mixed with MOPS buffer with or without Mg2+. The final pH
values are shown on the horizontal axis, and those of KMG-104-Ash
and TCtag peptide are 1 and 4 μM, respectively. The fluorescence
intensities were normalized to the value at pH 7.3 with 100 mM of
Mg2+.
Figure 6. Application to live cell imaging. (a) Fluorescence image of
HeLa cells stained with KMG-104-AsH, expressing TCtag-actin.
Intensive fluorescence on actin filaments (F-actin), weak fluorescence
in the cytoplasm, in which tagged G-actin might bind to KMG-104-
AsH (G-actin), and granular fluorescence, which correspond to the
distribution of actin (Nonspecific), were observed. The nonspecific
fluorescence was often observed in cells not expressing TCtag-actin.
The scale bar represents 20 μm. (b) Fluorescence intensities of KMG-
104-AsH were normalized to the initial value and compared before and
after FCCP treatment of various areas in HeLa cells (n = 5). On F-
actin and G-actin, KMG-104-AsH showed increased fluorescence after
administration of FCCP. On the other hand, no changes in the
fluorescence upon nonspecific staining were detected in TCtag-actin
expressing and nonexpressing cells. Averaged fluorescence intensities
measured for 1 min before stimulation and from 1 to 2 min after
stimulation were compared. Error bars indicate the standard error of
the mean (*P < 0.05 in Student’s t test).
Figure 5. Selectivity for the TCtag: fluorescence intensities of KMG-
104-AsH in the presence of the TCtag or other similar peptide
sequences (TCtag analogues) were compared in MOPS buffer (pH
7.3) containing EDT 250 μM. In the analogue peptide sequences, the
KMG-104-AsH binding sequence (CCPGCC) in the TCtag peptide
sequence (FLNCCPGCCMEP) was replaced by the peptide
sequences shown.
removed from the TCtag sequence (Figure S4, Supporting
Information). The selective recognition of the simultaneous
presence of Mg2+ and the TCtag by KMG-104-AsH indicates
that the binding affinities for Mg2+ and for the TCtag peptide of
the KMG-104 and FlAsH probes have been retained.
2378
dx.doi.org/10.1021/ja410031n | J. Am. Chem. Soc. 2014, 136, 2374−2381

Journal of the American Chemical Society
Article
to that of the commercially available FlAsH in TCtag−actin
expressing HeLa cells (Figure S6, Supporting Information).26
Both of those stained actins polymerized in filamentous form
(F-actin) and monomer actins (G-actin). In KMG-104-AsH-
loaded cells, granular fluorescence was also observed both in
TCtag expressing and nonexpresssing cells, which could be
attributed to the partial uptake of KMG-104-AsH into
intracellular compartments such as the endosome and
lysosome. Nevertheless, this nonspecific fluorescence did not
change in intensity in response to changes in Mg2+
concentration in the cytoplasm (Figure 6b). These results
demonstrated that changes in the fluorescence intensity of
KMG-104-AsH occur only when it is bound to the TCtag, and
the nonspecific TCtag-independent accumulation of KMG-104-
AsH does not affect the measurements of intracellular Mg2+
levels.
Next, the changes in Mg2+ concentration in the cytoplasm
and around mitochondria were compared. To eliminate the
effects of mitochondrial movement, the monomeric red
fluorescent protein mKeima was used as a reference. To
examine the Mg2+ sensitivity of KMG-104-AsH bound to
tagged protein in the intracellular environment, the cell
membrane was rendered permeable with digitonin,29 and
fluorescence ratios of KMG-104-AsH to mKeima were
measured at various Mg2+ concentrations (Figure S7,
Supporting Information). The Kd value was calculated to be
5.7 mM. Although this value is slightly higher than the one
found in MOPS buffer solution (1.7 mM), it is suitable to
measure the changes in intracellular Mg2+ concentrations.
Similar differences in Kd values measured in the intracellular
environment and in solution were also observed for Ca2+ in the
case of Calcium Green FlAsH.29 To localize the mKeima−
TCtag to the proximal region of the mitochondrial matrix, a
mitochondrial intermembrane space (MIS) localization signal31
was added to the N′-terminus of mKeima (Figure 7a). While
the mKeima−TCtag localized to the cytoplasm, the MIS-
targeted mKeima−TCtag localized to mitochondria (Figure
7b). The localization of KMG-104-AsH overlapped with the
distribution of the mKeima−TCtag. The time-courses of the
Mg2+ concentration changes in the cytoplasm and the MIS
were compared. Carbonyl cyanide p-(trifluoromethoxy)
phenylhydrazone (FCCP) initially induced a steep increase in
the Mg2+ concentration in the MIS, followed by a gradual
increase in cytoplasmic Mg2+ concentration (Figure 7c),
indicating the diffusion of Mg2+ from the mitochondria to the
cytoplasm in HeLa cells. No photobleaching was observed for
KMG-104-AsH during time-laps imaging of less than 10 min
(Figure S8, Supporting Information).
Figure 7. Mg2+ imaging in localized intracellular regions. (a) Design of
mKeima-TCtag and mKeima-TCtag with a mitochondrial intermem-
brane space (MIS) localization signal. (b) Fluorescence images of
cytoplasmic and MIS-targeted mKeima-TCtag. mKeima-TCtag was
detected as a diffuse pattern in the cytoplasm, whereas MIS targeted
mKeima-TCtag localized to mitochondria (red). The fluorescence of
KMG-104-AsH overlapped with the distribution of mKeima-TCtag
(green and merged). The scale bar indicates 20 μm. (c) Time-courses
of Mg2+ concentration in the cytoplasm (blue, n = 19) and the MIS
(red, n = 29). The ratio of the fluorescence of KMG-104-AsH to that
of mKeima was normalized to the initial value. Error bars indicate the
standard error mean (*P < 0.05 in Student’s t test).
Mg2+ in the whole cytoplasm and used to demonstrate Mg2+
mobilizations.35−37 In resent years, other groups also developed
unique Mg2+-selective fluorescent probes, such as a Mg2+ probe
suitable for two-photon microscopy, and detected intracellular
Mg2+ concentration changes.12,38,39 To clarify the mechanisms
and roles underlying intracellular Mg2+ concentration changes,
however, a novel type of probe for monitoring the
concentration changes in particular regions in cells is required.
In this study, KMG-104-AsH was developed by combining a
Mg2+-selective probe, KMG-104, and a fluorescent probe for
protein-labeling, FlAsH. KMG-104-AsH maintains the Mg2+
selectivity and the pH insensitivity of KMG-104 (Figures 3 and
4) and binds specifically to the TCtag peptide sequence similar
to FlAsH (Figure 5). It had no toxicity to cells when used at
concentrations relevant for staining (Figure S5, Supporting
Information). Therefore, we concluded that KMG-104-AsH is
an effective probe for measuring local changes in Mg2+
concentration around tagged proteins. In a previous study,
the Ca2+ indicator Calcium green was conjugated to FlAsH, and
DISCUSSION
■
Although Mg2+ plays critical roles in a number of physiological
processes, a clear understanding of the molecular mechanisms
underlying the cellular functions of Mg2+ has been hampered by
the lack of reliable probes for use in cell experiments. The
classical Mg2+ probes such as mag-fura-2 and magnesium-green
have a limited ability for distinguishing between signals
originating from Mg2+ or Ca2+ because of their relatively low
dissociation constant for Ca2+ (Kd/Ca) compared to that for
Mg2+ (Kd/Mg).32,33 KMG series dyes were developed as highly
selective intracellular free Mg2+ probes with an adequate Kd/Mg
for the detection of Mg2+ within the intracellular concentration
range, without responding to Ca2+.13,21,34 KMG-104 was shown
to be effective for detecting changes in the concentration of
2379
dx.doi.org/10.1021/ja410031n | J. Am. Chem. Soc. 2014, 136, 2374−2381

Journal of the American Chemical Society
Article
local Ca2+ dynamics around gap junctions in HeLa cells were
observed and analyzed.29 In this study, by conjugating KMG-
104 to FlAsH, a novel tool for visualizing intracellular local
dynamics of another essential divalent cation, Mg2+, has been
developed. These probes are expected to be powerful tools for
the detailed elucidation of cellular events.
CONCLUSION
■
To detect and image the local change in concentration of Mg2+
in cells, the novel fluorescence Mg2+ selective probe, KMG-104-
AsH, was designed and synthesized. It was derived from the
highly selective Mg2+ probe, KMG-104, and the FlAsH probe
that fluoresces only when it binds specifically to a TCtag
peptide genetically added to target proteins. We showed that
the probe has effective fluorescence properties, in that its
fluorescence increases drastically when it binds to both Mg2+
and TCtag. This effect results from the cancellation of a PET
effect and free rotations of arsenic groups. KMG-104-AsH was
also shown to be membrane-permeable, and it stains TC-tagged
actins and TC-tagged mKeima localized to the mitochondrial
intermembrane space in HeLa cells. We succeeded in detecting
the change in Mg2+ concentration induced by FCCP around
TC-tagged proteins. This probe is expected to be a strong tool
for the elucidation of little known molecular mechanisms and
critical roles of Mg2+ in cells.
KMG-104-AsH was successfully used to visualize actin
filaments transfected with TCtag peptides in HeLa cells (Figure
6a). The results indicated that KMG-104-AsH was sufficiently
membrane-permeable without acetoxymethylation of the
carboxyl group in the quinolizine derivative moiety, which is
necessary for loading KMG-104 into cells, and showed specific
intracellular binding to TCtag, similar to FlAsH. Depending on
the localization of tagged proteins, localization of KMG-104-
AsH to specific areas in cells was enabled. This allows to
measure changes in Mg2+ concentration in specific local regions
in cells. Therefore, TCtagged mKeima was localized to the
region proximal to the mitochondrial matrix, since the release
of Mg2+ from mitochondria upon depolarization of their inner
membrane potential was demonstrated in our previous
studies.35,36 Membrane depolarization caused a sharp increase
in Ca2+ concentration followed by a gradual increase in Mg2+
concentration in the cytoplasm of PC12 cells. However, the
concentration of Mg2+ in a small region around the
mitochondrial matrix might show sharper changes than in the
cytoplasm. Therefore, to test this hypothesis, a localization
signal sequence specific for the MIS was attached to the N′-
terminus of the mKeima−TCtag, and the change in Mg2+ levels
in the region proximal to the mitochondrial matrix was
measured. As shown in Figure 7c, Mg2+ levels showed an
initial steep increase in the MIS followed by a gradual increase
in the cytoplasm, indicating the diffusion of Mg2+ from the
mitochondrial matrix to the cytoplasm. In addition, the
combined use of KMG-104-AsH with mKeima enables to
eliminate the effects of movements of cells and organelles by
calculating the changes in relative fluorescence intensity of
KMG-104-AsH in reference to that of mKeima. Moreover, the
fluorescence ratio of KMG-104-AsH to mKeima correlated with
the intracellular Mg2+ concentration (Figure S7, Supporting
Information). Using this method, it was estimated that FCCP
induced an increase in Mg2+ concentration in the cytoplasm in
HeLa cells from 0.5 mM to 1.1 mM (Figure S9, Supporting
Information). These values are consistent with the Mg2+
concentration ranges evaluated with other methods,40 indicat-
ing that KMG-104-AsH allows quantitative measurements.
These are advantages of this method over KMG-104 and other
nonratiometric probes.
ASSOCIATED CONTENT
* Supporting Information
Supproting figures. This material is available free of charge via
■
S
the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
oka@bio.keio.ac.jp
■
Author Contributions
⊥These authors contributed equally to this work.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This research was supported by a Strategic Research
Foundation Grant-aided Project for Private Universities from
the Ministry of Education, Culture, Sport, Science, and
Technology, Japan (MEXT), 2008−2012 (No. S0801008),
and a Grant-in-Aid for Scientific Research, KAKENHI (Nos.
24240045 and 25750395).
REFERENCES
■
(1) Matsuda, H.; Saigusa, A.; Irisawa, H. Nature 1987, 325, 156−9.
(2) Romani, A. M.; Scarpa, A. Front Biosci. 2000, 5, D720−34.
(3) Politi, H. C.; Preston, R. R. Neuroreport 2003, 14, 659−68.
(4) Gwanyanya, A.; Amuzescu, B.; Zakharov, S. I.; Macianskiene, R.;
Sipido, K. R.; Bolotina, V. M.; Vereecke, J.; Mubagwa, K. J. Physiol.
2004, 559, 761−76.
(5) Rodriguez-Zavala, J. S.; Moreno-Sanchez, R. J. Biol. Chem. 1998,
273, 7850−5.
(6) Boska, M. D.; Welch, K. M.; Barker, P. B.; Nelson, J. A.; Schultz,
L. Neurology 2002, 58, 1227−33.
In this work, the TCtag-selective Mg2+ fluorescent probe,
KMG-104-AsH, was synthesized and applied to Mg2+ imaging
in local regions in HeLa cells. Changes in Mg2+ concentration
around mitochondria have been associated with apoptosis.41 It
indicates the possibility that imaging of Mg2+ concentration in
the local region around mitochondria reveals the contribution
of Mg2+ as an intracellular signal to mitochondria. Moreover, a
recent study revealed that Mg2+ influx via the selective Mg2+
channels MagT1 and TUSC3 is required for embryonic
development in zebrafish14 and that via MagT1 also acts as a
second messenger in T cells.42 These findings suggest that
changes in Mg2+ concentration around organelles and specific
proteins may play important roles in cellular processes. The
bifunctional KMG-104-AsH probe developed in this work is
expected to be a strong tool for the elucidation of the little-
known molecular mechanisms and critical roles of Mg2+ in cells.
(7) Yasui, M.; Kihira, T.; Ota, K. Neurotoxicology 1992, 13, 593−600.
(8) Keenan, D.; Romani, A.; Scarpa, A. FEBS Lett. 1996, 395, 241−4.
(9) Touyz, R.; Mercure, C.; Reudelhuber, T. Mol. Aspects Med. 2003,
24, 107−136.
(10) Sontia, B.; Touyz, R. M. Arch. Biochem. Biophys. 2007, 458, 33−
9.
(11) Hashimoto, T.; Nishi, K.; Nagasao, J.; Tsuji, S.; Oyanagi, K.
Brain Res. 2008, 1197, 143−51.
(12) Trapani, V.; Farruggia, G.; Marraccini, C.; Iotti, S.; Cittadini, A.;
Wolf, F. I. Analyst 2010, 135, 1855−66.
(13) Komatsu, H.; Iwasawa, N.; Citterio, D.; Suzuki, Y.; Kubota, T.;
Tokuno, K.; Kitamura, Y.; Oka, K.; Suzuki, K. J. Am. Chem. Soc. 2004,
126, 16353−60.
2380
dx.doi.org/10.1021/ja410031n | J. Am. Chem. Soc. 2014, 136, 2374−2381

Journal of the American Chemical Society
Article
(14) Zhou, H.; Clapham, D. E. Proc. Natl. Acad. Sci. U.S.A. 2009, 106,
15750−5.
(15) Jin, J.; Desai, B. N.; Navarro, B.; Donovan, A.; Andrews, N. C.;
Clapham, D. E. Science 2008, 322, 756−60.
(16) Griffin, B. A.; Adams, S. R.; Tsien, R. Y. Science 1998, 281, 269−
72.
(17) Adams, S. R.; Campbell, R. E.; Gross, L. A.; Martin, B. R.;
Walkup, G. K.; Yao, Y.; Llopis, J.; Tsien, R. Y. J. Am. Chem. Soc. 2002,
124, 6063−76.
(18) Bacci, J. P.; Kearney, A. M.; Van Vranken, D. L. J. Org. Chem.
2005, 70, 9051−3.
(19) Adams, S. R.; Tsien, R. Y. Nat. Protoc. 2008, 3, 1527−34.
(20) Otten, P. A.; London, R. E.; Levy, L. A. Bioconjugate Chem.
2001, 12, 203−12.
(21) Suzuki, Y.; Komatsu, H.; Ikeda, T.; Saito, N.; Araki, S.; Citterio,
D.; Hisamoto, D.; Kitamura, Y.; Kubota, T.; Nakagawa, J.; Oka, K.;
Suzuki, K. Anal. Chem. 2002, 74, 1423−8.
(22) de Silva, A. P.; Gunaratne, H. Q.; Gunnlaugsson, T.; Huxley, A.
J.; McCoy, C. P.; Rademacher, J. T.; Rice, T. E. Chem. Rev. 1997, 97,
1515−1566.
(23) Nagano, T.; Yoshimura, T. Chem. Rev. 2002, 102, 1235−70.
(24) Hirano, T.; Kikuchi, K.; Urano, Y.; Higuchi, T.; Nagano, T.
Angew. Chem., Int. Ed. 2000, 39, 1052−1054.
(25) Urano, Y.; Kamiya, M.; Kanda, K.; Ueno, T.; Hirose, K.;
Nagano, T. J. Am. Chem. Soc. 2005, 127, 4888−94.
(26) Martin, B. R.; Giepmans, B. N.; Adams, S. R.; Tsien, R. Y. Nat.
Biotechnol. 2005, 23, 1308−14.
(27) Mottram, L. F.; Maddox, E.; Schwab, M.; Beaufils, F.; Peterson,
B. R. Org. Lett. 2007, 9, 3741−4.
(28) Areephong, J.; Ruangsupapichart, N.; Thongpanchang, T.
Tetrahedron Lett. 2004, 45, 3067−3070.
(29) Tour, O.; Adams, S. R.; Kerr, R. A.; Meijer, R. M.; Sejnowski, T.
J.; Tsien, R. W.; Tsien, R. Y. Nat. Chem. Biol. 2007, 3, 423−31.
(30) Spagnuolo, C. C.; Vermeij, R. J.; Jares-Erijman, E. A. J. Am.
Chem. Soc. 2006, 128, 12040−1.
(31) Ozawa, T.; Natori, Y.; Sako, Y.; Kuroiwa, H.; Kuroiwa, T.;
Umezawa, Y. ACS Chem. Biol. 2007, 2, 176−86.
(32) Brocard, J. B.; Rajdev, S.; Reynolds, I. J. Neuron 1993, 11, 751−
7.
(33) Leyssens, A.; Nowicky, A. V.; Patterson, L.; Crompton, M.;
Duchen, M. R. J. Physiol. 1996, 496, 111−28.
(34) Shindo, Y.; Fujii, T.; Komatsu, H.; Citterio, D.; Hotta, K.;
Suzuki, K.; Oka, K. PLoS One 2011, 6, e23684.
(35) Kubota, T.; Shindo, Y.; Tokuno, K.; Komatsu, H.; Ogawa, H.;
Kudo, S.; Kitamura, Y.; Suzuki, K.; Oka, K. Biochim. Biophys. Acta
2005, 1744, 19−28.
(36) Shindo, Y.; Fujimoto, A.; Hotta, K.; Suzuki, K.; Oka, K. J.
Neurosci. Res. 2010, 88, 3125−32.
(37) Yamanaka, R.; Shindo, Y.; Hotta, K.; Suzuki, K.; Oka, K. FEBS
Lett. 2013, 587, 2643−8.
(38) Dong, X.; Han, J. H.; Heo, C. H.; Kim, H. M.; Liu, Z.; Cho, B.
R. Anal. Chem. 2012, 84, 8110−3.
(39) Kim, H. M.; Jung, C.; Kim, B. R.; Jung, S. Y.; Hong, J. H.; Ko, Y.
G.; Lee, K. J.; Cho, B. R. Angew. Chem., Int. Ed. 2007, 46, 3460−3.
(40) Romani, A. M. Arch. Biochem. Biophys. 2011, 512, 1−23.
(41) Wolf, F. I.; Trapani, V. Clin. Sci. (London) 2008, 114, 27−35.
(42) Li, F. Y.; Chaigne-Delalande, B.; Kanellopoulou, C.; Davis, J. C.;
Matthews, H. F.; Douek, D. C.; Cohen, J. I.; Uzel, G.; Su, H. C.;
Lenardo, M. J. Nature 2011, 475, 471−6.
2381
dx.doi.org/10.1021/ja410031n | J. Am. Chem. Soc. 2014, 136, 2374−2381