Intracellular Protein-Labeling Probes for Multicolor Single-Molecule Imaging of Immune Receptor-Adaptor Molecular Dynamics

Article abstract of DOI:10.1021/jacs.7b08262

Single-molecule imaging (SMI) has been widely utilized to investigate biomolecular dynamics and protein-protein interactions in living cells. However, multicolor SMI of intracellular proteins is challenging because of high background signals and other lim

Full text of DOI:10.1021/jacs.7b08262

Subscriber access provided by READING UNIV
Article
Intracellular protein-labeling probes for multicolor single-
molecule imaging of immune receptor-adaptor molecular dynamics
Ryota Sato, Jun Kozuka, Masahiro Ueda, Reiko Mishima, Yutaro Kumagai,
Akimasa Yoshimura, Masafumi Minoshima, Shin Mizukami, and Kazuya Kikuchi
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b08262 • Publication Date (Web): 09 Nov 2017
Downloaded from http://pubs.acs.org on November 9, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 14
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Intracellular protein-labeling probes for multicolor single-
molecule imaging of immune receptor-adaptor molecular dy-
namics
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Ryota Sato1‡, Jun Kozuka2‡, Masahiro Ueda2*, Reiko Mishima3, Yutaro Kumagai3*, Akimasa Yoshiꢀ
mura1, Masafumi Minoshima1, Shin Mizukami4 & Kazuya Kikuchi1,5*
1 Department of Material and Life Science, Graduate School of Engineering, Osaka University, Suita, Osaka 565ꢀ0871, Jaꢀ
pan
2 RIKEN Quantitative Biology, Suita, Osaka 565ꢀ0874, Japan
3 Quantitative Immunology Research Unit, WPIꢀImmunology Frontier Research Center, Osaka University, Suita, Osaka 565ꢀ
0871, Japan
4 Institute of Multidisciplinary Research for Advanced Materials, Tohoku University, Sendai, Miyagi, 980ꢀ8577, Japan
5 Chemical Imaging Techniques, WPIꢀImmunology Frontier Research Center, Osaka University, Suita, Osaka 565ꢀ0871,
Japan
KEYWORDS: fluorescent probes, singleꢀmolecule imaging, proteinꢀlabeling tag, protein dynamics, innate immune system.
ABSTRACT: Singleꢀmolecule imaging (SMI) has been widely utilized to investigate biomolecular dynamics and proꢀ
teinꢀprotein interactions in living cells. However, multicolor SMI of intracellular proteins is challenging because of
high background signals and other limitations of current fluorescence labeling approaches. To achieve reproducible
intracellular SMI, a labeling probe ensuring both efficient membrane permeability and minimal nonꢀspecific binding to
cell components is essential. We developed nearꢀinfrared (NIR) fluorescent probes for protein labeling, which specifiꢀ
cally binds to a mutant βꢀlactamase tag. By structural fineꢀtuning of cell permeability and minimized nonꢀspecific
binding, SiRcB4 enabled multicolor SMI in combination with a HaloTagꢀbased redꢀfluorescent probe. Upon addition
of both chemical probes at subꢀnanomolar concentrations, singleꢀmolecule imaging revealed the dynamics of TLR4
and its adaptor protein, TIRAP, which are involved in the innate immune system. Statistical analysis of the quantitative
properties and timeꢀlapse changes in dynamics revealed a proteinꢀprotein interaction in response to ligand stimulation.
Genetically encoded fluorescent proteins (FPs) are
widely used as fluorescence reporters. However, the
ability to use FPs for SMI and other photonꢀintensive
imaging applications is limited because of their poor
photostability10. Moreover, FPs have common limitaꢀ
tions in strictly controlling their expression levels. For
SMI, the labeling density of fluorescent molecules in
regions of interest must be controlled to observe indiꢀ
vidual fluorophores. It is challenging to optimize the
experimental conditions to observe an appropriate numꢀ
ber of fluorescent molecules. Dualꢀcolor SMI using two
FPs is particularly challenging, as it requires optimizaꢀ
tion of the expression levels of both FPs to determine
conditions under which FPs display similar fluorescence
intensities within a cell.
Introduction
Multicolor imaging is widely used in cell biology to
study the localization of specific biomolecules and track
their dynamics. In particular, visualization of ligandꢀ
receptorꢀadaptor interactions in response to ligand
stimulation is essential for investigating signal transducꢀ
tion in living cells. Recently, analysis of signaling pathꢀ
ways (e.g., that of epidermal growth factor receptor
(EGFR)1, 2) by multicolor imaging revealed the receptor
dynamics in response to ligand stimulation at singleꢀ
molecule resolution3. Singleꢀmolecule imaging (SMI)
enables tracking of hundreds or thousands of individual
molecules over a period of several seconds, and provides
direct information about molecular dynamics and interꢀ
actions in the cellular context, and consequently, about
dynamic and kinetic parameters of biological reactions
in living cells4–9. Multicolor SMI in particular, would be
a useful method for visualizing intermolecular interacꢀ
tions or analyzing molecular dynamics in living cells.
Over the past decade, proteinꢀlabeling approaches
with organic dyes have been developed. Particularly,
enzymeꢀbased ‘selfꢀlabeling tags’ (e.g., HaloTag11,
SNAPꢀtag12, and ACPꢀtag13) have been extensively studꢀ
ACS Paragon Plus Environment

Journal of the American Chemical Society
ied and applied in advanced imaging experiments, such
Page 2 of 14
Results
as superꢀresolution microscopy14. Recently, we develꢀ
oped a proteinꢀlabeling method named as the BLꢀtag
system, which is based on a mutant TEMꢀ1 βꢀlactamase.
This system enables covalent modification of βꢀlactam
derivatives with a functional molecule, such as a fluoroꢀ
phore15–20. One advantage of the BLꢀtag is that memꢀ
braneꢀpermeable probes can be synthesized19, 20. These
proteinꢀtagging approaches are superior to FPs because
the density of dyeꢀlabeled proteins can be simply conꢀ
trolled by adjusting the probe concentration and are useꢀ
ful for biomolecular tracking of membrane proteins at
1
2
3
4
5
6
7
8
Design and synthesis of SiRꢀbased labeling probes.
To develop NIRꢀfluorescent probes for cell imaging
based on the BLꢀtag, we selected SiRꢀMe and SiRꢀ
carboxyl as the NIR fluorophores, which possess excelꢀ
lent brightness, photostability, and cell permeability29, 30,
33, and bacampicillin as the BLꢀtag ligand for intracelluꢀ
lar protein labeling19. The membraneꢀpermeable SiRB
and SiRcB probes were synthesized by conjugating
bacampicillin with SiRꢀMe and SiRꢀcarboxyl (Figure 1a,
Schemes S1, S6). To evaluate labeling selectivity, fluoꢀ
rescence imaging with SiRB and SiRcB was performed
on HEK293T cells expressing BLꢀNLS, which is a fuꢀ
sion construct between the BLꢀtag and nuclear localizaꢀ
tion signal (NLS). BLꢀNLS was specifically labeled with
SiRcB (Figure S1a), whereas strong NIRꢀfluorescence
signals from SiRB were detected in both the nucleus and
the cytosol (Figure S3a), as also reported in a previous
paper29. This result indicates that SiRcB has sufficient
cell permeability and limited nonꢀspecific intracellular
accumulation for imaging of intracellular proteins.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
singleꢀmolecule resolution in living cells21–26
.
To further expand the SMI applications, development
of cellꢀpermeable proteinꢀlabeling probes with longerꢀ
wavelength excitation are required. Nearꢀinfrared (NIR)
fluorophores have many advantages in biological appliꢀ
cations such as their low phototoxicity, low spectral
overlap with cellular autofluorescence, and good tissue
penetration27, 28. NIRꢀemitting proteinꢀlabeling probes
can be used in combination with redꢀfluorescent probes
for multicolor SMI. Although various synthetic NIRꢀ
fluorophores with exceptional photostability and brightꢀ
ness have been reported, they tend to show nonꢀspecific
adsorption to cellular components because of their high
lipophilicity29. Thus, proteinꢀlabeling probes with NIR
dyes show offꢀtarget fluorescent spots in SMI applicaꢀ
tions and significantly interfere with the analysis of proꢀ
tein dynamics, despite the use of the protein tag system
like a SNAPꢀtag25. Introduction of a negatively charged
sulfonate group can increase dye hydrophilicity and deꢀ
crease nonꢀspecific adsorption30, but results in memꢀ
brane impermeability31, making it impossible to label
intracellular proteins. Recently, photostable Siꢀ
rhodamine (SiR) fluorophores were developed and apꢀ
To evaluate its capabilities for intracellular SMI, we
used SiRcB for SMI of the BLꢀtag anchored to the inner
leaflet of the plasma membrane. A fusion construct of
the BLꢀtag and the Lyn Nꢀterminus sequence (Lyn11ꢀBL)
was expressed in HEK293T cells, and SMI was perꢀ
formed by total internal reflection fluorescence microsꢀ
copy (TIRFM). However, following cell incubation with
SiRcB and washing, many bright spots were observed in
both Lyn11ꢀBL expressing and nonꢀtransfected cells
(Figure S5a, b). The spots were blinking (Movie S1),
probably because the carboxyl group of the dye was in
equilibrium with the nonꢀfluorescent spirolactone form.
Therefore, although SiRcB could specifically label the
target proteins, it was inadequate for SMI without furꢀ
ther modifications.
plied in liveꢀcell and superꢀresolution imaging29, 30, 32–37
.
However, reducing the high hydrophobicity of these
probes was still necessary to enhance the signalꢀtoꢀ
background (S/B) ratio for SMI in living cells. Thus,
analysis of proteinꢀprotein interactions and simultaneous
monitoring of the dynamics of two different proteins in
living cells remains challenging.
Improved design and synthesis of SiRcꢀbased labelꢀ
ing probes for intracellular SMI. Fluorescence signals
from SiRcB were specifically found in the cell nucleus,
where BLꢀNLS was expressed, because SiRcB remained
in its fluorescent zwitterionic form after covalent bindꢀ
ing to the BLꢀtag. By contrast, the free dye tended to
aggregate, resulting in unspecific binding to hydrophoꢀ
bic substances and the formation of nonꢀfluorescent spiꢀ
rolactone29. However, despite the nonꢀfluorescent nature
of SiRcB in hydrophobic environments, the observed
SiRcB fluorescent signals were sufficiently strong to
limit protein dynamics monitoring at the singleꢀmolecule
level by SMI. The offꢀtarget signals might be caused by
nonꢀspecific adsorption onto the glass surface or endogꢀ
enous proteins due to the hydrophobicity of the probe.
To increase hydrophilicity, sulfonate groups can be inꢀ
troduced into the fluorophore structure30, 31. However,
their negative charges cause electrostatic repulsion beꢀ
tween the fluorophores and the cell membrane.
Here, we developed NIR fluorescent probes for intraꢀ
cellular protein labeling and applied to multicolor SMI.
Control of the hydrophilicity by introducing a nonꢀ
charged oligoethylenoxy linker remarkably enhanced the
labeling S/B ratio to achieve SMI of intracellular proꢀ
teins. As results, our probe design enabled dual color
SMI of intracellular proteins in living cells without reꢀ
quiring optimization of the expression levels of two difꢀ
ferent proteins. Using our multicolor SMI platform, we
performed quantitative timeꢀlapse analysis of intracelluꢀ
lar protein dynamics at singleꢀmolecule resolution in the
context of innate immune system receptorꢀadaptor interꢀ
actions.
ACS Paragon Plus Environment

Page 3 of 14
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Design and fluorescence spectra of novel cell membrane permeable probes. (a) Structures of SiRꢀMe, SiRꢀcarboxyl, SiRB,
and SiRcB. (b) Structures of bacampicillinꢀbased probes. SiRcB2, SiRcB4, and SiRcB6 contain diꢀ, tetraꢀ, and hexaꢀethylene glycol, reꢀ
spectively. (c) The emission spectra of SiRcB, SiRcB2, SiRcB4, and SiRcB6 in PBS buffer (pH 7.4) containing 1.0% DMSO at 37 °C. The
probe concentration was 500 nM, and the excitation wavelength 654 nm. (d) Specific labeling of BLꢀNLS expressed in HEK293T cells.
The cells were incubated with 1.0 ꢀM SiRcB, SiRcB2, SiRcB4, or SiRcB6 for 30 min at 37 °C. Scale bar: 10 ꢀm.
To overcome this problem, we designed novel bacampiꢀ
cillinꢀbased labeling probes by introducing a hydrophilic
oligoethylenoxy linker, increasing hydrophilicity and the
distance between the fluorescent dye and the BLꢀtag
ligand in a stepwise manner: SiRcB(n) (n = 2, 4, 6)
(Figure 1b, Table 1). Moreover, since oligoethylenoxy
groups are both electroneutral and hydrophilic, the inꢀ
creased hydrophilicity of the probes was expected to
preclude their accumulation onto hydrophobic cellular
components, while maintaining their cell permeability.
These probes were synthesized by conjugation of
bacampicillin with oligoethylenoxy linkers, followed by
condensation with Siꢀcarboxyl (Schemes S3–6). Altꢀ
hough the optical properties of the open zwitterionic
forms of SiRcB(n) (n = 2, 4, 6) were nearly the same as
those of SiRcB (Table 1), the emission spectra in aqueꢀ
ous buffer showed fluorescence enhancement with inꢀ
creasing oligoethylenoxy linker length (Figure 1c). This
result indicates that the hydrophilic linker maintains the
fluorescent structure of the SiR dye by preventing probe
aggregation. As shown in Figure 1d, the SiRcB(n)
probes can label nuclear BLꢀtag. To evaluate labeling
selectivity, the S/B ratio was determined from the ratio
of nuclear to cytosolic fluorescence intensities. Howevꢀ
er, the S/B ratio of the SiRcB(n) probes was similar to
that of SiRcB probes, possibly because unreacted SiRcB
can form nonꢀfluorescent spirolactone, decreasing backꢀ
ground signals (Figure S1). Next, we compared the laꢀ
beling rates of intracellular (BLꢀNLS) or cell surface
(BLꢀGPI) BLꢀtag fusion proteins by the SiRcꢀbased
probes to evaluate their cell permeability (Supplemenꢀ
tary result, Figure S2). SiRcB2, SiRcB4, and SiRcB6
labeling rates were largely increased compared with the
SiRcB labeling rate. Additionally, the relative intensity
of SiRcB6 was lower than that of SiRcB4, especially for
BLꢀNLS labeling (Figure S2c, d), suggesting that inꢀ
creasing the probe hydrophilicity decreased its cell perꢀ
meability.
To determine whether SiRcꢀbased probes can be used
for SMI of intracellular proteins, we performed SMI of
Lyn11ꢀBL using three of the novel probes. Labeling with
SiRcB2, SiRcB4, and SiRcB6, unlike with SiRcB, reꢀ
sulted in numerous fluorescent spots on the images of
cells expressing Lyn11ꢀBL (Figure S5c, e, g), and few
spots in nonꢀtransfected cells (Figure S5d, f, h). The
bright spots in cells expressing Lyn11ꢀBL were mobile
(Movie S2), indicating diffusion of the labeled proteins
on the inner side of the plasma membrane. Tracking
analysis revealed that a significantly larger number of
fluorescent spots with higher diffusion coefficients was
detected upon labeling with SiRcB(n) (n = 2, 4, 6) than
with SiRcB (Figure 2a). These results clearly indicate
that introduction of the linker improved proteinꢀlabeling
efficiency and remarkably reduced background signals
due
to
nonꢀspecific
probe
adsorption.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 14
1
2
3
4
5
6
7
8
a CLogP
(open form) (closed form)
a CLogP
Linker length
(Å)
b λabs
(nm)
b λem
(nm)
b ε
probes
SiRcB
b Φfl
0.39
0.38
0.39
0.37
0.39
0.22
(Mꢀ1 cmꢀ1)
1.05
0.27
−0.08
−0.43
−0.88
−
7.98
7.20
6.85
6.45
5.86
−
0
9.6
16.7
23.8
−
654
651
651
651
645
554
671
670
670
670
661
579
100,000
100,000
100,000
100,000
100,000
70,000
SiRcB2
SiRcB4
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
SiRcB6
c SiRꢀcarboxyl
TMRꢀHaloTag ligand
−
Table 1. Physicochemical parameters and spectral properties of SiRcꢀbased probes and TMRꢀHaloTag ligand
a calculated by ChemDraw 15 (PerkinElmer), b measured in PBS containing 0.1% sodium dodecyl sulfate (SDS) (for SiRcB and SiRcB(n)
(n = 2, 4, 6) as open zwitterionic forms), c optical properties are described in reference 29
proteins of the innate immune system in living cells.
Tollꢀlike receptor 4 (TLR4)38 is a crucial sensor of lipoꢀ
polysaccharide (LPS), which is the major constituent of
the outer membrane of gramꢀnegative bacteria, and is
one of the pathogenꢀassociated molecular patterns
(PAMPs)38. The recognition of LPS by TLR4 requires
several other proteins, including LPS binding protein
(LBP), CD14, and MD2 (Figure 3a). TLR4 utilizes Tollꢀ
interleukinꢀ1 receptor (TIR) domainꢀcontaining adaptor
protein (TIRAP)39 and myeloid differentiation factorꢀ88
(MyD88) to activate the NFꢀκB pathway, and TIR doꢀ
Figure 2. Evaluation of nonꢀspecific adsorption of SiRcB and
SiRcB(n) (n = 2, 4, 6), and comparison of SiRcB4 and SiRcB6
labeling efficiency by TIRFM. (a) Distribution of the diffusion
the interferon regulatory transcription factorꢀ3 (IRF3)
coefficients of Lyn11ꢀBL labeled with 1.0 nM SiRcB and
SiRcB(n). The data shown are summed over three cells. (b) Comꢀ
parison of the number of bright spots per unit area between the
cells labeled with SiRcB4 and SiRcB6. Lyn11ꢀBL expressing
cells were incubated with 50 pM, 100 pM, 500 pM, and 1.0 nM
SiRcB4 or SiRcB6 and observed by TIRFM. The data shown are
mainꢀcontaining adaptor protein inducing IFN−β (TRIF)
and TRIFꢀrelated adaptor molecule (TRAM) to activate
pathway and induce the production of the appropriate
proꢀinflammatory cytokines38.
Here, we focused on the receptorꢀadaptor interaction
between TLR4 and TIRAP upon LPS stimulation to anaꢀ
lyze their molecular dynamics in living cells. TIRAP
was selected as the target adaptor protein because it loꢀ
calizes in the plasma membrane, where it interacts with
means ± standard error of the mean (S.E.M.) (N = 5).
TLR440. Thus, the dynamic interactions between TLR4
The number of fluorescent spots with lower diffusion
coefficients was slightly increased in SiRcB2ꢀlabeled
samples than in SiRcB4ꢀ and SiRcB6ꢀlabeled samples
(Figure 2a), indicating that the bright spots reflected
nonꢀspecific adsorption (Figure S5d). Moreover, when
the cells were incubated with subnanomolar concentraꢀ
tions of SiRcB6, the number of bright spots was lower
than that in SiRcB4ꢀlabeled cells (Figure 2b, Figure S5i,
j), indicating decreased membrane permeability. Based
on these results, SiRcB4 was the most suitable of the
four SiRc derivatives for intracellular protein SMI.
and TIRAP occur at the appropriate depth for SMI using
TIRFM. To analyze the TLR4ꢀTIRAP dynamics, the BLꢀ
tag proteinꢀlabeling system was used in combination
with a commercial HaloTag system. BLꢀtag and Haloꢀ
Tag were attached to the Nꢀterminus of TIRAP (BLꢀ
TIRAP) and the Cꢀterminus of TLR4 (TLR4ꢀHalo), reꢀ
spectively. To test the function of these fusion proteins,
we transfected HEK293T cells with BLꢀTIRAP or Heꢀ
magglutinin (HA)ꢀtagged TIRAP (HAꢀTIRAP), along
with a luciferase reporter for NFꢀκB. Overexpression of
BLꢀTIRAP or HAꢀTIRAP increased luciferase activity at
a comparable level, suggesting that fusion with BLꢀtag
did not perturb the capacity of TIRAP to activate NFꢀκB
(Figure S7a).
Analysis of receptorꢀadaptor interactions in reꢀ
sponse to ligand stimulation by confocal microscopy.
We performed multicolor imaging using the BLꢀtag and
SiRcB4 labeling system to study interactions between
ACS Paragon Plus Environment

Page 5 of 14
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Mechanism of multicolor imaging and computational analysis of SMI results. (a) Overview of LPS/TLR4 signaling. (b)
Imaging of HEK293T cells expressing TLR4ꢀHalo and BLꢀTIRAP by confocal microscopy. The cells were incubated with 100 nM Haloꢀ
TagꢀTMR Ligand and 500 nM SiRcB4 at 37 ºC for 30 min. Images were acquired before and after 10 min stimulation with 10 ꢀg/mL of
LPS. Scale bar: 20 ꢀm. (c) Multicolor SMI of HEK293T cells expressing TLR4ꢀHalo and BLꢀTIRAP. The cells were incubated with 50
pM HaloTagꢀTMR Ligand and 100 pM SiRcB4 for 30 min at 37 °C. Scale bar: 5 ꢀm. (d, e) Fluorescent labeling of TLR4 and TIRAP (d),
and TLR4(∆TIR)ꢀHalo and BLꢀTIRAP (e) with the HaloTagꢀTMR ligand and SiRcB4, respectively. (f–i) Transient response of the diffuꢀ
sion of TLR4ꢀHalo (f), and BLꢀTIRAP (g) in TLR4ꢀHalo and BLꢀTIRAP expressing cells, and TLR4(∆TIR)ꢀHalo (h), and BLꢀTIRAP (i)
in TLR4(∆TIR)ꢀHalo and BLꢀTIRAP expressing cells, upon LPS stimulation (10 ꢀg/mL).
N > 140,000 (5 cells).
Similarly, TLR4ꢀHalo and FLAGꢀtagged TLR4 (FLAGꢀ
TLR4) increased luciferase activity to a similar degree in
response to LPS stimulation, suggesting no effect of the
HaloTag on the TLR4 function as an LPS receptor (Figꢀ
ure S7b).
images were acquired before and after LPS stimulation.
Confocal microscopic images showed the localization of
TLR4 in plasma membranes and endoplasmic reticulum
and TIRAP in plasma membranes (Figure 3b), which are
consistent with those of previous studies40, 41. This result
confirms that the tags did not influence protein localizaꢀ
tion. However, localization of the fluorescence protein
fusions was not altered, even after ligand stimulation.
Thus, it is not possible to monitor TLR4ꢀTIRAP dynamꢀ
ics by conventional microscopy.
Then, the receptorꢀadaptor molecular dynamics inꢀ
duced by LPS stimulation were analyzed in living cells.
HEK293T cells expressing both target proteins were
labeled with SiRcB4 and HaloTagꢀTMR ligand, and
stimulated with 10 ꢀg/mL LPS for 10 min. Fluorescence
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 14
of TLR4 (TLR4(∆TIR)ꢀHalo). Since the TIR domain of
Live cell multicolor SMI of TLR4 and TIRAP, or
TLR4(∆TIR) and TIRAP, and statistical analysis of
protein dynamics. We used SiRcB4 and the HaloTagꢀ
TMR ligand to perform multicolor SMI in living cells
expressing BLꢀTIRAP and TLR4ꢀHalo, after stimulation
with 10 ꢀg/mL LPS, and quantitatively analyzed the
protein dynamics by statistical analysis. The dynamics
of fluorescentlyꢀlabeled TLR4ꢀHalo and BLꢀTIRAP
were successfully captured with an EMꢀCCD camera
(Figure 3c, Movies S3–6), demonstrating the capacity of
our proteinꢀlabeling system for multicolor SMI of intraꢀ
cellular proteins in living cells, using cellꢀpermeable
fluorescent probes (BLꢀtag and HaloTag probes). From
the statistical analysis of the intensities of each fluoresꢀ
cent spot, we verified that the detected mobile fluoresꢀ
cent spots originated from a single spot labeled with
SiRcB4 or HaloTagꢀTMR ligand (Figure S9).
TLR4 is necessary for TLR4ꢀTIRAP interactions39, we
expected that the TIRAP diffusion coefficients would
not change after LPS stimulation. Confocal imaging of
HEK293T cells transiently expressing BLꢀTIRAP and
TLR4(∆TIR)ꢀHalo showed that the localizations of the
proteins did not change regardless of LPSꢀstimulation
(Figure S8b). TIRFM SMI was then performed followꢀ
ing incubation with SiRcB4 and the HaloTag TMR ligꢀ
and (Figure S10). Singleꢀmolecule fluorescence spots
from TLR4(∆TIR) and TIRAP were captured, and the
data were statistically analyzed as described above (Taꢀ
ble S1). The results reveal that the relative ratio of slowꢀ
moving TIRAP molecules did not change even after ligꢀ
and stimulation (Figure 3i), indicating that since TIRAP
did not interact with TLR4(∆TIR) (Figure 3e), binding
of TIRAP to TLR4 through the TIR domain is essential
for changing its mobility. On the other hand, the populaꢀ
tion of slowꢀdiffusing TLR4(∆TIR) molecules still inꢀ
creased after stimulation (Figure 3h), indicating that the
change of TLR4 mobility upon LPS stimulation does not
require the TIR domain or interactions with TIRAP or
other intracellular molecules, and binding to LPS or othꢀ
er extracellular components, such as MDꢀ2 and CD14, is
sufficient.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Lateral molecular mobility is of particular importance
for investigating intracellular signal transduction mechꢀ
anisms because, in certain cases, it can be correlated
with molecular states and environmental conditions. For
example, intermolecular interactions are thought to deꢀ
crease diffusive mobility within the membrane42. Indeed,
altered mobility was noted for some membrane recepꢀ
tors, such as EGFR, after ligand stimulation and subseꢀ
quent binding to downstream signaling molecules, such
as Grb21, 43. Furthermore, intracellular signaling moleꢀ
cules exist in multiple states with different diffusion coꢀ
efficients44, 45. Here, we divided the molecular states into
two fractions (slow or fast state), and quantified the inꢀ
fluence of LPS stimulation for different time periods on
their ratio. The lateral mobility of diffusive molecules
was analyzed with a probability density function (PDF)
described in Eq. (1)46. Based on the PDF for a twoꢀstate
model, we estimated the ratio between states and calcuꢀ
lated the diffusion coefficients of each state, where state
1 and state 2 represent groups of fastꢀmoving and slowꢀ
moving molecules, respectively. Then, the rate of inꢀ
crease of the population of molecules in state 2 was calꢀ
culated using Eq. (2), and summarized as the time
course Rn min, which is defined as the ratio between slowꢀ
moving molecules after and before stimulation (Figure
3f, g, Table S1). As shown in Figure 3f, the TLR4 popuꢀ
lation with small diffusion coefficients increased upon
LPS stimulation (N > 140,000; 5 cells). Similarly, the
TIRAP population with slower diffusion rates increased
only after LPS stimulation. These results suggest that
LPS stimulation alters TLR4 and TIRAP mobility. Furꢀ
thermore, the diffusion rate of the slow TIRAP moleꢀ
cules was comparable to that of the slow TLR4 moleꢀ
cules (Table S1), implying interactions between TLR4
and TIRAP, which lead to decreased diffusion constants
upon LPS stimulation.
Discussion
SMI often suffers from offꢀtarget signals caused by
cellular autofluorescence and nonꢀspecific adsorption of
the dyes, even when imaging membrane proteins by
TIRFM25. Our designed probe with NIR fluorescence
overcame these limitations by utilizing an NIRꢀemitting
fluorescent dye and chemical modification to decrease
probe hydrophobicity. Introduction of an oligoethꢀ
ylenoxy group linker improved the hydrophilicity and
suppressed nonꢀspecific adsorption of the probes while
retaining cell permeability. The result of SMI showed
that labeling with SiRcB(n) (n = 2, 4, 6) remarkably reꢀ
duced the background signals and blinking observed for
SiRcB (Figure 2a; Figure S5, Movies S1, S2). In addiꢀ
tion, fineꢀtuning of the linker length is essential because
the high hydrophilicity of the probe interferes with celluꢀ
lar uptake, resulting in decreased labeling efficiency of
intracellular proteins (Figure 2b, Figure S2c, d). Thus,
stepwise control of probe hydrophilicity allows for effiꢀ
cient labeling of intracellular proteins with a remarkably
high S/B ratio. Our approach is critical for protein labelꢀ
ing with probes using protein tags in SMI. Moreover,
this method can be applied to fluorescent probes for use
in superresolution imaging based on singleꢀmolecule
localization microscopy.
Multicolor SMI is a useful method for visualizing inꢀ
termolecular interactions and analyzing molecular dyꢀ
namics in living cells. However, FPꢀbased labeling is not
optimal for SMI because of the necessity to control laꢀ
beling density by controlling FP expression levels; thus,
To confirm that the population increase in slowꢀ
moving TLR4 and TIRAP molecules after LPS stimulaꢀ
tion was due to TLR4ꢀTIRAP interactions, multicolor
SMI was performed using a TIRꢀdomain deletion mutant
ACS Paragon Plus Environment

Page 7 of 14
Journal of the American Chemical Society
dualꢀcolor SMI using two FPs with similar fluorescence other adaptor or downstream signalling molecules such
1
2
3
4
5
6
7
8
intensity is challenging, often making NIRꢀFPs impracꢀ
tical for use in SMI of intracellular proteins. Thus, we
used two orthogonal protein tag systems, the BLꢀtag and
HaloTag systems, with NIRꢀfluorescent probes for mulꢀ
ticolor SMI. In contrast to FPꢀbased systems, labeling
density can be easily optimized with two different probe
concentrations in the picomolar range (see Figure 2b,
Figure S5). Because these proteinꢀlabeling probes have
the advantage of efficient plasma membrane permeabilꢀ
ity, the experimental doses can be reduced to low conꢀ
centrations for labeling of intracellular proteins. The
combination of BLꢀtag and HaloTag probes with low
spectral overlap of highly photostable fluorophores enaꢀ
bled dualꢀcolor SMI of two intracellular proteins and
timeꢀlapse analysis of protein diffusion within the same
liveꢀcell context.
as MyD88.
Conclusion
In conclusion, we developed novel NIRꢀfluorescent
probes for intracellular protein labeling using the BLꢀtag
system and applied these probes in multicolor SMI.
SiRcB4 showed superior properties for labeling of intraꢀ
cellular proteins and an excellent S/B ratio, even at the
singleꢀmolecule level. Our multicolor SMI system enaꢀ
bled quantitative analysis of LPSꢀinduced TLR4ꢀTIRAP
interactions in living cells at the singleꢀmolecule level
for the first time. Our system may also be useful for
quantitative analysis of other receptors in response to
various PAMPs and investigation of the immune reꢀ
sponse of downstream adaptors localized at the cytoꢀ
plasmic membrane, such as TRAM48. Furthermore, this
probe can be used to study other intracellular proteinꢀ
protein interactions involved in signal transduction by
wideꢀfield fluorescence microscopy. Our intracellular
proteinꢀlabeling platform may be valuable for use in
screening studies based on quantitative singleꢀmolecule
analysis of molecular dynamics.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Analysis of protein dynamics based on multicolor SMI
showed an increase in the population of slowꢀmoving
TLR4 and TIRAP upon ligand stimulation by LPS (Figꢀ
ure 3f). The mobility of both proteins was decreased by
their interaction. This result is consistent with the curꢀ
rent model of TLR4 dynamics in response to ligand
stimulation40, which demonstrates the reliability of our
imaging system for detecting proteinꢀprotein interacꢀ
tions. In this study, we adopted a method for estimating
molecular diffusion from the position change of the
molecules by calculating the momentary travel distancꢀ
es. Coꢀlocalization analysis by dualꢀcolor simultaneous
measurement might be a more direct way to estimate the
binding rate between the molecule pair of interest at the
molecular level47. However, there are difficulties in coꢀ
localization analysis, such as aligning color channels or
a low possibility of events in which a molecule stained
by a dye interacts with another molecule also stained by
a dye because of partial staining of molecules. Thus, it is
difficult to distinguish intermolecular interactions from
only coꢀlocalization in a small space with a diameter of a
few hundred nanometers. Our method can reveal the
changes in molecular mobility upon the receptorꢀadaptor
interaction from the sequential image acquisition of difꢀ
ferent two colors even without two color simultaneous
measurement to detect molecular interaction directly
using molecular tracking.
Materials and Methods
Materials and analytical instruments. Unless othꢀ
erwise specified, all reagents were purchased from the
chemical suppliers Tokyo Chemical Industries (Tokyo,
Japan), Wako Pure Chemical (Osaka, Japan), or Sigmaꢀ
Aldrich Chemical Co. (St. Louis, MO, USA), and used
without further purification. Bacampicillin hydrochloꢀ
ride was a gift from NichiꢀIko Pharmaceutical Co. Ltd.
HaloTag TMR Ligand was purchased from Promega.
All labeling probes were dissolved in dimethyl sulfoxide
(DMSO) (biochemical grade; Wako Pure Chemical).
LPS from Salmonella minnesota Re595 was purchased
from Wako Pure Chemical (121ꢀ05291). Analytical thinꢀ
layer chromatography was performed on 60F254 silica
plates (Merck & Co., Inc., Kenilworth, NJ, USA) and
visualized under UV light. Silica gel column chromatogꢀ
raphy was performed using BWꢀ300 (Fuji Silysia Chemꢀ
ical Ltd., Kasugai, Japan). Nuclear magnetic resonance
(NMR) spectra were recorded on a JNMꢀAL400 instruꢀ
1
Statistical analysis showed that regardless of whether
the TIR domain was present or not, TLR4 diffusion
slowed down following stimulation, although the diffuꢀ
sion rate of TIRAP was dependent on the presence of the
TIR domain (Figure 3f–i, Table S1). These results indiꢀ
cate that TLR4 immobilization in response to LPS stimꢀ
ulation depends not only on interactions with TIRAP,
but also on clustering with MD2 and CD14. However,
the increased population of TLR4 with small diffusion
coefficients was more prominent than that of
TLR4(∆TIR), suggesting that the decrease in TLR4 difꢀ
fusion was enhanced by interactions with TIRAP, and
TIRAP immobilization may activate the interaction with
ment (JEOL, Tokyo, Japan) at 400 MHz for H NMR
and at 100 MHz for 13C NMR, or an AVANCE500HD
instrument (Bruker, Billerica, MA, USA) at 500 MHz
for 1H NMR and at 125 MHz for 13C NMR, using tetraꢀ
methylsilane as an internal standard. Mass spectra were
measured on a JMSꢀ700 mass spectrometer (JEOL) for
FAB and a LCTꢀPremier XE mass spectrometer (Waꢀ
ters, Milford, MA, USA), for electrospray ionization
(ESI). Highꢀperformance liquid chromatography
(HPLC) analyses were performed on an Inertsil ODSꢀ3
column (4.6 × 250 mm; GL Sciences) using an HPLC
system composed of a pump (PUꢀ2080; JASCO) and a
detector (MDꢀ2010 and FPꢀ2020; JASCO). Preparative
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 14
HPLC was performed with an Inertsil ODSꢀ3 column SiRB, SiRcB, SiRcB2, SiRcB4, or SiRcB6 in DMEM
1
2
3
4
5
6
7
8
(10.0 × 250 mm; GL Sciences, Inc.) using an HPLC sysꢀ
tem with a pump (PUꢀ2087; JASCO) and a detector
(UVꢀ2075; JASCO). Buffer A was composed of 0.1%
formic acid in H2O; Buffer B consisted of 0.1% formic
acid in acetonitrile. Fluorescence spectra were measured
using a Hitachi F4500 spectrometer. Ultravioletꢀvisible
absorbance spectra were measured using a Shimadzu
UV1650PC spectrometer. The pKmclꢀBLꢀGPI plasmid
was a kind gift from Dr. Atsushi Miyawaki.
for 30 min in a CO2 incubator. The cells were then
washed three times with HBSS and fluorescence microꢀ
scopic images were captured using appropriate filter
sets. Labeled cells were observed soon after the washing
step.
Confocal microscopic imaging of TLR4ꢀHalo,
TLR4(ꢁTIR)ꢀHalo, and BLꢀTIRAP. HEK293T cells
maintained in a glassꢀbottomed dish with DMEM conꢀ
taining 10% FBS at 37 °C under 5% CO2 were transꢀ
fected with the plasmids encoding TLR4ꢀHalo (or
TLR4(ꢁTIR)ꢀHalo), BLꢀTIRAP and MD2 as described
above. After 5 h, the culture medium was replaced with
DMEM and the cells were incubated at 37 °C for 24 h.
Next, the cells were washed three times with HBSS and
incubated with 500 nM SiRcB4 or 100 nM HaloTag
TMR Ligand for 30 min in a CO2 incubator. The cells
were then washed three times with HBSS and microꢀ
scopic images were acquired. Fluorescence images of
the cells in HBSS were captured using appropriate filter
sets. Labeled cells were observed soon after the washing
step. The fluorescence images were captured in preꢀ
stimulated cells and the cells after 10 min of LPS stimuꢀ
lation.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis. The synthetic procedures and characterizaꢀ
tion of compounds including SiRcB and SiRcB(n) (n =
2, 4, 6) are described in the Supporting Information.
Cell culture. HEK293T cells were cultured in highꢀ
glucose Dulbecco’s modified Eagle medium (DMEM) +
Gluta MaxꢀI (Invitrogen) supplemented with 10% fetal
bovine serum (FBS; Gibco), penicillin (100 U/mL), and
streptomycin (100 ꢀg/mL) (Invitrogen). Cells were inꢀ
cubated at 37 °C in a humidified atmosphere of 5% CO2.
A subculture was performed every 2–3 days from subꢀ
confluent (< 80%) cultures using
a
trypsinꢀ
ethylenediaminetetraacetic acid solution (Invitrogen).
Transfection of plasmids was carried out in a glassꢀ
bottomed dish (Matsunami) or 6ꢀwell (TPP) plate using
Lipofectamine 3000 (Invitrogen) according to a standard
protocol.
Micro cover glass cleaning. The 25ꢀmm glass coꢀ
verslips (Matsunami) were extensively cleaned to reꢀ
move background fluorescence. First, they were sonicatꢀ
ed in a solution of 0.1 M KOH for 30 min. Next, they
were sonicated in ethanol for 30 min three times. Coꢀ
verslips were stored in ethanol until use.
Confocal fluorescence microscopy. Fluorescence miꢀ
croscopic imaging was recorded using a confocal fluoꢀ
rescence microscopic imaging system including a fluoꢀ
rescence microscope (IX71, Olympus), a cooled CCD
camera (Cool Snap HQ, Roper Scientific) or an EMCCD
(iXon3, ANDOR TECHNOLOGY), a confocal scanner
unit (CSUꢀX1, Yokogawa Electric Corporation). Multiꢀ
spectral LED light sauce (Spectra X light engine, Luꢀ
mencor) was used to excite fluorophores at 377 ± 25 nm
(for Hoechst33342), 488 ± 3 nm (for MitoTracker®
Green FM in Figure S3), 542 ± 13.5 nm (for LysoTrackꢀ
er® Red DNDꢀ99), and 643 ± 10 nm (for SiRA, SiRB,
SiRB2, SiRcB, SiRcB2, SiRcB4, and SiRcB6). The filꢀ
ter sets were BP377 ± 25/DM405/BA447 ± 30 (for
Hoechst33342), BP488 ± 3/DM488/BA520 ± 17.5 (for
Set up of TIRFM. A TIRF microscope (TiꢀE, Nikon)
equipped with an EMꢀCCD camera (ImagEM C9100ꢀ13,
Hamamatsu Photonics) and a 60× oilꢀimmersion objecꢀ
tive (Apo TIRF 60× oil, NA = 1.49; Nikon) was used.
Laser diodes (Sapphire, COHERENT) were used to exꢀ
cite HaloTag TMR Ligand and SiRꢀcarboxylꢀbased
probes (SiRcB, SiRcB2, SiRcB4, and SiRcB6) at 561
and 640 nm, respectively. Specimens were excited with
an evanescent light. The fluorescence signals from Halꢀ
oTag TMR Ligand and SiRꢀbased probes were imaged
using dichroic mirrors (DM; Di01ꢀR561 and Di02ꢀR635,
Semrock) and barrier filters (BF; FF01ꢀ609/54 and
676/29, Semrock), respectively. The laser power density
is ~3.91 × 103 W/cm2, which was calculated from the
field of illumination (256 ꢀm2) and laser power (~10
mW). Image sequences (300 frames) were acquired with
an exposure time of 32.82 ms. Frame rate was 30.48 fps.
Experiments were performed at 37 °C.
MitoTracker®
Green
FM),
BP560
±
12.5/DM561/BA624 ± 20 (for LysoTracker® Red DNDꢀ
99 in Figure S3), or BP628 ± 10/DM635/BA692 ± 20
(for SiRA, SiRB, SiRB2, SiRcB, SiRcB2, SiRcB4, and
SiRcB6). The lens was UPlanSApo (60×, N.A. 1.35, Oil,
Olympus). The whole system was controlled using Metꢀ
aMorph 7.6 software (Molecular Devices).
Confocal microscopic imaging of BLꢀNLS.
HEK293T cells maintained in glassꢀbottomed dishes in
DMEM containing 10% FBS at 37 °C under 5% CO2
were transfected with the pcDNA3.1 (+)ꢀBLꢀNLS plasꢀ
mid using Lipofectamine 3000. After 5 h, the culture
medium was replaced with DMEM and the cells were
incubated at 37 °C for 24 h. Next, the cells were washed
three times with HBSS and incubated with 500 nM
Singleꢀmolecule imaging of Lyn11ꢀBL with a TIRF
microscope. HEK293T cells maintained on glass coꢀ
verslips in a 6ꢀwell plate were transfected with Lyn11ꢀBL
using Lipofectamine 3000. After 5 h, the culture medium
was replaced with DMEM and the cells were incubated
at 37 °C for 24 h. The culture medium was exchanged
with OptiꢀMEM (Gibco), and the cells were incubated at
37 °C for 2 h. Next, the cells were washed three times
ACS Paragon Plus Environment

Page 9 of 14
Journal of the American Chemical Society
with HBSS and incubated with 1.0 nM SiRcB or point and determined the diffusion coefficients D1 and
1
2
3
4
5
6
7
8
SiRcB(n) in OptiꢀMEM for 30 min in a CO2 incubator.
The cells were then washed three times with HBSS and
singleꢀmolecule imaging was performed using a TIRF
microscope.
D2 at the first time. Then, we calculated the population
of the state 1 at each time point, pbefore or pn min, by usꢀ
ing the obtained D1 and D2. Finally, we calculated the
increasing rate of the population of molecules in the
state 2 by using Eq. (2), and summarized as the time
course of Rn min, which is the ratio of the slowꢀmoving
molecules relative to the value of preꢀstimulation moleꢀ
cules.
Multicolor SMI of TLR4ꢀHalo, TLR4(∆TIR)ꢀHalo,
and BLꢀTIRAP. HEK293T cells maintained on glass
coverslips in a 6ꢀwell plate were coꢀtransfected with
TLR4ꢀHalo, BLꢀTIRAP, and MD2 using Lipofectamine
3000. After 5 h, the culture medium was replaced with
DMEM, and the cells were incubated at 37 °C for 24 h.
Next, the culture medium was exchanged with Optiꢀ
MEM and the cells were incubated at 37 °C for 2 h. The
cells were then washed three times with HBSS and inꢀ
cubated with SiRcB4 (100 pM) and HaloTag TMR Ligꢀ
and (50 pM) in OptiꢀMEM for 30 min in a CO2 incubaꢀ
tor. The cells were stimulated with LPS (10 ꢀg/mL) conꢀ
taining OptiꢀMEM, or LPS nonꢀcontaining OptiꢀMEM,
after washed three times with HBSS. SMI was perꢀ
formed before stimulation and 1, 5, 10, 15, 20, 25, 30
min after the stimulation using a TIRFM.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ASSOCIATED CONTENT
Supporting Information. Materials and Methods contains Chemꢀ
ical Synthesis; Detection of Protein Labeling; Evaluation of
Membrane Permeability; Plasmid Construction; Luciferase Assay;
and Statistical Analysis. This material is available free of charge
via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
* K.K. (kkikuchi@mls.eng.osakaꢀu.ac.jp), M.U. (masaꢀ
hiroueda@fbs.osakaꢀu.ac.jp), and Y.K. (ykumagai@biken.osakaꢀ
u.ac.jp).
Statistical PDF analysis. To describe lateral mobility
with state transitions, we focused on molecules that
move by simple diffusion, irrespective of whether they
adopt multiple states or not. If a mobile molecule has
two states but no state transitions, the population of the
molecule can be analyzed as a mixture of two simple
diffusion processes. Assuming that the proportion of
molecules in state 1 (the group of the fastꢀmoving moleꢀ
cules) is p, the PDF is written by a summation of PDFs
for two simple diffusion processes as published previꢀ
ously46 and calculated as
Author Contributions
‡R.S., and J.K. contributed equally.
Funding Sources
This research was supported by MEXT, Japan (Grants 25220207,
26102529, 16H00768, to K.K., and 24115513, 15H03120,
15H00818 to S.M.).
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENT
The authors would like to thank to S. Teraguchi for discussion
and to M. Hata (RIKEN Quantitative Biology), A. Yoshimura, E.
Kurumatani, and Y. Kimura (Osaka University, IFReC) for assisꢀ
tance in cell culturing and support experiments.
ꢄ
ꢂꢋ
4 ꢇꢈꢉ + ꢊ
1 − ꢄ
ꢅ4ꢆꢁꢇꢋꢉ + ꢊꢋꢃ
ꢁ ꢃ
ꢀ ꢂ =
ꢌ ꢍ−
ꢎ
ꢅ4ꢆꢁꢇꢈꢉ + ꢊꢋꢃ
ꢁ
ꢋꢃ
ꢂꢋ
4 ꢇꢋꢉ + ꢊ
+
ꢌ ꢍ−
ꢎ
References
ꢁ
ꢋꢃ
(1) LowꢀNam, S. T.; Lidke, K. A.; Cutler, P. J.; Roovers, R. C.;
Henegouwen, P. M. V. B.; Wilson, B. S.; Lidke, D. S. Nat. Struct.
Mol. Biol. 2011, 18, 1244–1249.
(2) Chung, I.; Akita, R.; Vandlen, R.; Toomre, D.;
Schlessinger, J.; Mellman, I. Nature 2010, 464, 783–787.
(3) Murakoshi, H.; Iino, R.; Kobayashi, T.; Fujiwara, T.;
Ohshima, C.; Yoshimura, A.; Kusumi, A. Proc. Natl. Acad. Sci.
USA 2004, 101, 7317–7322.
(4) Weiss, S. Nat. Struct. Biol. 2000, 7, 724–729.
(5) Ishijima, A.; Yanagida, T. Trends Biochem. Sci. 2001, 26,
438–444.
(6) Xie, S. Single Mol. 2001, 2, 229–236.
(7) Hinterdorfer, P.; Schütz, G.; Kienberger, F; Schindler, H. J.
Biotechnol. 2001, 82, 25–35.
(1)
where the population, p, is constant with time, meanꢀ
ing that two independent populations of molecules exist
irrespective of time. Here, σ is noise term, and D1 and D2
are the calculated diffusion coefficients of molecules,
which belong to state 1 and the state 2, respectively.
The ratio of molecules in state 2 (the group of slowꢀ
moving molecules) at each time point relative to the valꢀ
ue of preꢀstimulation molecules, Rn min, were calculated
as
1 − ꢄꢐ ꢑꢒꢐ
(8) Sako, Y.; Uyemura, T. Cell Struct. Funct. 2002, 27, 357–
365.
ꢏꢐ ꢑꢒꢐ
=
1 − ꢄꢓꢔꢕꢖꢗꢔ
(9) Funatsu, T.; Harada, Y.; Tokunaga, M.; Saito, K.; Yanagꢀ
ida, T. Nature 1995, 374, 555–559.
(10) Xia, T.; Li, N.; Fang, X. Annu. Rev. Phys. Chem. 2013, 64,
459–480.
(2)
where pn min and pbefore are the value of p in each time
point.
(11) Los, G. V.; Encell, L. P.; McDougall, M. G.; Hartzell, D.
D.; Karassina, N.; Zimprich, C.; Wood, M. G.; Learish, R.;
When we perform PDF analysis, we simultaneously
analyzed all images that were captured at each time
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 14
Ohana, R. F.; Urh, M.; Simpson, D.; Mendez, J.; Zimmerman, K.;
(31) Pauff, S. M.; Miller, S. C. Org. Lett. 2011, 13, 6196–6199.
1
2
3
4
5
6
7
8
Otto, P.; Vidugiris, G.; Zhu, J.; Darzins, A.; Klaubert, D. H.; Bulꢀ
leit, R. F.; Wood, K. V. ACS Chem. Biol. 2008, 3, 373–382.
(12) Keppler, A.; Gendreizig, S.; Gronemeyer, T.; Pick, H.;
Vogel, H.; Johnsson, K. Nat. Biotechnol. 2002, 21, 86–89.
(13) George, N.; Pick, H.; Vogel, H.; Johnsson, N.; Johnsson,
K. J. Am. Chem. Soc. 2004, 126, 8896–8897.
(14) Jones, S. A.; Shim, S. H.; He, J.; Zhuang, X. Nat.Methods
2011, 8, 499–505.
(15) Mizukami, S.; Watanabe, S.; Hori, Y.; Kikuchi, K. J. Am.
Chem. Soc. 2009, 131, 5016–5017.
(32) Egawa, T.; Hanaoka, K.; Koide, Y.; Ujita, S.; Takahashi,
N.; Ikegaya, Y.; Matsuki, N.; Terai, T.; Ueno, T.; Komatsu, T.;
Nagano, T. J. Am. Chem. Soc. 2011, 133, 14157–14159.
(33) Koide, Y.; Urano, Y.; Hanaoka, K.; Terai, T.; Nagano, T.
ACS Chem. Biol. 2011, 6, 600–608.
(34) Kim, E.; Yang, K. S.; Giedt, R. J.; Weissleder, R. Chem.
Commun. 2014, 50, 4504–4507.
(35) Kim, S.; Tachikawa, T.; Fujitsuka, M.; Majima, T. J. Am.
Chem. Soc. 2014, 136, 11707–11715.
(36) Grimm, J. B.; English, B. P.; Chen, J.; Slaughter, J. P.;
Zhang, Z.; Revyakin, A.; Patel, R.; Macklin, J. J.; Normanno, D.;
Singer, R. H.; Lionnet, T.; Lavis, L. D. Nat. Methods 2015, 12,
244–250.
(37) Thompson, A. D.; Omar, M. H.; RiveraꢀMolina, F.; Xi, Z.;
Koleske, A. J.; Toomre, D. K.; Schepartz, A. Angew. Chem. Int.
Ed. 2017, 56, 10408–10412.
(38) Akira, S.; Uematsu, S.; Takeuchi, O. Cell 2006, 124, 783–
801.
(39) Horng, T.; Barton, G. M.; Medzhitov, R. Nat. Immunol.
2001, 2, 835–841.
(40) Kagan, J. C.; Medzhitov, R. Cell 2006, 125, 943–955.
(41) Visintin, A.; Latz, E.; Monks, B. G.; Espevik, T.; Golenꢀ
bock, D. T. J. Biol. Chem. 2003, 278, 48313–48320.
(42) Gambin, Y.; LopezꢀEsparza, R.; Reffay, M.; Sierecki, E.;
Gov, N. S.; Genest, M. H. R. S.; Hodges, R.S.; Urbach, W. Proc.
Natl. Acad. Sci. USA 2006, 103, 2098–2102.
(43) Morimatsu, M.; Takagi, H.; Ota, K. G.; Iwamoto, R.; Yanꢀ
agida, T.; Sako, Y. Proc. Natl. Acad. Sci. USA 2007, 104, 18013–
18018.
(44) Hibino, K.; Watanabe, T. M.; Kozuka, J.; Iwane, A. H.;
Okada, T.; Kataoka, T.; Yanagida, T.; Sako, Y. ChemPhysChem
2003, 4, 748–753.
(45) Matsuoka, S.; Shibata, T.; Ueda, M. PLoS Comp. Biol.
2013, 9, e1002862–e1002862.
(46) Matsuoka, S.; Shibata, T.; Ueda, M. Biophys. J. 2009, 97,
1115–1124.
(47) Kozuka, J.; Yokota, H.; Arai, Y.; Ishii, Y.; Yanagida, T.
Nat. Chem. Biol. 2006, 2, 83–86.
(48) Rowe, D. C.; McGettrick, A. F.; Latz, E.; Monks, B. G.;
Gay, N. J.; Yamamoto, M.; Akira, S.; O’Neill, L. A.; Fitzgerald,
K. A.; Golenbock, D. T. Proc. Natl. Acad. Sci. USA 2006, 103,
6299–6304.
9
(16) Watanabe, S.; Mizukami, S.; Hori, Y.; Kikuchi, K. Bioꢀ
conjugate Chem. 2010, 21, 2320–2326.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(17) Yoshimura, A.; Mizukami, S.; Hori, Y.; Watanabe, S.;
Kikuchi, K. ChemBioChem 2011, 12, 1031–1034.
(18) Mizukami, S.; Yamamoto, T.; Yoshimura, A.; Watanabe,
S.; Kikuchi, K. Angew. Chem. Int. Ed. 2011, 50, 8750–8752.
(19) Watanabe, S.; Mizukami, S.; Akimoto, Y.; Hori, Y.; Kikuꢀ
chi, K. Chem. Eur. J. 2011, 17, 8342–8349.
(20) Mizukami, S.; Watanabe, S.; Akimoto, Y.; Kikuchi, K. J.
Am. Chem. Soc. 2012, 134, 1623–1629.
(21) Calebiro, D.; Rieken, F.; Wagner, J.; Sungkaworn, T.;
Zabel, U.; Borzi, A.; Cocucci, E.; Zurn, A.; Lohse, M. J. Proc.
Natl. Acad. Sci. USA 2013, 110, 743–748.
(22) Jacquier, V.; Prummer, M.; Segura, J. M.; Pick, H.; Vogel,
H. Proc. Natl. Acad. Sci. USA 2006, 103, 14325–14330.
(23) Liu, D. S.; Phipps, W. S.; Loh, K. H.; Howarth, M.; Ting,
A. Y. ACS Nano 2012, 6, 11080–11087.
(24) Suzuki, K. G.; Kasai, R. S.; Hirosawa, K. M.; Nemoto, Y.
L.; Ishibashi, M.; Miwa, Y.; .Fujiwara, T.; Kusumi, A. Nat. Chem.
Biol. 2012, 8, 774–783.
(25) Bosch, P. J.; Correa, I. R.; Sonntag, M. H.; Ibach, J.;
Brunsveld, L.; Kanger, J. S.; Subramaniam, V. Biophys. J. 2014,
107, 803–814.
(26) Latty, S. L.; Felce, J. H.; Weimann, L.; Lee, S. F.; Davis,
S. J.; Klenerman, D. Biophys. J. 2015, 109, 1798–1806.
(27) Weissleder, R. Nat. Biotechnol. 2001, 19, 316–317.
(28) Schneckenburger, H.; Weber, P.; Wagner, M.; Schickingꢀ
er, S.; Richter, V.; Bruns, T.; Strauss, W. S. L.; Wittig, R. J. Miꢀ
crosc. 2012, 245, 311–318.
(29) Lukinavičius, G.; Umezawa, K.; Olivier, N.; Honigmann,
A.; Yang, G.; Plass, T.; Mueller, V.; Reymond, L.; Corrêa I. R.
Jr.; Luo, Z.ꢀG.; Schultz, C.; Lemke, E. A.; Heppenstall, P.;
Eggeling, C.; Manley, S.; Johnsson, K. Nat. Chem. 2013, 5, 132–
139.
(30) Shieh, P.; Dien, V. T.; Beahm, B. J.; Castellano, J. M.;
WyssꢀCoray, T.; Bertozzi, C. R. J. Am. Chem. Soc. 2015, 137,
7145–7151.
ACS Paragon Plus Environment

Page 11 of 14
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 14
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Design and fluorescence spectra of novel cell membrane permeable probes. (a) Structures of SiR-
Me, SiR-carboxyl, SiRB, and SiRcB. (b) Structures of bacampicillin-based probes. SiRcB2, SiRcB4, and
SiRcB6 contain di-, tetra-, and hexa-ethylene glycol, respectively. (c) The emission spectra of SiRcB,
SiRcB2, SiRcB4, and SiRcB6 in PBS buffer (pH 7.4) containing 1.0% DMSO at 37 °C. The probe
concentration was 500 nM, and the excitation wavelength 654 nm. (d) Specific labeling of BL-NLS expressed
in HEK293T cells. The cells were incubated with 1.0 µM SiRcB, SiRcB2, SiRcB4, or SiRcB6 for 30 min at 37
°C. Scale bar: 10 µm.
194x110mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 13 of 14
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Evaluation of non-specific adsorption of SiRcB and SiRcB(n) (n = 2, 4, 6), and comparison of
SiRcB4 and SiRcB6 labeling efficiency by TIRFM. (a) Distribution of the diffusion coefficients of Lyn11-BL
labeled with 1.0 nM SiRcB and SiRcB(n). The data shown are summed over three cells. (b) Comparison of
the number of bright spots per unit area between the cells labeled with SiRcB4 and SiRcB6. Lyn11-BL
expressing cells were incubated with 50 pM, 100 pM, 500 pM, and 1.0 nM SiRcB4 or SiRcB6 and observed by
TIRFM. The data shown are means ± standard error of the mean (S.E.M.) (N = 5).
185x102mm (300 x 300 DPI)
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 14
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Mechanism of multicolor imaging and computational analysis of SMI results. (a) Overview of
LPS/TLR4 signaling. (b) Imaging of HEK293T cells expressing TLR4ꢀHalo and BLꢀTIRAP by confocal
microscopy. The cells were incubated with 100 nM HaloTagꢀTMR Ligand and 500 nM SiRcB4 at 37 ºC for 30
min. Images were acquired before and after 10 min stimulation with 10 ꢁg/mL of LPS. Scale bar: 20 ꢁm. (c)
Multicolor SMI of HEK293T cells expressing TLR4ꢀHalo and BLꢀTIRAP. The cells were incubated with 50 pM
HaloꢀTagꢀTMR Ligand and 100 pM SiRcB4 for 30 min at 37 °C. Scale bar: 5 ꢁm. (d, e) Fluorescent labeling
of TLR4 and TIRAP (d), and TLR4(∆TIR)ꢀHalo and BLꢀTIRAP (e) with the HaloTagꢀTMR ligand and SiRcB4,
respectively. (f–i) Transient response of the diffusion of TLR4ꢀHalo (f), and BLꢀTIRAP (g) in TLR4ꢀHalo and
BLꢀTIRAP expressing cells, and TLR4(∆TIR)ꢀHalo (h), and BLꢀTIRAP (i) in TLR4(∆TIR)ꢀHalo and BLꢀTIRAP
expressing cells, upon LPS stimulation (10 ꢁg/mL). N > 140,000 (5 cells).
226x185mm (300 x 300 DPI)
ACS Paragon Plus Environment