Selective covalent labeling of tag-fused GPCR proteins on live cell surface with a synthetic probe for their functional analysis

Article abstract of DOI:10.1021/ja910703v

Selective protein labeling with a small molecular probe is a versatile method for elucidating protein functions in living cells. In this paper, we report a covalent labeling method of tag-fused G-protein coupled receptor (GPCR) proteins expressing on cell surfaces utilizing small functional molecules. This method employs the selective and rapid reaction of a peptide tag and a molecular probe, which comprises the cysteine-containing short CA6D4×2 tag (CAAAAAADDDDGDDDD) and a tetranuclear Zn(II)-DpaTyr probe containing a reactive α-chloroacetyl moiety. The covalent labeling of tag-fused GPCRs such as bradykinin receptor (B2R) and acetylcholine receptor (m1AchR) selectively proceeded under physiological conditions during short incubation (10-30 min) with Zn(II)-DpaTyr probes bearing various functional groups. Labeling with fluorophore-appended Zn(II)-DpaTyr probes enabled visualization of the GPCRs on the surface of HEK293 cells by fluorescence. Labeling with the biotin-appended probe allowed introduction of a biotin unit into the GPCRs. This biotin label was utilized for fluorescence bioimaging studies and postlabeling blotting analysis of the labeled GPCRs by use of the specific biotin-streptavidin interaction. The utility of this labeling method was demonstrated in several function analyses of GPCRs, such as fluorescence visualization of the stimuli-responsive internalization of GPCRs and pH change in endosomes containing the internalized GPCRs.

Full text of DOI:10.1021/ja910703v

Published on Web 06/22/2010
Selective Covalent Labeling of Tag-Fused GPCR Proteins on
Live Cell Surface with a Synthetic Probe for Their Functional
Analysis
Hiroshi Nonaka,† Sho-hei Fujishima,† Sho-hei Uchinomiya,† Akio Ojida,†,‡ and
Itaru Hamachi*,†,§
Department of Synthetic Chemistry and Biological Chemistry, Graduate School of Engineering,
Kyoto UniVersity, Katsura Campus, Nishikyo-ku, Kyoto 615-8510, Japan, JST, PRESTO (Life
Phenomena and Measurement Analysis), Sanban-cho, Chiyoda-ku, Tokyo 102-0075, Japan, and
JST, CREST (Creation of Next-generation Nanosystems through Process Integration),
Sanban-cho, Chiyoda-ku, Tokyo 102-0075, Japan
Received December 20, 2009; E-mail: ihamachi@sbchem.kyoto-u.ac.jp
Abstract: Selective protein labeling with a small molecular probe is a versatile method for elucidating protein
functions in living cells. In this paper, we report a covalent labeling method of tag-fused G-protein coupled
receptor (GPCR) proteins expressing on cell surfaces utilizing small functional molecules. This method
employs the selective and rapid reaction of a peptide tag and a molecular probe, which comprises the
cysteine-containing short CA6D4x2 tag (CAAAAAADDDDGDDDD) and a tetranuclear Zn(II)-DpaTyr probe
containing a reactive R-chloroacetyl moiety. The covalent labeling of tag-fused GPCRs such as bradykinin
receptor (B2R) and acetylcholine receptor (m1AchR) selectively proceeded under physiological conditions
during short incubation (10-30 min) with Zn(II)-DpaTyr probes bearing various functional groups. Labeling
with fluorophore-appended Zn(II)-DpaTyr probes enabled visualization of the GPCRs on the surface of
HEK293 cells by fluorescence. Labeling with the biotin-appended probe allowed introduction of a biotin
unit into the GPCRs. This biotin label was utilized for fluorescence bioimaging studies and postlabeling
blotting analysis of the labeled GPCRs by use of the specific biotin-streptavidin interaction. The utility of
this labeling method was demonstrated in several function analyses of GPCRs, such as fluorescence
visualization of the stimuli-responsive internalization of GPCRs and pH change in endosomes containing
the internalized GPCRs.
molecular probes have been developed.3 Most of these labeling
methods rely on enzyme-catalyzed reactions, in which the
Introduction
Membrane proteins on cell surfaces play significant roles in
myriad biological processes. Most of these proteins are classified
into several groups based on their functions such as membrane
receptor protein, ion channel, or transporter. These proteins can
transduct biological signals or transport biological substances
from the outside to the interior of cells or vice versa in response
to extracellular signals and maintain cell homeostasis.1 The
specific labeling of a protein with a functional molecule such
as a luminescent marker, affinity handle, or activity modulator
is one of the most powerful techniques for the functional analysis
of these proteins. Unlike the use of fluorescence proteins or
luciferases as genetically incorporated luminescent markers,
labeling methods with small probes allow the post-translational
introduction of a variety of functional molecules into a target
protein at any appropriate time with controllable labeling
efficiency. This approach significantly facilitates the functional
analysis of proteins based on various analytical techniques.2
Recently, a number of protein labeling methods using small
enzyme is genetically incorporated into the target protein, thus,
permitting specific reaction with the labeling probe. The most
successful example to date is the SNAP-tag system developed
by Johnsson et al., which employs an irreversible enzymatic
reaction between O6-alkylguanine-DNA alkyltransferase and O6-
benzylguanine derivatives.4 Another protein labeling method
utilizes an artificial molecular recognition system composed of
a short peptide tag and a small molecular probe. The pioneering
work of Tsien et al. led to the development of the first specific
binding pair, consisting of a genetically encodable tetra-cysteine
(2) (a) O’Hare, H. M.; Johnsson, K.; Gautier, A. Curr. Opin. Biol. 2007,
17, 488. (b) Chen, I.; Ting, A. Y. Curr. Opin. Biotechnol. 2005, 16,
1. (c) Soh, N. Sensors 2008, 8, 1004. (d) Ferna´ndez-Sua´rez, M.; Ting,
A. Y. Nat. ReV. Mol. Cell Biol. 2008, 9, 929. (e) Giepmans, B. N. G.;
Adams, S. R.; Ellisman, M. H.; Tsien, R. Y. Science 2006, 312, 217.
(f) Terpe, K. Appl. Microbiol. Biotechnol. 2003, 60, 523.
(3) (a) Gronemeyer, T.; Godin, G.; Johnsson, K. Curr. Opin. Biotechnol.
2005, 16, 453. (b) Carrico, I. S.; Carlson, B. L.; Bertozzi, C. R. Nat.
Chem. Biol. 2007, 3, 321. (c) Los, G. V.; Wood, K. Methods Mol.
Biol. 2006, 356, 195. (d) Lin, C.-W.; Ting, A. Y. J. Am. Chem. Soc.
2006, 128, 4542. (e) Chen, I.; Howarth, M.; Lin, W.; Ting, A. Y.
Nat. Methods 2005, 2, 99. (f) George, N.; Pick, H.; Vogel, H.;
Johnsson, N.; Johnsson, K. J. Am. Chem. Soc. 2004, 126, 8896. (g)
Los, G. V.; et al. ACS Chem. Biol. 2008, 3, 373.
† Kyoto University.
‡ JST, PRESTO.
§ JST, CREST.
(1) (a) Schlessinger, J. Cell 2000, 103, 211–225. (b) Verkman, A. S.; Mitra,
A. K. Am. J. Physiol.: Renal. Physiol. 2000, 278, 13. (c) Torres, G. E.;
Gainetdinov, R. R.; Caron, M. G. Nat. ReV. Neurosci. 2003, 4, 13.
(4) Keppler, A.; Gendreizig, S.; Gronemeyer, T.; Pick, H.; Vogel, H.;
Johnsson, K. Nat. Biotechnol. 2003, 21, 86.
9
10.1021/ja910703v  2010 American Chemical Society
J. AM. CHEM. SOC. 2010, 132, 9301–9309 9301

A R T I C L E S
Nonaka et al.
Figure 1. (a) Molecular design of the labeling probes. (b) Sequences of the CA6D4 and CA6D4x2 tag tethered to a protein of interest.
tag (-Cys-Cys-X-X-Cys-Cys-) and
a
bisarsenical probe
probe system for the selective labeling of membrane proteins.7
(FlAsH).5 More recently, several tag-probe pairs have been
devised and applied to the functional analysis of membrane
proteins, mainly for fluorescence bioimaging studies.6,7 These
labeling methods benefit from the small molecular size of both
the tag and the probe: since these protein labels are small, they
are less likely to perturb protein function and protein-protein
interactions, in contrast to large fluorescent proteins and
enzymes.
Building upon this preliminary system, we further developed a
method for covalently labeling a tag-fused protein with a Zn(II)-
DpaTyr probe such as 1-2Zn(II) (Figure 1).8 In this method,
the desired binding complex is formed and a specific nucleo-
philic reaction takes place between the cysteine residue incor-
porated into the reactive D4 tag (CA6D4 tag; CAAAAAAD-
DDD) and the R-chloroacetyl group of the probe. This
nonenzymatic labeling method, based on artificial tag-probe
pairs, has several potential advantages over the previous enzyme-
catalyzed labeling methods: there is no need to incorporate a
large enzyme/protein domain into target proteins, a wider range
of labeling reagents are available, and labeling conditions are
independent of the enzyme properties. Although we demon-
strated the utility of this covalent labeling method in the in Vitro
modification of soluble proteins, its applicability to cellular
proteins and cell functional analyses was not successful. In this
manuscript, we report that by utilizing the multivalency effect,
our covalent labeling method is effective and selective for
covalently modifying G-protein coupled receptor (GPCR)
proteins fused with a short tag (14-amino acid) at their outer
We recently developed a complementary recognition pair
composed of an oligo-aspartate tag (D4 tag; DDDD) and
multinuclear Zn(II) complexes (Zn(II)-DpaTyrs) as a new tag-
(5) (a) Griffin, B. A.; Adams, S. R.; Tsien, R. Y. Science 1998, 281, 269.
(6) (a) Halo, T. L.; Appelbaum, J.; Hobert, E. M.; Balkin, D. M.;
Schepartz, A. J. Am. Chem. Soc. 2009, 131, 438. (b) Hauser, C. T.;
Tsien, R. Y. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 3693. (c) Lata,
S.; Gavutis, M.; Tampe´, R.; Piehler, J. J. Am. Chem. Soc. 2006, 128,
2365. (d) Guignet, E. G.; Hovius, R.; Vogel, H. Nat. Biotechnol. 2004,
22, 440. (e) Yano, Y.; Yano, A.; Oishi, S.; Sugimoto, Y.; Tsujimoto,
G.; Fujii, N.; Matsuzaki, K. ACS Chem. Biol. 2007, 3, 321.
(7) (a) Ojida, A.; Honda, K.; Shinmi, D.; Kiyonaka, S.; Mori, Y.; Hamachi,
I. J. Am. Chem. Soc. 2006, 128, 10452. (b) Honda, K.; Nakata, E.;
Ojida, A.; Hamachi, I. Chem. Commun. 2006, 4024. (c) Honda, K.;
Fujishima, S.; Ojida, A.; Hamachi, I. ChemBioChem 2007, 8, 1370.
(d) Ojida, A.; Fujishima, S.; Honda, K.; Nonaka, H.; Uchinomiya, S.;
Hamachi, I. Chem. Asian J. 2010, 5, 877.
(8) Nonaka, H.; Tsukiji, S.; Ojida, A.; Hamachi, I. J. Am. Chem. Soc.
2007, 129, 15777.
9
9302 J. AM. CHEM. SOC. VOL. 132, NO. 27, 2010

Selective Covalent Labeling of Tag-Fused GPCR Proteins
A R T I C L E S
Figure 2. Selective covalent labeling of a tag-fused GPCR protein on the surface of living cells.
Figure 3. Comparison of the labeling efficiency of the two labeling reactions: MBP-CA6D4x2 with the dimeric probe 2-4Zn(II) and MBP-CA6D4 with
the monomeric probe 1-2Zn(II). (a) In-gel fluorescence (upper) and CBB (lower) analyses of the labeling reactions. (b) Time-course plot of the fluorescence
signal intensity of MBP protein labeled with 2-4Zn(II) (9) and 1-2Zn(II) (b). Conditions: 0.5 µM MBP protein, 2 µM probe in 50 mM HEPES buffer,
100 mM NaCl, and 100 µM DTT, pH 7.2, 4 °C.
membrane N-termini (Figure 2). The high specificity and
stability of the covalent labeling, as well as the availability of
various functional probes, allows for fluorescence imaging with
various fluorophores as well as postlabeling analysis of the
labeled GPCRs. We further demonstrate that this labeling
method is applicable to several functional analyses of GPCRs,
such as the fluorescence visualization of agonist-stimulated
internalization of GPCRs and the subsequent pH change of the
endosome interior containing the internalized GPCRs.
100-fold stronger than the monomeric binding pair (Kapp ) ∼105
M-1). Prior to the development of GPCR labeling on cell
surfaces, labeling efficiency was preliminarily evaluated in in
Vitro reactions using soluble maltose binding protein (MBP)
tethered to a CA6D4x2 tag (CAAAAAADDDDGDDDD, MBP-
CA6D4x2) or a CA6D4 tag (CAAAAAADDDD, MBP-CA6D4)
at its C-terminus, both of which have an optimized number of
the alanine spacer (A6) between the reactive cysteine and the
D4 sequence.8 Figure 3 shows a comparison of the labeling
efficiency of the two reactions, that is, MBP-CA6D4x2 with
the dimeric probe 2-4Zn(II) and MBP-CA6D4 with the
monomeric probe 1-2Zn(II), at a lower protein concentration
(0.5 µM) than the previous labeling conditions which used g5
µM tag-fused protein. It is apparent by in-gel fluorescence
analysis that the labeling reaction of MBP-CA6D4x2 with
2-4Zn(II) rapidly proceeds and almost saturates within 60 min,
even at low temperature (4 °C). This is in contrast to the
monomer pair of MBP-CA6D4 and 1-2Zn(II), in which the
fluorescence intensity of the labeled MBP-CA6D4 is much
lower, even after 240 min. These results indicate that the
stronger binding of the dimer pair effectively facilitates the
labeling reaction, whereas the interaction of the monomer pair
is inefficient at lower concentrations due to the weak binding
affinity (Kapp ) ∼105 M-1). We, thus, decided to employ the
dimer pair for covalent modification of GPCR proteins express-
ing on living cell surfaces.
Results and Discussion
Design of Tag-Probe Pair for More Efficient Covalent
Labeling. In our previous report, we conducted labeling reactions
using a rather high concentration of the tag-fused protein (g5
µM) in Vitro because the binding affinity of the Zn(II)-DpaTyr
probe 1-2Zn(II) for the D4 tag (DDDD) is relatively weak (Kapp
) ∼105 M-1).8 We assumed that it would be necessary to
enhance the binding affinity of the tag-probe pair in order to
achieve effective labeling of a tag-fused GPCR that usually
expresses on cell surfaces at rather low concentration. We
previously reported that the binding affinity of the tag-probe
pair was significantly enhanced by doubling the number of each
binding site due to the multivalent effect.7,9 For example, the
binding affinity between the doubly repeated D4 tag (D4x2 tag;
DDDDGDDDD) and the dimeric Zn(II)-DpaTyr probe
2-4Zn(II) was increased up to 1.5 × 107 M-1, which is over
Labeling of GPCRs on Living Cell Surfaces. Bradykinin
receptor type2 (B2R) and porcine m1 type acetylcholine receptor
(m1AchR) were used as target GPCRs for our covalent labeling
method. B2R represents one of the most important GPCRs and
(9) (a) Mammen, M.; Choi, S.-K.; Whitesides, G. M. Angew. Chem., Int.
Ed. 1998, 37, 2754. (b) Gestwicki, J. E.; Cairo, C. W.; Strong, L. E.;
Oetjen, K. A.; Kiessling, L. L. J. Am. Chem. Soc. 2002, 124, 14922.
9
J. AM. CHEM. SOC. VOL. 132, NO. 27, 2010 9303

A R T I C L E S
Nonaka et al.
Figure 4. Covalent labeling of GPCR expression on the surface of HEK293 cells. Fluorescence imaging of CA6D4x2 tag-fused B2R (a), B2R lacking
CA6D4x2 tag (b), and CA6D4x2 tag-fused m1AchR (c) by labeling with TAMRA-appended probe 2-4Zn(II). (d) Fluorescence imaging of CA6D4x2
tag-fused B2R receptor by labeling with Oregon Green-appended probe 4-4Zn(II). In each labeling experiment, the fluorescence images were obtained
using two different channels corresponding to the probe (i) and a protein expression marker (ii) such as Cy5-appended antagonist peptide or EGFP. The
transmission image is shown in (iii), and the overlay image of (i) and (ii) is shown in (iv). Scale bar is 10 µm.
has been extensively studied with respect to its fundamental
cell biology, and as a therapeutic target for treating diseases.10
For the labeling experiment, a human B2R tethered to a
CA6D4x2 tag (-CAAAAAADDDDGDDDD-) at the exoplasmic
N-terminus was transiently expressed in HEK293 cells. We
employed TAMRA-appended probe 2-4Zn(II) for the labeling
reaction; this probe is cell impermeable and, thus, suitable for
labeling cell surface proteins. Short incubation (10 min) of the
cells with 2-4Zn(II) (1 µM) enabled fluorescence imaging of
the tag-fused B2R on the cell surfaces. However, this fluores-
cence intensity was significantly diminished upon washing with
HBS buffer containing a high concentration of pyrophosphate
(PPi, 1 mM), a strong competitive binder for Zn(II)-DpaTyrs.
This result indicates that the 2-4Zn(II) probe binds reversibly
through metal-ligand interactions, and does not form a stable
covalent bond with the cysteine residue of the tag. After a series
of trials, we obtained durable fluorescence images when labeling
was conducted after short treatment (10 min) with the reducing
agent, tris-carboxyethylphosphine (TCEP) (Figure 4). The
effectiveness of this reductive pretreatment strongly suggests
that the cysteine residue of the tag, which would be deactivated
under extracellular oxidative conditions, is reactivated as a
nucleophile. Fluorescence remained after extensive washing with
a high concentration of inorganic pyrophosphate (1 mM),
indicating that B2R was covalently labeled. A 10-min labeling
period was thus sufficient to obtain a fluorescence signal of B2R
by confocal laser-scanning microscopy. The labeling specificity
for B2R was confirmed by co-staining with a B2R antagonist
peptide appended with a Cy5 fluorophore. Fluorescence labeling
detectable by the Cy5 and TAMRA channels predominantly
overlapped, as shown in Figure 4a. A control labeling experi-
ment with B2R lacking the CA6D4x2 tag produced negligible
fluorescence labeling in the TAMRA channel (Figure 4b), which
was quantitatively evaluated to be less than 3% relative to the
intensity observed in Figure 4a. Another control labeling
experiment using the tag-fused B2R and a nonreactive probe
lacking an R-chloroacetyl group 3-4Zn(II) also afforded no
detectable fluorescence labeling in the TAMRA channel (Figure
S1). The selective fluorescence detection of the tag-fused B2R by
2-4Zn(II), together with the two control experiments, indicated
that, on living cell surfaces, the covalent labeling reaction occurs
site-specifically between the tag fused to B2R and the R-chloro-
acetyl group of the probe. The cell viability assay using trypan
blue revealed that the cells maintain a high viability rate (>98%)
within 24 h after the labeling with 2-4Z(II) (Figure S2), indicating
that the labeling treatment with TCEP and the probe did not exhibit
significant cell toxicity and cell death induction.
(10) (a) Leeb-Lundberg, L. M. F.; Marceau, F.; Mu¨ller-Esterl, W.; Pettibone,
D. J.; Zuraw, B. L. Pharmacol. ReV. 2005, 57, 27. (b) Marceau, F.;
Regoli, D. Nat. ReV. Drug DiscoVery 2004, 3, 845.
This labeling method is applicable to another GPCR,
m1AchR,11 which has a CA6D4x2 tag and an enhanced green
9
9304 J. AM. CHEM. SOC. VOL. 132, NO. 27, 2010

Selective Covalent Labeling of Tag-Fused GPCR Proteins
A R T I C L E S
Figure 5. (a) Fluorescence imaging of CA6D4x2 tag-fused B2R by labeling with biotin-appended probe 5-4Zn(II) and subsequent visualization with
SA-HiLyte647. The fluorescence images were obtained using two different channels corresponding to HiLyte647 (i) and TAMRA appended to the B2R
antagonist (ii). The transmission image is shown in (iii), and the overlay image of (i) and (ii) is shown in (iv). Scale bar is 10 µm. (b) Blotting analyses of
biotin-labeled B2R by chemiluminescence detection using streptavidin-HRP conjugate (lanes 1-4) and Western blotting using anti-B2R antibody (lanes
5-8). The two bands marked with asterisk derive from endogenous biotinated proteins. The labeling reaction was conducted in the presence (lanes 2, 4, 6,
and 8) or absence (lanes 1, 3, 5, and 7) of PPi (1 mM).
fluorescent protein (EGFP) domain at the exoplasmic N-
terminus. Fluorescence labeling of m1AchR with 2-4Zn(II) was
carried out by the same procedure as used for B2R labeling.
As shown in Figure 4c, clear fluorescence labeling in the
TAMRA channel was detected on the surface of HEK293 cells
expressing m1AchR, which co-localized well with EGFP
fluorescence. In the control experiment, HEK293 cells express-
ing m1AchR which lacked the CA6D4x2 tag showed negligible
fluorescence labeling of 2-4Zn(II) (Figure S3), demonstrating
again the tag-specific labeling of GPCRs.
noncovalent tag-probe interactions, clearly demonstrating the
advantage of the present method in protein analyses.
Evaluation of Labeling Efficiency of GPCR. Careful analysis
of fluorescence images of living cells allowed quantitative
evaluation of the labeling rate of GPCRs. After treatment of
HEK293 cells transiently expressing B2R with 2-4Zn(II), the
pixel-by-pixel scatterplot between two fluorescence channels
corresponding to TAMRA of the probe and Cy5 of the B2R
antagonist peptide gave a distribution graph (Figure 6a), which
was analyzed by linear curve-fitting to give the slope value.
Since the slope, which is the averaged fluorescence intensity
of the probe based on the antagonist fluorescence, increased as
the labeling reaction proceeded, we can simply define the slope
as the relative labeling ratio (r). The r values obtained at
different labeling times are shown in Figure 6b. For the labeling
reaction between the CA6D4x2 tag-fused B2R with 2-4Zn(II),
the r value increased significantly in a time-dependent manner
and reached a plateau after 30 min. The change in the r value
agrees well with a theoretical saturation curve based on pseudo-
first-order reaction. After 60 min, reductive treatment with TCEP
followed by a second labeling with a different probe, 3-4Zn(II),
did not alter fluorescence labeling of B2R, suggesting that the
labeling reaction with 2-4Zn(II) was essentially quantitative.13
In contrast, for the first labeling reaction with the nonreactive
probe, 3-4Zn(II), which lacks an R-chloroacetyl group, a
negligible change was induced in the r value during the reaction,
This protein labeling method allowed for the introduction of
various functional molecules into the tag-fused GPCR. When
we used Zn(II)-DpaTyr probe 4-4Zn(II), which contains
Oregon Green 488 as the fluorophore, labeling of tag-fused B2R
with 4-4Zn(II) proceeded smoothly to provide clear fluores-
cence images in the Oregon Green channel (Figure 4d). Proteins
labeled with distinct fluorescence colors are essential for
multicolor imaging in cells. The biotin-appended probe
5-4Zn(II) is also available for GPCR labeling; this was
confirmed by treatment of biotin-labeled B2R with streptavidin
modified with the near-infrared HiLyte647 dye, SA-HiLyte647.
A bright and clear fluorescence image was obtained due to
strong florescence from the multiple HiLyte647 dye molecules
appended to streptavidin (Figure 5a). The biotin unit is also
useful for the postlabeling analysis of GPCRs. For example,
HEK293 cells displaying labeled B2R were lysed and subjected
to biotin blotting. Figure 5b shows a distinctive smear band in
the range from 75 to 50 kDa (lane 1), which was coincident
with the Western blotting pattern obtained using anti-B2R
antibody (lane 5).12 Control labeling experiments using PPi as
an effective inhibitor of the tag-probe binding (lanes 2 and 4)
or B2R lacking a CA6D4x2 tag (lanes 3 and 4) gave negligible
bands upon biotin blotting analysis. These results indicate that
the labeling reaction with 5-4Zn(II) specifically occurs with
tag-fused B2R. It is noteworthy that such postlabeling analysis
is practically impossible with reversible labeling based on
(13) To confirm the labeling reaction efficiency for the overexpressed tag-
fused B2R in live cells, we quantitatively evaluated its expression
number using fluorescence microscopy analysis (Supporting Informa-
tion, Table S1, and Figure S4). On the basis of the local concentration
of B2R on the cell surface, the number of B2R was estimated to be
6.3 ( 1.3 × 104/cell, which is comparable to a reported value (1.43
× 104/cell) evaluated in a cell line of HEK293 stably expressing the
native B2R (Lamb, M. E. et al., Biochem. J. 2001, 355, 741.). In the
fluorescence microscopy analysis for the quantification, the strong
fluorescence of the Cy5-appended B2R antagonist was selectively
detected on the cell surface, indicating that B2R is not homogeneously
distributed in solution. Indeed, we evaluated the local concentration
of B2R on the cell surface as 25.0 ( 5.2 µM, whereas its formal
concentration in solution is calculated to be a significantly lower value
(57 ( 12 nM), suggesting that B2R is concentrated by ca. 500-fold
on the cell surface. These results may imply that the substantially high
concentration of B2R on the cell surface largely contributes to the
efficient labeling of B2R.
(11) Van Koppen, C. J.; Kaiser, B. Pharmacol. Ther. 2003, 98, 197.
(12) (a) Michineaul, S.; Muller, L.; Pizard, A.; Alhenc-Gelas, F.; Rajerison,
R. M. Biol. Chem. 2004, 385, 49. (b) Blaukat, A.; Pizard, A.; Rajerison,
R. M.; Alhenc-Gelas, F.; Muller-Esterl, W.; Dikic, I. FEBS Lett. 1999,
451, 337.
9
J. AM. CHEM. SOC. VOL. 132, NO. 27, 2010 9305

A R T I C L E S
Nonaka et al.
Figure 6. (a) Pixel-by-pixel scatterplot analysis of fluorescence images of cells using two fluorescence channels corresponding to TAMRA appended to
2-4Zn(II) and Cy5 appended to the B2R antagonist peptide. Linear curve-fitting analysis of the scatterplot gives the slope value defined as the relative
labeling efficiency (r) at each labeling time (0, 10, 30, and 60 min). (b) Time-course of the relative labeling efficiency (r) in the labeling reaction of tag-fused
B2R with 2-4Zn(II) (9), 1-2Zn(II) (b), and 3-4Zn(II) (2).
Figure 7. Time-lapse fluorescence imaging of the internalization of CA6D4x2 tag-fused B2R (a) and m1AchR (b) labeled with 2-4Zn(II) by agonist
stimulation with bradykinin (200 nM) and calbachol (100 µM), respectively. Scale bar is 5 µm.
indicating no labeling of B2R. A slight increase in the r value
was observed during the labeling reaction with the monomeric
probe 1-2Zn(II) after 60 min (r ) 0.13), but this value is
approximately 12-fold less than that obtained with 2-4Zn(II)
after the same reaction time (r ) 1.56). This result indicates
that the labeling efficiency of the monomer-type probe
1-2Zn(II), is much lower than that of the dimer probe
2-4Zn(II), consistent with the in Vitro labeling experiment using
the tag-fused MBP a s shown in Figure 3. We, thus, concluded
that the strong binding between the CA6Dx2 tag and a
tetranuclear Zn(II)-DpaTyr probe such as 2-4Zn(II), supported
by the multivalency effect, is essential for the efficient covalent
labeling of B2R expressing on cell surfaces.
tory mechanism of GPCR-mediated cell signaling.14 We ex-
amined fluorescence visualization of GPCR internalization to
demonstrate the utility of this labeling method for analyzing
dynamic cell functions involving GPCR. Figure 7a shows time-
lapse imaging of agonist-induced internalization of CA6D4x2
tag-fused B2R labeled with TAMRA-appended probe 2-4Zn(II)
in HEK293 cells. Cell stimulation with bradykinin (200 nM), a
native agonist peptide of B2R,15 clearly induced the intracellular
movement of fluorescent endosome particles containing labeled
B2R after 30 min; by 60 min, these particles were localized in
the cytosol. Such dynamic fluorescence change was not observed
during the same time frame (60 min) in the absence of
Fluorescence Visualization of GPCR Internalization. It is
well-known that many GPCRs expressing on the cell surface
migrate into the cytosol as components of endosome particles,
induced by strong stimulation with the corresponding agonists.
This phenomenon, called internalization, is an important regula-
(14) Pierce, K. L.; Premont, R. T.; Lefkowitz, R. J. Nat. ReV. Mol. Cell
Biol. 2002, 3, 63.
(15) (a) Howl, J.; Langel, U.; Hawtin, S. R.; Valkna, A.; Yarwood, N. J.;
Saar, K.; Wheatley, M. FASEB J. 1997, 11, 582–590. (b) Bachvarov,
D. R.; Houle, S.; Bachvarova, M.; Bouthillier, J.; Adam, A.; Marceau,
F. J. Pharmacol. Exp. Ther. 2001, 297, 19–26.
9
9306 J. AM. CHEM. SOC. VOL. 132, NO. 27, 2010

Selective Covalent Labeling of Tag-Fused GPCR Proteins
A R T I C L E S
Figure 8. Pulse-chase visualization of internalized B2R involving endosomes. (a) Schematic representation of the labeling procedure (i-iv) and the merged
fluorescence image in each step (I-V); (i) labeling of CA6D4x2 tag-fused B2R with 2-4Zn(II) and 5-4Zn(II), (ii) first internalization induced by bradykinin,
(iii) fluorescence visualization of the biotin-labeled B2R with SA-HiLyte647, and (iv) second internalization induced by bradykinin. (b and c) Fluorescence
images of labeled B2R in steps IV (b) and V (c). The images were obtained using two different channels corresponding to TAMRA appended to 2-4Zn(II)
(i) and SA-HiLyte647 (ii). The transmission image is shown in (iii), and the overlay image of (i) and (ii) is shown in (iv). Scale bar is 5 µm.
stimulation (Figure S5), indicating that the phenomena shown
in Figure 7a is agonist-dependent internalization. Agonist-
induced internalization of m1AchR was also clearly visualized
(Figure 7b). In the time-lapse imaging, many fluorescent
endosome particles containing m1AchR labeled with 2-4Zn(II)
were gradually internalized into the cytosol of HEK293 cells
upon stimulation with carbachol (100 µM).16 These two
examples demonstrate that our protein labeling method is useful
for fluorescence monitoring of GPCR internalization. Signifi-
cantly, fundamental functions such as agonist binding ability
are not noticeably compromised, and the agonist-induced
conformational changes involving stimulus-responsive oligo-
merization, essential to GPCR internalization, are also unaf-
fected. To further confirm whether the labeling with the probe
exerts any influence on the B2R function, we measured the
receptor activated Ca2+ signaling of the HEK293 cells. The
intracellular Ca2+ concentration ([Ca2+]i), which was monitored
by the ratio of the fura-2 emission excited at 340 and 380 nm
(ex340/ex380), increased upon the addition of bradykinin (100
nM). The ([Ca2+]i) responses observed in the B2R labeled with
2-4Zn(II) were nearly identical to those in the nonlabeled B2R
(Figure S6), indicating that the labeling with the probe does
not interfere with the original activity of B2R in the intracellular
signaling. Interestingly, only a subtle [Ca2+]i response was
observed in the case of the B2R fused with EGFP at the
exoplasmic N-terminus, indicative of functional impairment of
the B2R against the agonist stimulation. These results suggest
an advantage of the present labeling method over fluorescent
proteins fusion method in the point of less disturbance of protein
function, which might be ascribed to the smaller size of the tag
and probe pair than EGFP.
We next performed pulse-chase visualization of the internal-
ized B2R by labeling using two different probes (Figure 8).
First, the CA6D4x2 tag-fused B2R expressing on HEK293 cells
was simultaneously labeled with the TAMRA-appended probe
2-4Zn(II) and the nonfluorescent biotin-appended probe
5-4Zn(II) (step II in Figure 8a). The cells were then treated
with bradykinin to induce agonist-dependent internalization (step
III in Figure 8a) followed by SA-HiLyte647 to fluorescently
visualize B2R labeled with 5-4Zn(II). B2R on the cell surfaces
was detectable by both the TAMRA (green) and HiLyte647 (red)
fluorescence channels, whereas the endosome particles contain-
ing the internalized B2R were selectively visualized by TAMRA
fluorescence inside the cells (step IV in Figure 8a,b). A second
bradykinin stimulation was then conducted to induce the second
internalization (step V in Figure 8a), followed by detection of
the newly internalized endosomes by fluorescence of HiLyte647
(red) inside the cells (Figure 8c-ii). In the merged image of the
two fluorescence channels (Figure 8c-iv), the first internalized
endosomes were visualized as green, whereas the second
internalized endosomes were selectively visualized as yellow
particles due to co-labeling of B2R with both green and red
fluorophores. Therefore, pulse-chase visualization using two
different probes allows for fluorescent discrimination of B2R-
containing endosomes internalized at different times.
Fluorescence Detection of Interior Environment of
Endosome. It is known that the interior environment of endo-
somes is different from the environment in the cytosol. For
example, endosomal pH gradually decreases from neutral to pH
5-6 between the early and late stages of endocytosis.17
Fluorescence analysis of the interior environment of endosomes
(16) Shockley, M. S.; Burford, N. T.; Sadee, W.; Lameh, J. J. Neurochem.
1997, 68, 601–609.
9
J. AM. CHEM. SOC. VOL. 132, NO. 27, 2010 9307

A R T I C L E S
Nonaka et al.
Figure 9. Time-lapse fluorescence detecting endosomal pH change using a pH-responsive fluorophore, Cypher5E, linked to tag-fused B2R. Each column
shows a transmission image, the fluorescence channel for HiLyte488, the fluorescence channel for Cypher5E, and the ratio image of the two fluorescence
channels (CypHer5E/HiLyte488) from left to right.
Conclusion
by incorporation of environmentally sensitive probes is possible
using our GPCR labeling method. We attempted to ratiometri-
cally detect the change in endosomal pH using a pH-responsive
fluorophore (Figure 9). After introducing a biotin unit into tag-
fused B2R by 5-4Zn(II), simultaneous labeling with two
different streptavidins (SAs) (appended to pH-responsive
CypHer5E or to pH-insusceptible HiLyte488) was conducted
to afford the dual-emission pH sensing system. Initially (0 h),
clear fluorescence due to HiLyte488 was observed on the cell
surfaces, whereas no fluorescence of CypHer5E was detected
because the fluorophore existed in the nonfluorescent proton-
free form (pKa ) 7.3) under the slightly basic buffered
conditions (pH 8.0).18 After internalization stimulated by
bradykinin, fluorescence of CypHer5E inside the cells gradually
intensified over a 2-h period, co-localizing with that of Hi-
Lyte488. This result indicates that SA-CypHer5E fluorescently
senses the decrease in endosomal pH due to the formation of the
fluorescent protonated form. The acidic pH shift inside the
endosome particles was also detectable by ratiometric analysis of
the two fluorescence images (CypHer5E/HiLyte488), which more
precisely visualizes the distribution of acidic endosomes inside cells
compared to the single channel image due to CypHer5E.
In conclusion, we demonstrated a useful labeling method for
selective covalent modification of GPCR proteins expressing
on cell surfaces. The advantages of this nonenzymatic labeling
method include (i) a simple labeling procedure with a short
reaction time (10-30 min); (ii) flexible availability of the
functional molecular probes, which enables fluorescence imaging
of GPCRs with various fluorophores and postlabeling biotin-
blotting analysis; and (iii) small molecular size of the tag and
probe, which allows agonist-induced internalization without
significant perturbation of GPCR functions. We also demon-
strated the applicability of this labeling method to fluorescence
analysis of GPCR internalization, and analysis of environmental
changes occurring in the interior of endosomes. To our
knowledge, our labeling method is the first to allow functional
modification of cell surface proteins by a nonenzymatic reaction
between a short peptide tag and a small molecular probe.
Moreover, this labeling method has potential for other cell
functional analyses, such as detection of GPCR oligomerization
by using a labeling probe with a photo cross-linking unit and
analysis of dynamic structural changes in membrane-bound
proteins using ESR probes. In addition to such fundamental
researches, the present GPCR labeling and imaging in living
cells may be expectedly useful for more practical applications
such as convenient high-throughput screening of drug candidates
targeting GPCR-involved cell functions. Research along these
lines is currently underway.
(17) (a) Urano, Y.; Asanuma, D.; Hama, Y.; Koyama, Y.; Barrett, T.;
Kamiya, M.; Nagano, T.; Watanabe, T.; Hasegawa, A.; Choyke, P. L.;
Kobayashi, H. Nat. Med. 2009, 15, 104–109. (b) Maxfield, F. R.;
McGraw, T. E. Nat. ReV. Mol. Cell Biol. 2004, 5, 121. (c) Hanyaloglu1,
A. C.; Zastrowe, M. V. Annu. ReV. Pharmacol. Toxicol. 2008, 48,
537.
Experimental Section
(18) Adie, E. J.; Kalinka, S.; Smith, L.; Francis, M. J.; Marenghi, A.;
Cooper, M. E.; Briggs, M.; Michael, N. P.; Milligan, G.; Game, S.
BioTechniques 2002, 33, 1152.
General Methods. Unless otherwise noted, all proteins/enzymes
and reagents were obtained from commercial suppliers and used
9
9308 J. AM. CHEM. SOC. VOL. 132, NO. 27, 2010

Selective Covalent Labeling of Tag-Fused GPCR Proteins
A R T I C L E S
without further purification. Streptavidin-HiLyte647 and HiLyte488
were obtained from AnaSpec, Inc. Bradykinin and carbachol were
obtained from Wako Pure Chemical Industries.
containing 1% protease inhibitor cocktail set III (Novagen) at 4 °C
for 60 min. The insoluble material was removed by centrifugation
at 12 000 rpm for 10 min. The supernatant was mixed with 2×
sample buffer for membrane protein (pH 6.8 125 mM Tris ·HCl,
20% glycerol, 4% SDS, 0.01% bromophenol blue, 100 mM DTT)
and subjected to SDS-PAGE (10% T, 0.33% C) followed by
electrotransfer onto an Immun-Blot PVDF membrane (Bio-Rad).
The immunodetection of the B2R was performed with anti-B2R
antibody (1:200 dilution, Santa Cruz Biotechnology) and horserad-
ish peroxidase (HRP)-conjugated anti-rabbit IgG antibody (1:5000
dilution, Santa Cruz Biotechnology). The chemiluminescence signal
using Chemi-Lumi One (Nacalai Tesque) was detected with
ChemiDoc XRS (BIO-RAD). After removing anti-B2R antibody
by washing with the stripping buffer (pH 2.0 2.5 mM glycine, 1%
SDS), the membrane was treated with HRP-conjugated streptavidin
(1: 5000 dilution; Invitrogen) and subjected to the chemilumines-
cence detection.
Pulse-Chase Labeling of B2R Internalization. The tag-fused
B2R expressing on HEK293 cell surface was labeled by the
treatment with 1 mL of HBS (pH 7.4) containing 1.0 µM 2-4Zn(II)
and 2.0 µM 5-4Zn(II). The dish was incubated for 10 min at room
temperature. After removing the probe solution, the cells were
washed with 1.0 mL of PBS containing 1 mM PPi. For the first
internalization, the cells were stimulated with 200 nM bradykinin
for 60 min at 25 °C. After washing the bradykinin solution, the
cells were treated with 1 mL of PBS containing 5 µg of Streptavidin
(SA)-HiLyte647. For the second internalization, the cells were
stimulated by 200 nM bradykinin for 30 min at 25 °C. The CLSM
analysis was performed with a ×100 lens and appropriate emission
filter for TAMRA and SA-HiLyte647.
Covalent Labeling of the CA6D4-Tagged MBP and
CA6D4x2-Tagged MBP. A solution of MBP (500 nM in 50 mM
HEPES, 100 mM NaCl, 0.1 mM DTT, pH 7.2) was mixed with
1-2Zn(II) or 2-4Zn(II) (2 µM in final concentration). The mixture
was incubated at 4 °C for 2 h, in which 40 µL of the mixture was
sampled at the appropriate time and treated with Laemmli buffer
to quench the labeling reaction. The sampling solutions were heated
at 95 °C for 3 min, and subjected to SDS-PAGE. The in-gel
fluorescence analysis was performed with LAS-4000 lumino image
analyzer (FUJIFILM) by EPI mode (520 nm excitation, 575DF20
filter).
Cell Culture and Recombinant Protein Expression in
HEK293 Cells. HEK293 cells were cultured in high glucose
Dulbecco’s Modified Eagle Medium (DMEM, 4.5 g of glucose/L)
supplemented with 10% fetal bovine serum (FBS), penicillin (100
units/mL), streptomycin (100 µg/mL), and amphotericin B (250
ng/mL). Cells were maintained at 37 °C in a humidified atmosphere
of 5% CO2 in air. A subculture was performed every 3-4 days
from subconfluent (<80%) cultures using a trypsin-EDTA solution.
Transfection of cDNA plasmids was carried out in a 35 mm glass-
bottomed dish (Iwaki) using Lipofectamine LTX (Invitrogen)
according to a general protocol. The cells were subjected to labeling
experiments after 44-52 h of the transfection.
General Procedure for GPCR Labeling and Fluorescence
Imaging. The HEK293 cells expressing the tag fused B2R (∼1 ×
106) were incubated with 1 mL of HEPES-buffered saline (HBS,
containing 107 mM NaCl, 6 mM KCl, 1.2 mM MgSO4, 2 mM
CaCl2, 11.5 mM glucose, 20 mM HEPES, adjusted to pH 7.4 with
NaOH) containing 0.5 mM TCEP for 10 min at room temperature.
After removing TCEP solution, the cells were treated with 1 mL
of HBS (pH 7.4) containing the labeling probe (0.5 µM) for 10
min at room temperature. The probe solution was removed and
the cells were washed with 1.0 mL of Phosphate Buffered Saline
(PBS) containing 1 mM PPi to remove the unreacted probe. After
being filled with PBS containing Cy5-appended B2R antagonist
peptide (0.5 µM), the cells were analyzed using a confocal laser
scanning microscope (CLSM; Olympus, FLUOVIEW FV1000)
equipped with a ×100 lens and the appropriate emission filters for
Oregongreen488, EGFP, TAMRA, and Cy5. In the case of the
imaging study using Streptavidin(SA)-HiLyte647, the cells labeled
with 5-4Zn(II) (1.0 µM) were further treated with SA-HiLyte647
(5 µg in 1 mL PBS) and subjected to the imaging analysis.
Blotting Analysis of B2R Labeling. HEK293 cells expressing
the tag fused B2R were labeled with 5-4Zn(II) by the procedure
described above. The cells were collected by cell scraper and lysed
with 200 µL of RIPA lysis buffer (pH 7.6 25 mM Tris ·HCl, 150
mM NaCl, 0.1% SDS, 1% Nonidet P-40, 1% Deoxycholic acid)
Sensing of pH Change in Endosomes Containing GPCR. The
tag-fused B2R expressing on HEK293 cell surface was labeled by
the treatment with 1 mL of HBS (pH 7.4) containing 1.0 µM
5-4Zn(II). The dish was incubated for 20 min at room temperature.
After removing the probe solution, the cells were washed with 1.0
mL of PBS containing 1 mM PPi and treated with 1 mL of PBS
containing 2 µg of SA-HiLyte488 and 5 µg of SA-CypHer5E. After
washing with PBS, the cells were stimulated with 200 nM
bradykinin for 60 min at 25 °C. The CLSM analysis was performed
with a ×100 lens and appropriate emission filter for TAMRA and
SA-HiLyte647.
Supporting Information Available: Detailed synthetic pro-
cedures and compound characterizations, plasmid constructions,
evaluation of cell toxicity and expression number of receptor
protein, functional analysis of the labeled receptor, and the
complete ref 3g. This material is available free of charge via
the Internet at http://pubs.acs.org.
JA910703V
9
J. AM. CHEM. SOC. VOL. 132, NO. 27, 2010 9309