Design of coordination interaction of Zn(II) complex with oligo-aspartate peptide to afford a high-affinity tag-probe pair

Article abstract of DOI:10.1246/bcsj.20150037

A complementary recognition pair consisting of a genetically encodable peptide tag and a small molecular probe isa powerful tool to specifically label and manipulate a protein ofinterest under biological conditions. In this study, we report the redesign of a tag-probe pair comprising an oligo-aspartate peptide tag (such as DDDD) and a binuclear zinc complex. Isothermal-titration calorimetry screening of binding between the series of peptides and zinc complexes revealed that the binding affinity was largely influenced by subtle changes of the ligand structure of the probe. However, the binding was tolerant to differences of the tag peptide sequence. Of those tested, a pair containing a peptide tag (DDAADD) and a binuclear zinc complex possessing 4-chloropyridines (3-2Zn(II)) showed the strongest binding affinity (Ka = 3.88 × 10⁵ M^-1), which was about 10-fold larger than the conventional pair of D4-peptide tag (DDDD) and 1-2Zn(II) containing nonsubstituted pyridines (K_a = 3.73 × 10⁴ M^-1). The strong binding of this new complementary recognition pair enabled the rapid covalent labeling of a tag-fused maltose binding protein with a fluorescent zinc complex, demonstrating its potential utility in protein analysis.

Full text of DOI:10.1246/bcsj.20150037

Selected Paper
Design of Coordination Interaction of Zn(II) Complex with
Oligo-Aspartate Peptide to Afford a High-Affinity Tag-Probe Pair
Hirokazu Fuchida,1 Shigekazu Tabata,1 Naoya Shindo,1 Ippei Takashima,1
Qiao Leng,1 Yuji Hatsuyama,1 Itaru Hamachi,2 and Akio Ojida*1
1Graduate School of Pharmaceutical Sciences, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812-8582
2Department of Synthetic Chemistry and Biological Chemistry, Graduate School of Engineering, Kyoto University,
Katsura, Nishikyo-ku, Kyoto 615-8510
E-mail: ojida@phar.kyushu-u.ac.jp
Received: January 27, 2015; Accepted: February 25, 2015; Web Released: March 4, 2015
A complementary recognition pair consisting of a genetically encodable peptide tag and a small molecular probe is a
powerful tool to specifically label and manipulate a protein of interest under biological conditions. In this study, we report
the redesign of a tag-probe pair comprising an oligo-aspartate peptide tag (such as DDDD) and a binuclear zinc complex.
Isothermal-titration calorimetry screening of binding between the series of peptides and zinc complexes revealed that the
binding affinity was largely influenced by subtle changes of the ligand structure of the probe. However, the binding was
tolerant to differences of the tag peptide sequence. Of those tested, a pair containing a peptide tag (DDAADD) and a
binuclear zinc complex possessing 4-chloropyridines (3-2Zn(II)) showed the strongest binding affinity (Ka = 3.88 ©
105 M¹1), which was about 10-fold larger than the conventional pair of D4-peptide tag (DDDD) and 1-2Zn(II) containing
nonsubstituted pyridines (Ka = 3.73 © 104 M¹1). The strong binding of this new complementary recognition pair enabled
the rapid covalent labeling of a tag-fused maltose binding protein with a fluorescent zinc complex, demonstrating its
potential utility in protein analysis.
A complementary recognition pair of a short epitope peptide
and a small molecular probe is a versatile analytical tool for a
wide variety of protein research. To date, several peptide tag-
probe pairs, such as oligo-histidine tag (His-tag) with Ni(II)-
NTA (nitrilotriacetic acid)1 and tetracysteine motif (-Cys-Cys-
X-X-Cys-Cys-) with bisarsenical ligand (FlAsH),2 have been
reported. These pairs are widely used for protein purification
using affinity column chromatography, protein/peptide immo-
bilization on a microtiter plate, and bioimaging of proteins of
interest (POI) under live cell conditions. In contrast to pro-
tein labeling techniques using fluorescent proteins (FPs)3 and
enzyme reactions such as SNAP-tag (a protein tag based on
human O6-alkylguanine-DNA alkyltransferase),4 the use of a
genetically encodable peptide tag/small molecular probe pair is
a flexible approach for a wide variety of POIs. The pair shows
little or no interference with the function of the protein, because
its size (1-3 kDa) is smaller than FP (27 kDa) or SNAP-tag
(20 kDa). However, the availability of such a useful comple-
mentary tag-probe pair is limited, mainly because of the diffi-
culty of rationally designing a recognition pair, where the
careful optimization of a probe structure and peptide tag
sequence is required to achieve a desirable strong affinity and
selective interaction.
oligo-aspartate peptide tag.5 Based on the coordination inter-
action between the zinc complex and the aspartate residues of
the peptide, the DpaTyr-Zn(II) probe showed a strong binding
affinity (Ka = 6.9 © 105 M¹1) to D4-tag (Asp-Asp-Asp-Asp) in
neutral aqueous conditions (50 mM HEPES, pH 7.2). However,
this binding affinity decreased to 4.9 © 104 M¹1 under high salt
concentrations (50 mM HEPES, 100 mM NaCl, pH 7.2) that
are relevant to biological environments.6 Despite this weak
binding, there has been little effort to improve the binding
affinity by optimizing the probe structure and the peptide tag
sequence. In this study, we report the redesign of the coor-
dination interaction between the zinc complex and oligo-
aspartate peptide. We found that modification of the ligand
structure of the zinc complex largely influenced the binding
affinity with tetra-aspartate aspartate peptides. Among the
evaluated tag-probe pairs, 3-2Zn(II) containing 4-chloro-
pyridines displayed the strongest binding affinity toward a
peptide with a DDAADD sequence (Ka = 3.88 © 105 M¹1).
This strong binding pair was successfully used in the rapid
covalent labeling of a tag-fused maltose binding protein (MBP)
protein with a fluorescent zinc complex.
Results and Discussion
We recently developed a binuclear Dpa (2,2¤-dipicolyl-
amine)-Zn(II) complex possessing an L-tyrosine scaffold
(DpaTyr-Zn(II)) (Figure 1) as a selective binding probe for an
Design of the Binuclear Zn(II) Complex. We initially
synthesized a series of binuclear Zn(II) complexes based on a
p-cresol scaffold, each of which possessed four different
784 | Bull. Chem. Soc. Jpn. 2015, 88, 784–791 | doi:10.1246/bcsj.20150037
© 2015 The Chemical Society of Japan

R
R
/min
N
N
R
R
N
N
N
N
N ZnII
N
O ZnII
ZnII ZnII
N
N
O
N
N
3Cl-
3Cl-
Me
OMe
BocHN
O
1-2Zn(II) : R = H
2-2Zn(II) : R = OMe
3-2Zn(II) : R = Cl
DpaTyr-Zn(II)
R
R
N
S
S
N
R
R
N
S
N
S
N
N
N
N
N
N
N
N
ZnII ZnII
ZnII ZnII
N
N
N
O
N
O
3Cl-
3Cl-
Me
Me
4-2Zn(II) : R = H
5-2Zn(II) : R = SO2N(CH3)2
6-2Zn(II) : R = Me
Figure 1. Molecular structures of the binuclear zinc
complexes.
Figure 2. ITC titration curve (upper) and processed data
(lower) of 1-2Zn(II) with DDAADD peptide (Ac-
YADDAADD-NH2). [1-2Zn(II)] = 50 ¯M, [DDAADD] =
2 mM (1.5 ¯L © 25 injections), 50 mM HEPES, 100 mM
NaCl, pH 7.2, 25 °C.
heteroaromatic rings as the zinc coordination units. To evaluate
the substitution effect of the pyridine rings, we prepared three
Dpa (2,2-dipicolylamine)-type zinc complexes 1-2Zn(II), 2-
2Zn(II), and 3-2Zn(II), each of which contained nonsubsti-
tuted, 4-chloro-, and 4-methoxypyridine rings, respectively.
It was expected that a binuclear zinc complex comprising
five-membered heteroaromatic rings could also function as a
binding probe for the oligo-aspartate tag. Thus, we prepared a
thiazole-type complex 4-2Zn(II) possessing 4-thiazoles as zinc
coordination units. We also prepared imidazole derivatives 5-
2Zn(II) and 6-2Zn(II), which contained 1-methylimidazoles
and 1-(N,N-dimethylsulfamoyl)imidazoles as the coordination
units, respectively. The syntheses and characterization of these
zinc complexes are described in Supporting Information
(Schemes S1-S6).
Isothermal-Titration Calorimetry Screening of the Pep-
tide Tag-Probe Pair. To screen the binding of the tag-probe
pair, we initially evaluated the binding affinities of the binu-
clear Zn(II) complex 1-2Zn(II) with a series of oligo-aspartate
peptide tags (DDDD, DDADD, DDAADD, and DAAAD) by
isothermal-titration calorimetry (ITC). Figure 2 shows a typi-
cal ITC titration curve between DDAADD and 1-2Zn(II), and
Table 1 summarizes the results of the ITC experiments. The
binding affinity of the conventional pair between 1-2Zn(II)
possessing nonsubstituted pyridines and DDDD (D4-tag) was
3.73 © 104 M¹1, which was comparable to that of the origi-
nal DpaTyr-Zn(II)/D4-tag pair. Interestingly, 1-2Zn(II) could
bind to the alanine-inserted tetra-aspartate tags (DDADD and
DDAADD) with moderate binding affinities (Ka = 3.18 © 104
and 4.23 © 104 M¹1, respectively), which were comparable to
that of DDDD. Furthermore, 1-2Zn(II) displayed weak binding
to the di-aspartate peptide DAAAD (Ka = 3.18 © 103 M¹1),
which contains two aspartate residues at the i, i + 4 position.
The binding process of 1-2Zn(II) with the peptides was exo-
thermic enthalpy driven (¦H < 0) in all cases, but its binding
with D4 and DDADD was also driven by the positive entropy
change (¦S > 0). This result implies that the bindings of the
latter two cases involve the release of hydration waters, which
favors the increased binding affinities of 1-2Zn(II) with the
peptides.7
The binding affinities of the other binuclear zinc complexes
containing oligo-aspartate peptides were further evaluated by
ITC (Table 1). The binding affinities of 2-2Zn(II) possessing 4-
methoxypyridines were comparable to those of 1-2Zn(II) for all
cases tested. However, 3-2Zn(II) possessing 4-chloropyridines
exhibited a stronger binding affinity than 1-2Zn(II). In par-
ticular, the binding affinity with DDAADD was 3.88 ©
105 M¹1, about 10-fold larger than the conventional binding
pair of 1-2Zn(II) and D4 (Ka = 3.73 © 104 M¹1). The thermo-
dynamic data suggest that the large negative enthalpy change
(¦H = ¹7.0 kcal mol¹1) of the 3-2Zn(II)/DDAADD pair con-
tributes to their stronger binding affinity compared with the 1-
2Zn(II)/D4 pair (¦H = ¹3.35 kcal mol¹1). A comparison of
the thermodynamic data of the bindings of the different probes
with DDAADD, 2-2Zn(II) (¦H = ¹6.54 kca mol¹1, T¦S =
¹0.38 kcal mol¹1) showed similarities with 1-2Zn(II) (¦H =
¹6.53 kcal mol¹1, T¦S = ¹0.22 kcal mol¹1), while the binding
of 3-2Zn(II) involved a favorable positive entropy change
(T¦S = 0.66 kcal mol¹1) as well as a slightly larger negative
Bull. Chem. Soc. Jpn. 2015, 88, 784–791 | doi:10.1246/bcsj.20150037
© 2015 The Chemical Society of Japan | 785

Table 1. Summary of ITC Screening to Evaluate the Binding between Zinc Complexes and Oligo-Aspartate Peptidesa),b)
DDDDc) (D4-tag)
DDADDd)
DDAADDe)
DAAADf)
1-2Zn(II)
2-2Zn(II)
3-2Zn(II)
4-2Zn(II)
5-2Zn(II)
6-2Zn(II)
Ka
(3.73 « 0.04) © 104
¹3.35 « 0.21
2.69
0.82 « 0.04
(2.98 « 0.44) © 104
¹2.46 « 0.34
3.64
0.85 « 0.10
(2.08 « 0.22) © 105
¹1.34 « 0.04
5.90
0.76 « 0.02
(5.44 « 0.31) © 104
¹2.67 « 0.05
3.78
(3.18 « 0.31) © 104
¹2.56 « 0.21
3.58
0.95 « 0.06
(2.59 « 0.31) © 104
¹3.50 « 0.50
2.52
0.75 « 0.09
(1.23 « 0.13) © 105
¹1.56 « 0.07
5.39
0.64 « 0.02
(3.64 « 0.17) © 104
¹3.84 « 0.08
2.38
(4.23 « 0.19) © 104
¹6.53 « 0.21
¹0.22
0.87 « 0.02
(3.26 « 0.22) © 104
¹6.54 « 0.48
¹0.38
0.72 « 0.04
(3.88 « 0.37) © 105
¹7.00 « 0.12
0.66
0.78 « 0.01
(7.56 « 0.23) © 104
¹8.43 « 0.07
¹1.77
(3.18 « 0.10) © 103
¹5.47 « 0.15
¹0.69
1.03 « 0.02
(3.28 « 0.16) © 103
¹2.10 « 0.11
2.70
0.98 « 0.04
(2.99 « 0.19) © 103
¹3.80 « 0.13
2.31
0.86 « 0.02
(8.12 « 0.11) © 103
¹6.49 « 0.04
¹1.15
¦H
T¦S
n
Ka
¦H
T¦S
n
Ka
¦H
T¦S
n
Ka
¦H
T¦S
n
0.88 « 0.01
(2.71 « 0.56) © 104
¹1.04 « 0.17
5.01
0.93 « 0.13
<103
0.87 « 0.02
(2.06 « 0.32) © 104
¹1.37 « 0.21
4.50
0.93 « 0.12
<103
0.92 « 0.01
(2.36 « 0.21) © 104
¹4.25 « 0.5
1.72
0.85 « 0.08
<103
0.96 « 0.00
(1.76 « 0.11) © 103
¹4.33 « 0.44
0.10
Ka
¦H
T¦S
n
0.90 « 0.08
g)
Ka
®
¦H
T¦S
n
a) n: stoichiometry, Ka: binding constant (M¹1), ¦H: enthalpy (kcal mol¹1), T¦S: entropy (kcal mol¹1). b) Measurement
conditions: 50 mM HEPES, 100 mM NaCl, pH 7.2, 25 °C. c) D4 peptide: Boc-(Asp)4-NH2. d) DDADD peptide: Ac-
YADDADD-NH2. e) DDAADD peptide: Ac-YADDAADD-NH2. f) DAAAD peptide: Ac-DAAAD-NH2. g) Not
measured.
enthalpy change (¦H = ¹7.00 kcal mol¹1) compared with 1-
2Zn(II), both of which contribute to the strong binding affinity
of 3-2Zn(II) with DDAADD. Thus, a subtle structural change
of the probe markedly influenced the binding affinity of the
tag-probe pair by perturbation of the binding thermodynamics.
This suggests that a delicate structural design is required to
achieve a tighter binding with the probe that operates under
aqueous conditions.8
In the series of zinc complexes possessing the five-membered
heteroaromatic rings, the thiazole-type complex 4-2Zn(II)
showed a slightly higher binding affinity than 1-2Zn(II). The
binding affinities of 5-2Zn(II) possessing 1-(N,N-dimethyl-
sulfamoyl)imidazoles were slightly weaker than 1-2Zn(II). The
binding processes of these zinc complexes are exothermic
enthalpy driven (¦H < 0) in all cases, as also observed for
the pyridine-type probes. Interestingly, 6-2Zn(II) possessing N-
methylimidazoles showed a very weak binding affinity with all
of the tetra-aspartate peptides (Ka < 103 M¹1). The weak bind-
ing of 6-2Zn(II) cannot be explained by the ITC result because
of the lack of thermodynamic data. However, we speculated that
the weak binding of 6-2Zn(II) might be attributable to the
stronger basicity of the N-methylimidazole rings compared with
5-2Zn(II), which contains the more electron-deficient 1-(N,N-
dimethylsulfamoyl)imidazoles. The calculated pKa value of the
conjugated acid of 1-methylimidazole is 7.01, which is signifi-
cantly larger than that of 1-(N,N-dimethylsulfamoyl)imidazole
(pKa = 2.41).9 Thus, the strong basicity of 1-methylimidazole
might reduce the Lewis acidity of the coordinated zinc ions of
6-2Zn(II), resulting in its weak binding with the tetra-aspartate
peptide.
Fluorescence Evaluation of the Binding of the Peptide
Tag-Probe Pair. Binding of the zinc complexes with the
oligo-aspartate peptide was evaluated by fluorescence titration
to confirm the results of the ITC screening. For fluorescence
titration, we prepared a fluorescent DDAADD peptide (hc-
KDDAADD) labeled with 7-hydroxycoumarin at the ε-amino
group of the lysine reside (Figure 3a). When 3-2Zn(II) was
added to the aqueous solution of hc-KDDAADD (5 ¯M in 50
mM HEPES, 100 mM NaCl, pH 7.2), the fluorescence of the 7-
hydroxycoumarine decreased to approximately 20% of its origi-
nal intensity (Figure 3b). This fluorescence change might be
due to the heavy metal ion effect of zinc ions and/or the micro-
pH change induced by the probe binding. The change of fluo-
rescence intensity at 446 nm evaluated by curve fitting analysis
was 5.1 © 105 M¹1, which is comparable to that observed in the
ITC experiment (Ka = 3.88 © 105 M¹1) (Figure 3c). The bind-
ing stoichiometry between 3-2Zn(II) and hc-DDAADD was
evaluated by fluorescence Job’s plot, which clearly indicates
that their binding stoichiometry is 1:1 (Figure S1). Fluorescence
titration was performed with the other zinc complexes and the
results are summarized in Table 2. The binding affinity of 1-
2Zn(II) was 5.87 © 104 M¹1, approximately 10-fold lower than
that of 3-Zn(II), confirming the affinity-enhancing effect of the
chloride groups on its pyridine rings. The binding affinity of the
thiazole-type complex 4-2Zn(II) was 1.18 © 105 M¹1, a slightly
higher value than that of 1-2Zn(II). Regarding the imidazole-
786 | Bull. Chem. Soc. Jpn. 2015, 88, 784–791 | doi:10.1246/bcsj.20150037
© 2015 The Chemical Society of Japan

CD Conformation Analysis of Peptide Binding with the
Probe. It is reasonable to assume that the zinc complexes
induced a folded peptide structure in the binding complex,
especially in the case of longer peptides such as DDAADD and
DAAAD, both of which contain aspartate residues that are
involved in binding at the distal positions. To clarify this, we
initially performed circular dichroism (CD) measurements of
the DAAAD peptide possessing the two aspartate residues at the
i, i + 4 position (Figure 4). DAAAD showed a negative Cotton
peak around 200 nm in the absence of 3-2Zn(II), indicative of
the nonfolded random structure in its free form. In the binding
complex of DAAAD with 3-2Zn(II), the negative Cotton peak
around 200 nm disappeared and the strong induced CD (i-CD)
peaks appeared in the wide wavelength region from 350 to
190 nm. The large i-CD observed at over 300 nm (ªmax = 310
nm), which corresponds to the absorption region of the zinc-
phenolate moiety, suggested that a chiral environment was
introduced at the coordination sphere of 3-2Zn(II) upon binding
to the peptide. However, these Cotton peaks did not involve the
representative CD patterns of the folded peptide such as α-helix
and β-turn. Almost the same trend was observed in the CD
measurement of the binding complexes of 3-2Zn(II) with tetra-
aspartate peptides such as DDDD, DDADD, and DDAADD
(Figure S2). These peptides lost their nonfolded random struc-
tures and showed unique i-CD patterns distinct from each other
in the binding complexes with 3-2Zn(II). Although these CD
patterns might reflect the folded peptide structures, they did not
involve the typical Cotton peaks of α-helix or β-turn. Thus, the
CD analysis was unable to suggest the conformation of the
folded peptides binding with the zinc complex.
a
b
hc-KDDAADD : Ac-K( -hc)-ADDAADD-NH2
O
N
hc =
H
HO
O
O
600
400
200
0
400
450
500
550
600
wavelength/nm
c
1.0
0.75
0.5
Covalent Labeling of the Tag-Fused Protein with the
Zinc Complex.
The utility of the strong binding affinity
¹1
of DDAADD-tag/3-2Zn(II) probe pair (Ka = 3.88 © 105 M
determined by ITC experiment) was assessed by the binding-
induced covalent labeling of the tag-fused protein.10 For this
purpose, we prepared MBP tethered to a cysteine-containing
CA6DDAADD-tag (CAAAAAADDAADD) and a fluorescent
Dpa-Zn(II) complex possessing a reactive chloroacetyl group.11
It was expected that their complementary strong binding would
enhance the nucleophilic reaction between the reactive tag and
the probe to produce the covalently labeled MBP (Figure 5a).
The synthesis of the reactive zinc complex 7-2Zn(II), which
possesses a fluorescent coumarin unit, is described in Scheme 1.
Compound 9 was conjugated with Boc-Lys-OH to give 10,
which was deprotected by TFA and then conjugated with α-
chloroacetyl O-succinimidyl ester to yield the reactive fluores-
cent unit 11. The ligand unit was synthesized from the bis(o-
chloromethyl)cresol 1212 and the dipicolylamine 13 to give 14,
which was deprotected by treatment with hydrazine to yield
15. Finally, both units were conjugated and the product was
purified by HPLC to give the pure ligand 7. When 7-2Zn(II)
(5 ¯M) (Figure 5c), which was prepared by the treatment of 7
with 2 equiv of ZnCl2, was added to an aqueous solution of
the CA6DDAADD-tag-fused MBP (1 ¯M in 50 mM HEPES,
100 mM NaCl, pH 7.2), rapid labeling occurred. In-gel fluores-
cence analysis showed that the fluorescent band corresponding
to MBP labeled with 7-2Zn(II) gradually increased within 30
min (Figures 5b and 5d). MALDI-TOF mass analysis identified
a new peak corresponding to the covalent adduct of the tag-
0.25
0
0
5
10
15
20
3-2Zn(II)/µM
Figure 3. Fluorescence titration of the 7-hydroxycoumarin-
appended oligo-aspartate peptide (hc-KDDAADD) with
zinc complexes. (a) Structure of the hc-KDDAADD
peptide. (b) Fluorescence spectral change of the hc-
KDDAADD peptide (5 ¯M) upon addition of 3-2Zn(II)
(0-20 ¯M). (c) Curve-fitting analysis of the fluorescence
emission change at 446 nm. The error bars represent
standard deviations obtained by the triplicate experiments.
Measurement conditions: 50 mM HEPES, 100 mM NaCl,
pH 7.2, 25 °C, -ex = 400 nm.
type complexes, 5-2Zn(II) showed a slightly weaker binding
affinity (Ka = 2.76 © 105 M¹1) than 1-2Zn(II), whereas the
binding affinity of 6-2Zn(II) was very weak (Ka < 103 M
)
¹1
under the aqueous conditions. The order of the binding affinities
of the zinc complexes shown in Table 2 is the same as that ob-
tained by the ITC experiments (Table 1), suggesting the validity
of the present binding data obtained by the two distinct methods.
Bull. Chem. Soc. Jpn. 2015, 88, 784–791 | doi:10.1246/bcsj.20150037
© 2015 The Chemical Society of Japan | 787

Table 2. Summary of the Binding Constants (Ka, M¹1) of the Zinc Complexes with the hc-KDDAADD Peptide as
Determined by Fluorescence Titrationa),b)
1-2Zn(II)
2-2Zn(II)
3-2Zn(II)
4-2Zn(II)
5-2Zn(II)
6-2Zn(II)
¹1
Ka/M
5.87 © 104
(4.23 © 104)
6.25 © 104
(3.26 © 104)
5.14 © 105
(3.88 © 105)
1.18 © 105
(7.56 © 104)
2.76 © 104
(2.36 © 104)
<103
(<103)
a) Measurement conditions: 50 mM HEPES, 100 mM NaCl, pH 7.2, 25 °C. b) The binding constants in parentheses are
the values determined in the ITC experiment (Table 1).
sion that further optimization of the probe structure and peptide
sequence will provide a tag-probe pair with a better binding
affinity. The present data reveal that a subtle difference in the
probe structure resulted in a large difference in peptide binding
affinity. This result suggests there is scope to develop a new
tag-probe pair with a strong binding affinity suitable for protein
analysis and manipulation under crude biological conditions.
3
2
1
0
Experimental
Synthesis of the Zinc Complexes. The zinc complexes 1-
5-2Zn(II) were prepared by treatment of the ligand 1-5 with an
aqueous solution of ZnCl2 (2 equiv) in MeOH at rt for 30 min.
After removal of the solvent by evaporation, the residue was
dissolved in distilled H2O, and the solution was filtered through
cellulose acetate filter and lyophilized. The solid was collected
by filtration, washed with AcOEt, and dried in vacuo to give
a powder. In the case of 6-2Zn(II), the ligand 6 was difficult
to handle in air due to its hygroscopic property. Thus, the zinc
complex was prepared by the treatment of the aqueous solution
of 6 with 2 equiv of ZnCl2, in which the concentration of 6 was
determined by UV titration (-abs = 295 nm) with ZnCl2 in the
solution (50:50, v/v) of CH3CN and 50 mM HEPES, 100 mM
NaCl, pH 7.2.
1-2Zn(II) (88%) Anal Calc for C33H34N6O¢2Zn¢3Cl¢HCl¢
2H2O: C, 47.23; H, 4.56; N, 10.01%. Found: C, 47.26; H, 4.22;
N, 10.01%. ESI-MS m/z 751.13 [C33H33ClN6OZn2 + 2OH +
Na]+.
2-2Zn(II) (94%) Anal Calc for C37H42N6O5¢2Zn¢3Cl¢HCl¢
H2O: C, 47.21; H, 4.71; N, 8.93%. Found: C, 47.20; H, 4.48;
N, 8.99%. ESI-MS m/z 871.16 [C37H41ClN6O5Zn2 + 2OH +
Na]+.
3-2Zn(II) (92%) Anal Calc for C33H30Cl4N6O¢2Zn¢3Cl¢
HCl¢H2O: C, 41.33; H, 3.36; N, 8.76%. Found: C, 41.64; H,
3.33; N, 8.34%. ESI-MS m/z 888.96 [C33H29Cl5N6OZn2 +
2OH + Na]+.
4-2Zn(II) (94%) Anal Calc for C25H26N6OS4¢2Zn¢3Cl¢
HCl¢H2O: C, 35.52; H, 3.34; N, 9.94%. Found: C, 35.40; H,
3.18; N, 9.90%. ESI-MS m/z 772.94 [C25H25ClN6OS4 +
2OH + Na]+.
5-2Zn(II) (86%) Anal Calc for C33H50N14O9S2¢2Zn¢3Cl¢
HCl¢H2O: C, 32.87; H, 4.35; N, 16.04%. Found: C, 32.87; H,
4.35; N, 15.92%. ESI-MS m/z 1135.08 [C33H49N14O9S4 +
2OH + Na]+.
-1
-2
-3
200
250
300
350
wavelength/nm
Figure 4. CD titration of DAAAD peptide (50 ¯M) with 3-
2Zn(II) (0, 25, 50, and 100 ¯M). Measurement conditions:
H2O, pH 7.4, 25 °C.
fused MBP with 7-2Zn(II) (Figure S3). However, none of the
fluorescent bands were detected by in-gel fluorescence analysis
when the labeling reaction was conducted in the presence of a
high concentration of inorganic pyrophosphate (PPi, 2.5 mM),
a strong binder of the binuclear zinc complex (Figure S2).5c,12
This result suggests that the selective binding between the tag
and the probe is critical to the covalent labeling of the tag-fused
MBP. Binding-induced MBP labeling was also conducted using
the conventional D4-tag/Dpa-Zn(II) probe pair (Figure 5b). In-
gel fluorescence analysis revealed that the labeling of CA6D4
(CAAAAAADDDD)-tag-fused MBP with 8-2Zn(II) was slow-
er than that of the new tag-probe pair; the initial reaction rate
(¦F, min¹1) of MBP labeling using the CA6D4-tag/8-2Zn(II)
probe pair was 0.59, which was 2.7-fold smaller than that using
CA6DDAADD-tag/7-2Zn(II) probe pair (¦F = 1.61 min¹1).13
This result demonstrates an advantage of the new tag-probe pair
over the previous one for binding-induced protein labeling, in
which the strong binding affinity of the new pair facilitates the
rapid and efficient labeling of the tag-fused protein under low
concentration conditions.
Conclusion
Synthesis of the Labeling Probe. Synthesis of 10:
A
We developed a new tag-probe pair consisting of a
DDAADD-tag and 3-2Zn(II) probe, which shows approxi-
mately 10-fold stronger binding affinity than the conventional
pair of DDDD and 1-2Zn(II). We now plan further application
of this tag-probe pair for fluorescence imaging in living cells
and affinity purification of a tag-fused protein. We also envi-
solution of Boc-Lys-OH (500 mg, 2.03 mmol), 9 (717 mg, 2.00
mmol), and DIEA (523 ¯L, 3.00 mmol) in dry DMF (15.0 mL)
was stirred at rt for 3.5 h. After removal of the solvent under
reduced pressure, the residue was diluted with water and
extracted with AcOEt. The extracts were washed with brine,
dried over MgSO4, and concentrated by evaporation. The
788 | Bull. Chem. Soc. Jpn. 2015, 88, 784–791 | doi:10.1246/bcsj.20150037
© 2015 The Chemical Society of Japan

a
b
c
7-2Zn(II) + CA6DDAADD-MBP
8-2Zn(II) + CA6D4-MBP
5
10 20 30 40 60 90
time (min)
0
5
10 20 30 40 60 90
0
FL
CBB
d
2.0
1.5
R
R
R
R
N
N
N
N
ZnII ZnII
N
O
O
N
3Cl-
H
N
1.0
0.5
0
Cl
N
H
O
O
O
N
HN
O
7-2Zn(II) : R = Cl
8-2Zn(II) : R = H
0
30
60
90
time/min
Figure 5. (a) Schematic illustration of the covalent labeling of the CA6DDAADD-tag-fused protein with the reactive binuclear zinc
complex. (b) In-gel fluorescence (FL) and CBB analyses of the covalent labeling of CA6DDAADD-tag-fused maltose binding
protein (MBP) with 7-2Zn(II) (left) and CA6D4-tag-fused MBP with 8-2Zn(II) (right). Labeling conditions: [tag-fused MBP] =
1 ¯M, [zinc complex] = 5 ¯M, 50 mM HEPES, 100 mM NaCl, pH 7.2, 18 °C. (c) Structures of the zinc complexes used for the
covalent labeling of tag-fused MBP. (e) Time trace plot of the labeling reactions shown in Figure 5b; CA6DDAADD-tag-fused
MBP with 7-2Zn(II) ( ), CA6D4-tag-fused MBP with 8-2Zn(II) ( ). The error bars represent standard deviations obtained by the
triplicate experiments.
residue was purified by flash column chromatography on SiO2
(CHCl3:MeOH = 20:1) to give 10 (558 mg, 56%) as a yellow
amorphous powder. H NMR (400 MHz, CDCl3): ¤ 8.92-8.89
(1H, s), 3.42-3.39 (6H, m), 2.02 (2H, brs), 1.59-1.44 (4H, m),
1.21-1.18 (6H, t, J = 7.0 Hz). ESI-MS m/z found 412.19 for
C20H27N3O5 [M + Na]+.
1
(1H, t, J = 5.6 Hz), 8.71 (1H, s), 7.42-7.40 (1H, d, J = 9.2 Hz),
6.64-6.61 (1H, dd, J = 2.0 Hz, 9.2 Hz), 6.48-6.47 (1H, d, J =
2.4 Hz), 5.31-5.29 (1H, d, J = 8.0 Hz), 4.27 (1H, brs), 3.46-
3.41 (6H, q, J = 7.2 Hz), 1.91-1.79 (2H, m), 1.66-1.62 (2H,
m), 1.46 (2H, m), 1.42 (9H, s), 1.24-1.20 (6H, t, J = 7.0 Hz).
ESI-MS m/z found 534.22 for C25H35N3O7 [M + 2Na]+.
Synthesis of 11: To a cooled (0 °C) solution of 10 (200 mg,
0.41 mmol) in dry CH2Cl2 (3 mL) was added dropwise TFA
(3 mL). The reaction mixture was stirred at rt for 2 h. After
removal of the solvent under reduced pressure, the reside was
dissolved in toluene and evaporated to give the deprotected
product (as mono TFA salt, 207 mg, quant.) as a yellow amor-
A solution of the deprotected product (100 mg, 0.21 mmol),
¡-chloroacetyl N-hydroxysuccinimidyl ester (50 mg, 0.26
mmol), and DIEA (89 ¯L, 0.51 mmol) in dry DMF (5 mL) was
stirred at rt for 1 h. After removal of the solvent under reduced
pressure, the residue was purified by flash column chromatog-
raphy on SiO2 to give 11 (88 mg, 91%) as a yellow solid.
1H NMR (400 MHz, CDCl3): ¤ 8.97 (1H, brs), 8.69 (1H, s),
7.43-7.41 (1H, d, J = 8.8 Hz), 6.64-6.62 (1H, d, J = 6.8 Hz),
6.47 (1H, s), 4.56 (1H, s), 4.04 (2H, s), 3.46-3.41 (6H, m),
1.98-1.86 (2H, m), 1.64 (2H, m), 1.45 (2H, m), 1.24-1.20 (6H,
t, J = 7.0 Hz). ESI-MS m/z found 488.16 for C22H28ClN3O6
[M + Na]+.
1
phous powder. H NMR (400 MHz, CDCl3): ¤ 9.14-9.12 (1H,
t, J = 2.6 Hz), 8.53 (1H, s), 7.98 (2H, brs), 7.38-7.36 (1H, d,
J = 8.8 Hz), 6.64-6.62 (1H, d, J = 9.2 Hz), 6.43 (1H, s), 4.08
Synthesis of 14: A solution of 12 (26 mg, 0.075 mmol),14
13 (48 mg, 0.18 mmol), KI (3.0 mg, 0.018 mmol), and K2CO3
(32 mg, 0.23 mmol) in dry DMF (2 mL) was stirred at rt over-
Bull. Chem. Soc. Jpn. 2015, 88, 784–791 | doi:10.1246/bcsj.20150037
© 2015 The Chemical Society of Japan | 789

O
O
H
H
N
N
Boc
OH
Cl
OH
O
O
O
O
1) TFA, CH2Cl2
Boc-Lys-OH
DIEA, DMF
N
O
O
N
O
O
O
N
O
2) -choroacetyl-O-
HN
N
O
HN
succinimidyl ester
(56%)
DIEA, DMF
O
O
(91%, 2 steps)
9
10
11
Cl
Cl
Cl
Cl
N
N
Cl OH Cl
N
N
Cl
N2H4-H2O
Cl
Cl
Cl
N
OH N
N
OH N
N
N
O
N
K2CO3, KI
N
H
+
THF
DMF
O
Cl
Cl
(52%)
(99%)
NH2
O
N
O
12
13
14
15
Cl
Cl
N
N
OH
N
N
Cl
Cl
N
O
N
•
1) WSC HCl, HOBt
DIEA, DMF
11
+
15
2) HPLC purification
(9%)
H
N
Cl
N
H
O
O
O
O
N
7
HN
Scheme 1. Synthesis of the labeling probe.
night. After removal of the solvent under reduced pressure,
the residue was dissolved in sat. NaHCO3 and extracted with
AcOEt. The extracts were dried over MgSO4 and concentrated
by evaporation. The residue was purified by flash column
chromatography on SiO2 (CHCl3:MeOH:NH3 = 500:10:1) to
give 14 (32 mg, 52%) as a colorless amorphous powder.
1H NMR (400 MHz, CDCl3): ¤ 8.39-8.37 (4H, d, J = 5.6 Hz),
7.81-7.79 (2H, m), 7.66-7.78 (2H, m), 7.50 (4H, s), 7.11-7.10
(4H, d, J = 5.2 Hz), 4.71 (2H, s), 3.83 (8H, s), 3.77 (4H, s).
ESI-MS m/z found 836.13 for C41H33Cl4N7O3 [M + Na]+.
Synthesis of 15: A solution of 14 (32 mg, 0.039 mmol) and
hydrazine monohydrate (145 ¯L) in THF (2.0 mL) was stirred
at rt overnight. After removal of the solvent by evaporation,
the residue was dissolved in sat. NaHCO3 and extracted with
CHCl3. The extracts were dried over MgSO4 and concentrated
by evaporation to give 15 (27 mg, 99%) as a yellow oil.
1H NMR (400 MHz, CDCl3): ¤ 8.42-8.29 (4H, m), 7.50-7.45
(4H, m), 7.13-7.09 (4H, m), 7.05 (2H, s), 3.86 (8H, s), 3.79
(4H, s), 3.73 (2H, s). ESI-MS m/z found 684.14 for
C33H31Cl4N7O6 [M + H]+.
NaHCO3, the mixture was extracted with AcOEt. The extracts
were dried over MgSO4 and concentrated by evaporation. The
residue was purified by HPLC (column; YMC-Actus Triart
C18, 20 © 250 mm) using CH3CN (0.1% TFA)/H2O (0.1%
TFA) solvent system with a linear gradient mode (CH3CN
(0.1% TFA)/H2O (0.1% TFA) = 50/50 ¼ 80/20 (0 ¼ 50
min)) to give 7 (as tetra TFA salt, 4.9 mg, 9%) as a yellow
solid. 1H NMR (400 MHz, CDCl3): ¤ 8.83 (1H, s), 8.65 (1H, s),
8.48-8.47 (4H, d, J = 4.8 Hz), 7.43 (4H, s), 7.42-7.39 (1H, d,
J = 9.2 Hz), 7.25 (4H, s), 7.14 (2H, s), 6.63-6.61 (1H, d,
J = 9.2 Hz), 6.47 (1H, s), 4.44 (1H, m), 4.21 (8H, s), 4.07 (4H,
s), 4.06 (2H, s), 4.05 (2H, s), 3.46-3.40 (6H, m), 1.94-1.80
(2H, m), 1.64-1.56 (2H, s), 1.44-1.40 (2H, m), 1.24-1.20
(6H, t, J = 7.2 Hz). ESI-TOF-MS m/z found 1131.31 for
C57H59Cl5N8O6 [M + H]+.
Isothermal Titration Calorimetry (ITC) Measurement.
ITC titration was performed with MicroCal Auto-iTC200 (GE
Healthcare). All measurements were conducted at 298 K. Typi-
cally, a solution of the peptide (2-7 mM) in a buffer solution
(50 mM HEPES, 100 mM NaCl, pH 7.2) was injected stepwise
(1.5 ¯L © 25 times) to a solution of the metal complex of
DpaTyr (50-300 ¯M) in the same solvent system. The mea-
sured heat flow was recorded as function of time and converted
into enthalpies (¦H) by integration of the appropriate reaction
Synthesis of 7: A solution of 11 (27 mg, 0.057 mmol), 15
(26 mg, 0.038 mmol), WSC¢HCl (22 mg, 0.11 mmol), HOBt¢
H2O (17 mg, 0.11 mmol), and DIEA (30 ¯L, 0.17 mmol) in dry
DMF (1 mL) was stirred at rt for 3 h. After dilution with sat.
790 | Bull. Chem. Soc. Jpn. 2015, 88, 784–791 | doi:10.1246/bcsj.20150037
© 2015 The Chemical Society of Japan

peaks. Dilution effects were corrected for by subtracting the
result of a blank experiment with an injection of the HEPES
solution into cell under identical experimental conditions. The
binding parameters (Kapp, ¦H, ¦S, n) were evaluated by apply-
ing one site model using the software Origin 7 (Origin Lab).
References
1
a) S. Uchinomiya, H. Nonaka, S. Wakayama, A. Ojida, I.
Hamachi, Chem. Commun. 2013, 49, 5022. b) M. Taki, F. Asahi, T.
Hirayama, Y. Yamamoto, Bull. Chem. Soc. Jpn. 2011, 84, 386.
c) M. Kamoto, N. Umezawa, N. Kato, T. Higuchi, Chem.®Eur. J.
2008, 14, 8004. d) C. T. Hauser, R. Y. Tsien, Proc. Natl. Acad. Sci.
U.S.A. 2007, 104, 3693. e) S. Lata, M. Gavutis, R. Tampé, J.
Piehler, J. Am. Chem. Soc. 2006, 128, 2365. f) E. G. Guignet, R.
Hovius, H. Vogel, Nat. Biotechnol. 2004, 22, 440.
Fluorescence Titration.
Fluorescence spectra were
recorded on a Perkin-Elmer LS55 spectrofluorophotometer.
The titration experiments were performed at 25 °C using a
solution of hc-DDAADD peptide in buffer solution (50 mM
HEPES, 100 mM NaCl, pH 7.2) in a quartz cell. The concen-
tration of the hc-DDAADD peptide was determined based on
the reported extinction coefficient of 7-hydroxycoumarine-3-
carboxamide (¾370nm = 12300 M¹1 cm¹1). The fluorescence
emission spectra (-ex = 400 nm) were measured after the addi-
tion of a freshly prepared aqueous solution of the zinc com-
2
a) A. Rutkowska, C. H. Haering, C. Schultz, Angew. Chem.,
Int. Ed. 2011, 50, 12655. b) O. Tour, S. R. Adams, R. A. Kerr,
R. M. Meijer, T. J. Sejnowski, R. W. Tsien, R. Y. Tsien, Nat. Chem.
Biol. 2007, 3, 423. c) B. A. Griffin, S. R. Adams, R. Y. Tsien,
Science 1998, 281, 269.
3
N. C. Shaner, P. A. Steinbach, R. Y. Tsien, Nat. Methods
2005, 2, 905.
plexes with a micropipet. Fluorescence titration curves (-em
=
4
a) M. J. Hinner, K. Johnsson, Curr. Opin. Biotechnol. 2010,
446 nm) were analyzed using nonlinear least-square curve-
fitting to evaluate the binding constant (Ka, M¹1).
21, 766. b) A. Keppler, S. Gendreizig, T. Gronemeyer, H. Pick, H.
Vogel, K. Johnsson, Nat. Biotechnol. 2002, 21, 86.
Circular Dichroism (CD) Measurement. CD spectra were
recorded using a JASCO J-720W spectropolarimeter. The
aqueous stock solution of the 3-2Zn(II) was added into the
solution of the peptide (50 ¯M) in water (pH 7.4, adjusted by
10 mM NaOH), and the CD spectrum was measured from 350
to 190 nm (scan speed; 50 nm min¹1) at 25 °C using a quartz
cell (0.1 cm path length). Each spectrum represents the average
of 10 scans with smoothing to reduce noise.
5
a) S. Uchinomiya, A. Ojida, I. Hamachi, Inorg. Chem.
2014, 53, 1816. b) H. Nonaka, S. Fujishima, S. Uchinomiya, A.
Ojida, I. Hamachi, J. Am. Chem. Soc. 2010, 132, 9301. c) A. Ojida,
K. Honda, D. Shinmi, S. Kiyonaka, Y. Mori, I. Hamachi, J. Am.
Chem. Soc. 2006, 128, 10452.
6
A. Ojida, S. Fujishima, K. Honda, H. Nonaka, S.
Uchinomiya, I. Hamachi, Chem.®Asian J. 2010, 5, 877.
7
a) M. Rekharsky, Y. Inoue, S. Tobey, A. Metzger, E.
Anslyn, J. Am. Chem. Soc. 2002, 124, 14959. b) S. Kubik, R.
Kirchner, D. Nolting, J. Seidel, J. Am. Chem. Soc. 2002, 124,
12752.
Covalent Labeling of MBP Protein. A solution of the
tag-fused MBP (1 ¯M final concentration) in degassed 50 mM
HEPES, 100 mM NaCl, 20 ¯M TCEP, pH 7.2 was mixed with
the zinc complex (7-Zn(II) or 8-2Zn(II), 5 ¯M in final con-
centration). The reaction mixture (total volume 100 ¯L) was
incubated at 18 °C for 90 min, in which time the mixture (8 ¯L)
was sampled at the appropriate reaction time and then imme-
diately mixed with 10 mM inorganic pyrophosphate (2 ¯M)
to quench the reaction. After addition of the sampling buffer
and heating at 95 °C for 2 min, the samples were subjected to
SDS-PAGE. The in-gel fluorescence and CBB analyses were
performed with LAS-4000 lumino image analyzer (FUJIFILM)
by EPI (365 nm excitation, Y515Di filter) and DIA-W mode,
respectively.
8
a) F. Huang, C. Cheng, G. Feng, J. Org. Chem. 2012, 77,
11405. b) J. A. Drewry, S. Fletcher, H. Hassan, P. T. Gunning, Org.
Biomol. Chem. 2009, 7, 5074.
9
The pKa values were calculated by Advance Chemistry
Development Software (ACD/pKa, Ver 11.02).
10 H. Nonaka, S. Tsukiji, A. Ojida, I. Hamachi, J. Am. Chem.
Soc. 2007, 129, 15777.
11 The number of alanine residues inserted between the oligo-
aspartate unit and the cysteine were previously optimized, in which
the spacer with the six alanine residues gave the fastest covalent
labeling with a binuclear zinc complex (see Ref. 10).
12 The binding affinity (Ka) of 3-2Zn(II) with PPi was evalu-
ated to be 2.57 © 106 M¹1 by ITC experiment (50 mM HEPES, 100
mM NaCl, pH 7.2). This value was larger than that of 1-2Zn(II)
(Ka = 1.47 © 106 M¹1) with PPi under the same conditions.
13 The labeling yield of CA6DDAADD-MBP with 7-2Zn(II)
was roughly estimated to be 33% after 10 min based on CBB
analysis, in which the relative band intensities of the labeled and
unlabeled MBPs were measured. This labeling yield is higher than
that of CA6D4-MBP with 8-2Zn(II) (22% after 10 min), the trend
of which agrees with the result of the fluorescence analysis.
14 A. Johansson, M. Abrahamsson, A. Magnuson, P. Huang,
J. Mårtensson, S. Styring, L. Hammarström, L. Sun, B. ¡kermark,
Inorg. Chem. 2003, 42, 7502.
We appreciate the technical support from the Research
Support Center, Graduate School of Medical Sciences, Kyushu
University. This work was performed under the Cooperative
Research Program of “Network Joint Research Center for
Materials and Devices.” A. O. acknowledges Takeda Science
Foundation for financial support.
Supporting Information
Experimental data of Job’s plot and CD measurement, and
the protein labeling reaction. Syntheses of the zinc complexes
and the peptides, and expression of the tag-fused MBPs. This
material is available electronically on J-STAGE.
Bull. Chem. Soc. Jpn. 2015, 88, 784–791 | doi:10.1246/bcsj.20150037
© 2015 The Chemical Society of Japan | 791