Alkyne-tag Raman imaging for visualization of mobile small molecules in live cells

Article abstract of DOI:10.1021/ja308529n

Alkyne has a unique Raman band that does not overlap with Raman scattering from any endogenous molecule in live cells. Here, we show that alkyne-tag Raman imaging (ATRI) is a promising approach for visualizing nonimmobilized small molecules in live cells. An examination of structure-Raman shift/intensity relationships revealed that alkynes conjugated to an aromatic ring and/or to a second alkyne (conjugated diynes) have strong Raman signals in the cellular silent region and can be excellent tags. Using these design guidelines, we synthesized and imaged a series of alkyne-tagged coenzyme Q (CoQ) analogues in live cells. Cellular concentrations of diyne-tagged CoQ analogues could be semiquantitatively estimated. Finally, simultaneous imaging of two small molecules, 5-ethynyl-2′-deoxyuridine (EdU) and a CoQ analogue, with distinct Raman tags was demonstrated.

Full text of DOI:10.1021/ja308529n

Subscriber access provided by FORDHAM UNIVERSITY
Article
Alkyne-tag Raman imaging for visualization
of mobile small molecules in live cells
Hiroyuki Yamakoshi, Kosuke Dodo, Almar Flotildes Palonpon, Jun
Ando, Katsumasa Fujita, Satoshi Kawata, and Mikiko Sodeoka
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/ja308529n • Publication Date (Web): 01 Dec 2012
Downloaded from http://pubs.acs.org on December 10, 2012
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 11
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
Alkyne-tag Raman imaging for visualization of mobile small
molecules in live cells
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
Hiroyuki Yamakoshi,†,‡ Kosuke Dodo,†,‡ Almar Palonpon,†,§ Jun Ando,†,‡ Katsumasa Fujita,†,§ Satoshi
Kawata,§, ‡ and Mikiko Sodeoka*,†,‡
†Sodeoka Live Cell Chemistry Project, ERATO, JST, 2ꢀ1, Hirosawa, Wakoꢀshi, Saitama 351ꢀ0198, Japan
‡RIKEN Advanced Science Institute, 2ꢀ1, Hirosawa, Wakoꢀshi, Saitama 351ꢀ0198, Japan
§ Department of Applied Physics, and Osaka University, 2ꢀ1, Yamadaoka, Suita, Osaka, 565ꢀ0871, Japan
ABSTRACT: Alkyne has a unique Raman band that does not overlap with Raman scattering from any endogenous molecule in
live cells. Here, we show that alkyneꢀtag Raman imaging (ATRI) is a promising approach for visualizing nonꢀimmobilized small
molecules in live cells. An examination of structureꢀRaman shift/intensity relationships revealed that alkynes conjugated to an aroꢀ
matic ring and/or to a second alkyne (conjugated diynes) have strong Raman signals in the cellular silent region and can be excelꢀ
lent tags. Using these design guidelines, we synthesized and imaged a series of alkyneꢀtagged coenzyme Q (CoQ) analogues in live
cells. Cellular concentrations of diyneꢀtagged CoQ analogues could be semiꢀquantitatively estimated. Finally, simultaneous imaging
of two small molecules, 5ꢀethynylꢀ2’ꢀdeoxyuridine (EdU) and a CoQ analogue, with distinct Raman tags was demonstrated.
nous molecules show no Raman scattering. Thus, alkyne can
INTRODUCTION
be used as a tag to image target molecules by Raman microsꢀ
copy. Focusing on this idea, we developed alkyneꢀtag Raman
imaging (ATRI), which is expected to be applicable to a wide
range of molecules. As a proof of concept, we successfully
imaged an alkyneꢀtagged cell proliferation probe, EdU (5ꢀ
ethynylꢀ2’ꢀdeoxyuridine), in live cells by Raman microscopy3
(Figure 1b). EdU was covalently incorporated into DNA as a
mimic of thymidine, and its localization in the nucleus was
clearly visualized. We considered that ATRI should be availaꢀ
ble for imaging a wide range of mobile nonꢀcovalentꢀbondꢀ
forming small molecules, such as specific lipids and drug canꢀ
didate molecules, in live cells, as well as for estimation of the
cellular concentration of such molecules, and for multiꢀcolor
imaging of small molecules in live cells. For these purposes, a
basic knowledge of the relationship between structure and
Raman shift/intensity of alkynes is essential.
Imaging of molecules in live cells is an important tool in biolꢀ
ogy, chemical biology and pharmaceutical and medical sciencꢀ
es. In particular, the use of fluorescent labels has made possiꢀ
ble the sensitive detection of specific molecules in living cells.
For example, green fluorescent protein (GFP) and related proꢀ
teins have been widely used as labels for imaging various proꢀ
teins. However, the use of fluorescent tags is problematic for
bioactive small molecules, because sensitive fluorescent dyes,
such as derivatives of fluorescein, BODIPY and many others,
have a molecular weight comparable to or even larger than the
parent small molecules (Figure 1a). Consequently, fluorescent
tags may alter the biological activity, cellular localization and
dynamics of the parent small molecules.1 To address this issue,
click chemistry has recently been employed to introduce a
bioorthogonal azide or alkynyl group as a tag. However, a
fluorescent group is usually introduced after fixation of the
cells for detection of the bioorthogonal tag.
In this paper, we first describe an examination of the structureꢀ
Raman shift/intensity relationship of various alkynes. Based
on the results, we propose guidelines for designing appropriate
alkyneꢀtagged molecules for ATRI. Furthermore, we discuss
other candidate small Raman tags that exhibit Raman peaks in
the cellular silent region (Figure 1c). Using the proposed deꢀ
sign guidelines for alkyne tags, we synthesized a series of moꢀ
bile smallꢀmolecular alkyneꢀtagged coenzyme Q (CoQ) anaꢀ
logues, and examined their spatial localization in live cells by
Raman microscopy. By calibrating the alkyne Raman intensity
with concentration, the cellular concentration of the alkyneꢀ
tagged CoQ analogues was semiꢀquantitatively estimated.
On the other hand, Raman microscopy can visualize the localꢀ
ization of molecules without a fluorescent tag because of its
ability to detect molecular vibrations.2 Recent developments in
Raman microscopes have made it possible to obtain highꢀ
contrast Raman images of live cells in a reasonable time. But,
the observable molecular species have been limited to those
species existing in large amounts in cells, such as proteins and
lipids. Raman signals from small molecules tend to be weak,
and may easily be masked by intense overlapping Raman sigꢀ
nals from dominant intracellular species. However, the alkyne
moiety shows a distinct, strong Raman scattering peak in a
cellular silent region (1800 ~ 2800 cmꢀ1), where most endogeꢀ
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 11
1
2
3
4
5
6
7
8
9
slitꢀscanning excitation and detection optics. A 532 nm laser
(Verdi Vꢀ18, Coherent, Santa Clara, CA) was used for excitaꢀ
tion. The laser beam was shaped into a line by a series of cyꢀ
lindrical lenses and focused into the sample by a 60X/1.2 NA
water immersion objective lens (UPLSAPO 60XW, Olympus,
Tokyo, Japan). The backscattered Raman signals from the
illuminated line were collected by the same objective lens,
passed through several Raman edge filters (Semrock, Rochesꢀ
ter, NY) and imaged at the entrance slit of a dispersive specꢀ
trograph (Bunkoh Keiki, MKꢀ300, Tokyo, Japan). The Raman
signals were then dispersed by a grating and detected with a
cooled CCD camera (Pixis 400, Princeton Instruments, Trenꢀ
ton, NJ) to obtain the Raman spectra (600ꢀ3000 cmꢀ1) simultaꢀ
neously from multiple points in the line.
a)
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
To acquire a Raman image, the laser line was scanned in one
direction across the sample using a singleꢀaxis galvano mirror
(GSI Lumonics, Billerica, MA) with a step size of ~0.33 ꢀm,
which is larger than the lateral resolution (0.27 ꢀm) of our
setup. The slit width of the spectrograph was 40 ꢀm. The laser
intensity at the sample plane and exposure time of each line
are indicated in the figure captions. The laser intensity was
calculated by obtaining the ratio of the measured laser power
at the sample position and the area of the illumination line
(width = 0.61λ/NA, length was measured from the brightfield
image). All reported image acquisition times take into account
the spectral data transfer rate of the CCD camera to the PC,
which is about ~3 s/line.
b )
c)
O
N
Candidate Raman-tags
NH
R1
R2
C
C
R
N
C
HO
O
alkyne
nitrile
R1
O
R2
D
N
N N
C
OH
R
R3
deuterium
EdU
(5-ethynyl-2'-deoxyuridine)
azide
All Raman hyperspectral data sets were postꢀprocessed using
the singular value decomposition (SVD) technique for noise
reduction and a modified polyfit fluorescence technique for
removal of the autofluorescence baseline signal. Finally, the
Raman image was constructed by displaying the Raman intenꢀ
sity of the vibrational band of interest at each spatial position.
All data processing was performed using inꢀhouse image proꢀ
cessing software. To remove any nonresonant Raman contriꢀ
butions to the alkyne images, we subtracted the Raman image
at a neighboring nonresonant wave number from the Raman
image at the alkyne peak position.
Figure 1. Alkyne tag for specific detection of small molecules by
ATRI (alkyneꢀtag Raman imaging). (a) Comparison of fluorescent
tag and alkyne tag. (b) Structure of EdU. (c) Possible bioorthogoꢀ
nal functional groups showing a Raman peak in the cellular silent
region.
Finally, simultaneous multicolor imaging of two small moleꢀ
cules, EdU and a CoQ analogue, with distinct alkyne tags was
demonstrated. Our results indicate that ATRI would be a valuꢀ
able tool for the experimental study of a wide range of small
molecules in living cells.
All cell samples were grown on a quartz substrate. Loading
concentration of the Ramanꢀtagged molecules and incubation
times are indicated in the figure captions. Prior to Raman imꢀ
aging, the medium of the cell sample was replaced with a
HEPESꢀbuffered Tyrode solution composed of NaCl (150
mM), glucose (10 mM), HEPES (10 mM), KCl (4.0 mM),
MgCl2 (1.0 mM), CaCl2 (1.0 mM) and NaOH (4.0 mM).
EXPERIMENTAL SECTION
Cell culture. HeLa human cervical cancer cell line was culꢀ
tured in DMEM supplemented with 10% fetal bovine serum
(FBS) and antibiotics (penicillin / streptomycin).
Reagents. EdU (5ꢀethynylꢀ2’ꢀdeoxyuridine) was purchased
from Invitrogen.
Other methods. Experimental procedures for chemical synꢀ
thesis and compound data including NMR spectra are given in
the Supplementary information.
Raman spectra of compounds. Raman spectra of compounds
were obtained using a slitꢀscanning Raman microscope
(RAMANꢀ11; Nanophoton, Osaka, Japan) with excitation at
532 or 660 nm. Samples were installed on a glassꢀbottomed
dish (Matsunami, multiꢀwell glassꢀbottomed dish No. 1S). The
laser output was focused into the sample by a 60X/1.2 NA
water immersion objective lens (UPLSAPO 60XW, Olympus,
Tokyo, Japan). The slit width of the spectrograph was 50 ꢀm.
The light intensity at the sample plane was 3 mW/ꢀm2 for 532
nm and 0.3 mW/ꢀm2 for 660 nm.
RESULTS
Relative Raman intensities of various alkynes. To design
optimum tag molecules for ATRI, it is indispensable to know
the Raman shift and intensity characteristics of various types
of alkynes. Although there has been great progress in theoretiꢀ
cal studies of Raman spectroscopy, quantitative data on Raꢀ
man intensities of various alkyneꢀcontaining molecules are
still very limited, partly because absolute Raman intensities
are not easy to measure,4 and the signal intensity from the
same molecule can change depending on the measurement
conditions and instrument.5 Therefore, measurement of relaꢀ
Raman imaging. Raman imaging experiments were perꢀ
formed on a slitꢀscanning Raman microscope built in our laꢀ
boratory. Basically, we modified an inverted Nikon Eclipse
microscope (TE2000ꢀU, Nikon, Tokyo, Japan) and introduced
ACS Paragon Plus Environment

Page 3 of 11
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
tive intensities with respect to a standard molecule under the
Raman shift of these diynes also varied depending on the subꢀ
stitution pattern. Although the peaks of the bisarylꢀsubstituted
diynes (Group F) appeared at similar wave numbers (2210 ~
2220 cmꢀ1) to those of the bisarylꢀsubstituted alkynes (group
C), peaks of the other types of diynes (Groups D and E) apꢀ
peared at higher wave numbers compared to those of the simꢀ
ple alkynes (Groups A and B).
same conditions is generally used for comparing Raman intenꢀ
sities between different molecules. In the 1960’s, Alaune et al.
examined the Raman intensities of several alkynes using carꢀ
bon tetrachloride as an internal standard.6 Since our aim is
imaging of bioactive compounds in live cells, a method to
measure the relative Raman intensity of very small amounts of
target molecules is needed. In this study, we employed a simꢀ
ple method for evaluating relative Raman intensity using a
Raman microscope instead of a normal Raman spectrophoꢀ
tometer. EdU was used as a standard because it has already
been used in live cell imaging,3 and it is chemically stable and
nonꢀvolatile. Briefly, the Raman spectrum of a mixture of test
compound and EdU in DMSO was measured, and the relative
Raman intensity versus EdU (RIE) was calculated based on
the intensity of the two alkyne peaks and the molar ratio of the
molecules (Supplementary Figure 1). Although DMSO shows
multiple peaks in the Raman spectrum, none of them appear in
the cellular silent region. DMSO is commonly used for preparꢀ
ing stock solutions of bioactive molecules, so it would be conꢀ
venient if stock solutions could be used directly for this measꢀ
urement.
The results of the evaluation of Raman shifts and Raman inꢀ
tensities of different alkynes enabled us to develop guidelines
for designing alkyneꢀtagged molecules for ATRI. The introꢀ
duction of the smallest ethynyl group at an appropriate posiꢀ
tion in the aromatic ring would be a good choice for aromatic
compounds. On the other hand, diyne would be a suitable
sensitive tag for nonꢀaromatic compounds such as lipids.
Diyne may not be a common functional group, but it can easiꢀ
ly be prepared from terminal alkyne by crossꢀcoupling reaction
(see Supplementary materials).7 Simple twoꢀcarbon elongation
can afford a 5 times stronger Raman signal. We anticipated
that probes with small alkyne tags designed according to these
guidelines would be likely to retain the biological activity of
the parent molecule, and at the same time, have a sufficiently
strong Raman intensity to be detectable with currently availaꢀ
ble Raman microscopes.
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
By using the above method, Raman shifts and relative Raman
intensities of 89 alkynes were evaluated. Figure 2 summarizes
the results (for details, see Supplementary Tables 1 ~ 6). Alꢀ
kynes were divided into 14 groups based on the pattern of
substituents, and the data were plotted in a matrix of RIE and
Raman shift. The results indicated that conjugation of alkyne
to an aromatic ring greatly increases the Raman intensity of
the alkyne (RIE: Group A < B < C), which is in agreement
with previous reports.6 In addition, the type and position of the
substituent in the aromatic ring of πꢀconjugated groups, such
as phenyl and carbonyl groups, influenced the alkyne intensity.
For example, the RIE of 4ꢀethynylbenzaldehyde is clearly
greater than that of 2ꢀethynylbenzaldehyde, though the only
difference is the position of the formyl group. Formyl and
ester substituents at the 4ꢀposition of the phenyl group resulted
in enhanced intensity, which suggests the importance of extenꢀ
sion of πꢀorbitals in the direction of alkyne stretching (Supꢀ
plementary Table 2). It is noteworthy that iodoacetylene derivꢀ
atives also showed much higher intensity [(ioꢀ
doethynyl)benzene vs (bromoethynyl)benzene]. Besides the
Raman intensity, Raman shift also varied greatly depending on
the substitution pattern. The Raman shifts of terminal alkynes
(Groups A1 and B1) were generally observed at lower wave
numbers (2080 ~ 2120 cmꢀ1). On the other hand, those of inꢀ
ternal alkynes (Groups A2, B2, and C) were observed at highꢀ
er wave numbers (2200 cmꢀ1 <). Notably, the Raman peaks of
trimethylsilylꢀ or halogenꢀsubstituted alkynes appeared beꢀ
tween 2150 and 2200 cmꢀ1 (Group A3 and B3). The Raman
shifts of trimethylsilyl and iodide were observed at 2150~
2180 cmꢀ1, whereas that of bromide was near 2200 cmꢀ1.
In addition to the alkynyl group, other functional groups, such
as nitrile, azide and deuterium, are expected to show Raman
scattering in the cellular silent region (Figure 1c). Nitrile
seems to be a promising candidate as a tag because the nitrile
group is small and is often used in medicinal chemistry.8 Azꢀ
ide is also an attractive functional group because it is widely
used as a tag for click chemistry. Deuterium is expected to be
a good Raman tag because of its small size and limited effects
on biological activities. Indeed, Raman imaging of fully deuꢀ
terated fatty acids has already been reported.9 We compared
these groups on the same scaffold to answer the question “Is
alkyne the best tag?” To this end, alkyneꢀ, nitrileꢀ, and azideꢀ
tagged hexanoic acids were synthesized and their Raman inꢀ
tensities were compared with that of commercial hexanoic
acidꢀd11 (Figure 3a and Supplementary Scheme 1). Figure 3a
shows the RIE and Raman shifts of the tagged hexanoic acids.
As expected, all the tags showed a Raman peak in the cellular
silent region (1800 ~ 2800 cmꢀ1), indicating their potential
value as Raman tags. Compared with alkyneꢀtagged hexanoic
acid, the RIE of nitrileꢀtagged hexanoic acid was slightly less
than half, whereas the RIE of azideꢀtagged hexanoic acid was
quite low. In contrast, hexanoic acidꢀd11 produced multiple
signals that overlapped with the EdU peak, and hence evaluaꢀ
tion of its RIE was difficult (Raman spectra of tagged hexanoꢀ
ic acids in DMSO are shown in Supplementary Figure 2). Furꢀ
thermore, we compared the Raman signal of tagged hexanoic
acids in a mixture with nitrile as a benchmark, because the
Raman shift of nitrile was easy to differentiate among the tags
(Figure 3b). The results accord well with the RIE values. To
further examine the relative intensity of deuterium, the Raman
spectrum of acetonitrileꢀd3 was measured (Supplementary
Figure 3). The Raman intensity derived from the three identiꢀ
cal CꢀD bonds was comparable with that of nitrile. The small
size and relatively high Raman intensity of deuterium appear
promising, but the complexity of the signal severely limits its
potential for multicolor imaging. In view of the above results,
we concluded that alkyne tags are the most suitable for Raman
imaging among the functional groups evaluated.
We also investigated conjugated diynes, butadiynes, as potenꢀ
tial Raman tags for ATRI. Interestingly, conjugated diynes
(Group D ~ F) showed much higher intensity (approximately 5
times greater) and higher wave numbers (2200 cmꢀ1 <) comꢀ
pared to simple alkynes (Group A vs D, B vs E, and C vs F).
As expected, conjugation of the diyne to an aromatic ring reꢀ
sulted in enhanced intensity (RIE: Group D < E < F). Bisarylꢀ
substituted diynes (Group F) was found to show the greatest
RIE, being about 25 times more intense than EdU on average.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 11
1
2
3
4
a)
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
b)
Raman shift
(cm-1
Raman shift
(cm-1
R: alkyl, alkenyl, acyl etc.
Ar: aryl, X: SiMe3, Br, or I
number of average
Group
compounds
RIE
representative example
HO
RIE
)
)
A1
A2
8
0.16
2084 ~ 2113
2221 ~ 2241
0.16
2111
2223
R
R
R
H
R
X
HO
HO
5
5
0.11
0.40
0.11
0.17
(CH2)CH3
SiMe3
2164 ~ 2195
2172
A3
22
8
0.81
1.0
2.1
6.2
2091 ~ 2107
2198 ~ 2248
2157 ~ 2194
2216 ~ 2219
0.49
0.63
0.93
4.2
2102
2228
2157
2219
B1
B2
B3
C
Ar
Ar
Ar
Ar
H
Ph
Ph
R
(CH2)3CO2Me
SiMe3
5
Ph
X
5
Ar
Ph
Ph
HO
2
0.65
1.4
2222 ~ 2225
2247 ~ 2255
0.84
1.7
2225
2251
D1
D2
R
R
H
R
HO
HO
13
CMe3
SiMe3
2
1.7
2217 ~ 2222
1.6
2222
D3
R
X
1
6
1
5
4.8
4.1
7.8
25
2208
4.8
5.3
7.8
27
2208
2237
2207
2219
E1
E2
E3
F
Ar
Ar
Ar
Ar
H
Ph
Ph
Ph
2237 ~ 2249
2207
CMe3
R
SiMe3
Ph
X
2217 ~ 2221
Ar
Ph
CHO
Other examples of B1 and B3
OHC
Ph
Br
Ph
I
RIE / Raman shift (cm-1
)
0.73 / 2100
1.3 / 2103
0.85 / 2194
2.8 / 2164
Figure 2. StructureꢀRaman shift/intensity relationship of alkynes. (a) Plot of relative Raman intensity versus EdU (RIE) and Raman shift
of various types of alkynes. (b) Alkynes are divided into 14 groups according to the substitution pattern. Average RIE, range of Raman
shift, and representative alkynes of each group are shown.
Although the Raman intensity of nitrile was lower than that of
alkyne, it might still be useful as a Raman tag. Since inforꢀ
mation about the relative Raman intensity of nitriles is limꢀ
ited,10 we examined the relative Raman intensities of 28 niꢀ
triles (Figure 3c and Supplementary Table 7). As was the case
with alkyne, conjugation to an aromatic ring amplified the
intensity (Group G1 < G2). Nevertheless, the intensity of the
nitrile signal was generally weaker than that of the correspondꢀ
ing terminal alkyne (Group A1 vs G1 and B1vs G2) (Suppleꢀ
mentary Figure 4).
ACS Paragon Plus Environment

Page 5 of 11
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
hydrophobic to be incorporated into cells exogenously, many
lowꢀmolecularꢀweight analogues, such as decylubiquinone and
idebenone, have been developed and investigated.13 In many
cases, the length of the side chain affects the biological activiꢀ
ty, indicating that their cellular uptake and accumulation might
be dependent upon their hydrophobicity.14 Nevertheless, inꢀ
formation about cellular uptake of CoQ analogues is quite
limited,15 mainly because of the lack of a quantitative detecꢀ
tion method for small molecules in living cells. Therefore,
CoQ analogues would be suitable target molecules to demonꢀ
strate the value of ATRI.
a)
O
O
CN
HO
HO
alkyne
nitrile
relative intensity
vs EdU (RIE)
0.15
0.062
2242
Raman shift (cm-1
)
2110
O
D
D D D
O
HO
CD3
N3
HO
D D D D
azide
deuterium
2124
First, we designed and synthesized several alkyneꢀtagged CoQ
analogs, AltQ1~8, with different types of alkyne tag and apꢀ
propriate hydrophobicity (i.e., CLogP values similar to those
of idebenone and decylubiquinone) (Figure 4a and Suppleꢀ
mentary Scheme 2). HeLa cells were incubated in the presence
of 20 ꢀM AltQs for 1 hr, and averaged Raman spectra were
obtained by irradiation at 532 nm with scanning at 3 seꢀ
conds/line. The averaged Raman spectra obtained from 10 x
10 pixels (3.6 ꢀm x 3.6 ꢀm) in the cytoplasmic regions of 21
living cells and the relative Raman intensity are shown in Figꢀ
ure 4b and Figure 4c, respectively (averaged Raman spectra of
the extracellular region are shown in Supplementary Figure 5).
The Raman signals of alkynes with RIE values larger than ca.
0.5 (AltQ1ꢀ6) were easily detected in live cells (Figure 4c).
Furthermore, we observed a rough correlation between the
relative Raman intensity of the AltQs in DMSO and those in
cells, except for AltQ1 (Figure 4c). The observed Raman sigꢀ
nal of AltQ1 in cells was much weaker than expected from the
RIE value measured in DMSO. This may be due to poor upꢀ
take of AltQ1, which has a high ClogP value. This point will
be discussed in detail in the next section. The alkyne peak of
AltQ8 with a simple terminal alkyne was not observed under
these conditions (3 sec/line irradiation, 20 ꢀM concentration),
but the peak was clearly observed when the cells were treated
with 60 ꢀM AltQ8 for 1 hr and scanned at 5 sec/line. Hence,
even the weakestꢀintensity class of alkyne (RIE = 0.17) can
still be used as a Raman tag, if it is sufficiently accumulated in
cells and scanning conditions are set appropriately. The results
obtained with the alkyneꢀtagged CoQ analogs in cells were
consistent with the RIE values obtained in DMSO, validating
our guidelines for designing alkyne tags.
relative intensity
vs EdU (RIE)
0.022
2096
Raman shift (cm-1
)
b)
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
c)
Figure 3. Alkyne is the best tag among the candidate groups. (a)
Relative Raman intensities and Raman shifts of tagged hexanoic
acids. (b) Comparison of Raman intensity at 500 mM after mixing
equal amounts of two hexanoic acid derivatives in DMSO. (c)
StructureꢀRaman shift/intensity relationship of nitriles. G1: aliꢀ
phatic nitriles; G2: aromatic nitriles.
After confirming the Raman shifts/intensities of the alkyneꢀ
tagged CoQ analogs, we use ATRI to visualize the distribution
of CoQ molecules in cells. Endogenous CoQ10 is known to
exist mainly in mitochondria, where it is synthesized and
works as an electron transporter in the respiratory chain.
Moreover, the cellular distribution of exogenously added
CoQ10 was examined by differential centrifugation,15b and
CoQ10 was reported to be enriched mainly in the mitochondrial
fraction. These results imply the existence of a molecular
mechanism of mitochondrial accumulation of CoQꢀrelated
compounds. However, the distribution of shortꢀchain CoQ
analogues in living cells has not been studied due to the diffiꢀ
culty in fluorescent labeling of CoQ. Therefore, we attempted
Raman imaging of AltQ in live cells. Figure 5 shows Raman
imaging of AltQ4 in live HeLa cell as a representative examꢀ
ple (see also Supplementary Figure 6 and 7). Live HeLa cells
were treated with 6 ꢀM AltQ4, and after 50 min, a Raman
image of the cells was obtained using 532 nm excitation.
Overall, not only alkyne but also nitrile and deuterium appear
to be suitable for use as Raman tags, but among them, alkyne
has clear advantages in terms of high signal intensity, narrow
linewidth and suitable wave number of the signal.
Alkyne-tag Raman imaging of CoQ analogues. With these
basic structureꢀRaman intensity relationships in hand, we set
out to test the validity of our approach for the live cell imaging
of mobile small molecules, focusing on lipids as typical moꢀ
bile molecules that are not immobilized under normal fixation
conditions.11 In particular, we are interested in the mitochonꢀ
drial lipid CoQ, which is an essential and abundant molecule
in cells, functioning as an electron transporter in the mitoꢀ
chondrial respiratory chain and an endogenous antioxidant.12
Because endogenous CoQ species, such as CoQ10, are too
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 11
1
2
3
a)
CoQ10
4
5
6
7
8
9
R =
H
9
O
O
CLogP: >12
MeO
MeO
decylubiquinone
CLogP: 5.4
R =
R =
R
idebenone
CLogP: 3.4
OH
AltQ1 (Group C)
AltQ2 (Group E2)
O
MeO
MeO
Ph
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
R =
R =
R
Ph
O
RIE: 7.7
CLogP: 5.2
RIE: 5.3
CLogP: 3.6
AltQs: CLogP = 3.3~5.2
AltQ5 (Group B2)
AltQ3 (Group D3)
TMS
AltQ4 (Group D2)
Ph
R =
R =
R =
RIE: 1.8
CLogP: 3.3
RIE: 1.4
CLogP: 3.6
RIE: 1.2
CLogP: 3.4
AltQ6 (Group D1)
AltQ7 (Group A3)
AltQ8 (Group A1)
R =
R =
R =
I
RIE: 0.46
CLogP: 4.7
RIE: 0.26
CLogP: 4.7
RIE: 0.17
CLogP: 4.1
b)
c)
Figure 5. Raman imaging of AltQ4 in living HeLa cells. A: Raꢀ
man image obtained from HeLa cells treated with 6 ꢀM AltQ4.
Images at 752, 2258, and 2851 cmꢀ1 are shown at the lower panꢀ
els. The overlay images at 752, 2258, and 2851 cmꢀ1 (top panels)
were assigned to the red, green, and blue channels, respectively.
The light intensity at the sample plane was 3 mW/ꢀm2. The expoꢀ
sure time for each line was 5 sec. The total number of lines of
exposure was 95. The image acquisition time was 13 min. B: Raꢀ
man image obtained from control HeLa cells under the same conꢀ
ditions.
Images were reconstructed from the distribution of Raman
peaks at 752, 2258, and 2851 cmꢀ1, which were assigned to
cytochrome c, AltQ4 (diyne tag), and lipid molecules (CH2
stretching), respectively. As is evident in Figure 5, no signal at
2258 cmꢀ1 was observed in the nonꢀtreated control HeLa
cells.3,16 In contrast, AltQ4ꢀtreated cells showed a clear signal
at 2258 cmꢀ1 due to the alkyne. AltQ4 was coꢀlocalized with
cytochrome c to a considerable extent, indicating that AltQ4
was mainly accumulated in mitochondria. However, the accuꢀ
mulation of AltQ4 seemed not to be specific to mitochondria
under these conditions, and it was also coꢀlocalized with lipids.
This is consistent with the reported distribution of CoQ.15b
These results confirm that ATRI is a promising tool for studyꢀ
ing cellular uptake and accumulation of small molecules.
Figure 4. Relative Raman intensity of AltQs in live HeLa cells.
(a) Structures, RIE and predicted CLogP values of AltQs. CLogP
values were predicted by ChemBioDraw Ultra 12.0. (b) Average
Raman spectra of cytoplasmic region (3.6 ꢀm x 3.6 ꢀm) of 21
cells cultured with AltQs. Sample concentration was 20 ꢀM, the
light intensity at the sample plane was 3 mW/ꢀm2, and the expoꢀ
sure time for each line was 3 sec with the exception of AltQ8HL
(HL: high concentration and long exposure). In the case of
AltQ8HL, the sample concentration was 60 ꢀM. The light intensiꢀ
ty at the sample plane was 3 mW/ꢀm2, and the exposure time for
each line was 5 sec. Spectra were vertically offset for ease of
viewing. (c) Relative Raman intensity of AltQs in DMSO and in
cells.
ACS Paragon Plus Environment

Page 7 of 11
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
a)
Estimation of cellular concentration of CoQ analogues.
O
O
Following the successful imaging of AltQs, we set out to estiꢀ
mate the concentration of AltQs in cells by using Raman miꢀ
croscopy. Cellular accumulation of fluorescenceꢀlabeled molꢀ
ecules is expected to be markedly influenced by the fluoresꢀ
cent moiety due to its bulkiness relative to the parent comꢀ
pound. In contrast, a diyneꢀtag (molecular weight: 48) is conꢀ
sidered to be sufficiently small and may have relatively little
influence on the properties of the parent compound. Thereꢀ
fore, the dynamics of cellular accumulation of the tagged
compound may well reflect that of the parent compound.
MeO
MeO
R
compound
RIE
CLogP
R
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
AltQ9
1.6
1.4
1.3
1.2
1.3
1.4
1.6
1.4
1.4
3.1
3.6
4.1
4.7
5.2
5.7
6.2
6.8
7.3
AltQ4
AltQ10
AltQ11
AltQ12
AltQ13
AltQ14
AltQ15
AltQ16
As previously mentioned, the efficiency of cellular uptake of
CoQ analogs is believed to be dependent on their ClogP valꢀ
ues. To confirm this, we prepared a series of diyneꢀtagged
CoQ analogues, AltQ4 and AltQ9 ~ 16, which have side
chains of different lengths (C9 ~ C17) (Figure 6a and Suppleꢀ
mentary Scheme 3). As expected, the relative Raman intensiꢀ
ties of the alkyne in these analogues were similar, whereas the
ClogP values increased with increasing length of the side
chain.
b)
For the estimation of cellular accumulation, HeLa cells were
incubated in the presence of 20 ꢀM diyneꢀtagged CoQ anaꢀ
logues for 1 hr, and averaged Raman spectra were obtained by
excitation at 532 nm with scans of 10 seconds/line. Cellular
concentration of diyneꢀtagged CoQ analogues was derived by
measuring the alkyne signal intensity of the averaged Raman
spectra from 10 x 10 pixels (3.6 ꢀm x 3.6 ꢀm) in the cytoꢀ
plasmic regions of 21 living cells and using a calibration curve
obtained from the Raman spectra of diyneꢀtagged CoQ anaꢀ
logues in solution (Figure 6b and Supplementary Figure 8).
Figure 6. Relative Raman intensity of diyneꢀtagged analogues in
live HeLa cells. (a) Structures, RIE, and predicted CLogP value of
diyneꢀtagged analogues. CLogP values were predicted using
ChemBioDraw Ultra 12.0. (b) Estimation of cellular concentraꢀ
tion of AltQs. Averaged Raman spectra of cytoplasmic region (3.6
ꢀm x 3.6 ꢀm) of 21 cells cultured with AltQs. Sample concentraꢀ
tion was 20 ꢀM. Incubation time was 60 min. The light intensity
at the sample plane was 3 mW/ꢀm2, and the exposure time for
each line was 10 sec.
The cytoplasmic concentration of the shortest analogue, AltQ9,
was estimated to be about 2 mM, which is up to 100 times
higher than the treatment concentration, indicating efficient
uptake and accumulation of this molecule. The estimated cyꢀ
toplasmic concentrations of the analogues apparently deꢀ
creased in proportion to the chain length (and ClogP value).
These results are in agreement with reported findings that the
cellular activity of shorter side chain analogues such as
decylubiquinone and idebenone is greater than that of exogeꢀ
nously added superhydrophobic CoQ10.17 Remarkably, the
observed differences in accumulation of CoQ analogues imply
that a slight change in the side chain length of CoQ analogues,
even a oneꢀcarbon difference, greatly affects the efficiency of
cellular uptake. Although, as shown in Figure 5, the distribuꢀ
tion of the molecules in cells is not uniform and the calibration
curve in DMSO may not accurately reflect the cellular enviꢀ
ronment, ATRI can nonetheless provide approximate values of
concentration. This ability to semiꢀquantitatively estimate celꢀ
lular accumulation of small molecules represents a major adꢀ
vantage of ATRI.
Two different alkyne peaks at 2122 and 2248 cmꢀ1, correꢀ
sponding to EdU and AltQ2, respectively, were easily discrimꢀ
inated (Supplementary Figure 9). Raman images reconstructed
from the distribution of Raman signals at 2122 and 2248 cmꢀ1
clearly indicated different localization patterns of these two
molecules (Figure 7). EdU was localized in the nucleus,
whereas AltQ2 was localized in the cytoplasm.18 An advantage
of using Raman microscopy is that recording of the Raman
spectrum at various regions of interest in the cell is possible.
Therefore, multiple Raman peaks derived from exogenous
molecules as well as various endogenous biomolecules such as
cytochrome c can be detected and imaged simultaneously. In
comparison, Raman imaging of control HeLa cells confirmed
that the signals at 2122 and 2248 cmꢀ1 were derived from alꢀ
kyneꢀtagged molecules, and that the cytochrome c signal was
from the endogenous compound. Since we already know the
structureꢀRaman shift relationships of various alkyne moleꢀ
cules, simultaneous multiꢀcolor imaging of more than two
alkyneꢀtagged molecules should be possible by selecting apꢀ
propriate combinations of alkyne tags. This would make ATRI
a useful imaging platform for studying interactions of multiple
small molecules in cells.
Two-color alkyne-tag Raman imaging. Finally, we attemptꢀ
ed simultaneous twoꢀcolor imaging using ATRI. For this purꢀ
pose, we selected two alkyneꢀtagged molecules, EdU and
AltQ2, which show alkyne signals at different wave numbers.
HeLa cells were treated with 40 ꢀM EdU for 1 day, then exꢀ
cess EdU was washed out and Raman images of live HeLa
cells were obtained in the presence of 2 ꢀM AltQ2 with exciꢀ
tation at 532 nm and scanning at 10 sec/line.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 11
1
2
3
molecules should not present much difficulty (see also Supꢀ
plementary materials).
4
5
6
7
8
9
In this report, the Raman intensities of various alkyne tags
were evaluated using EdU as standard. In order for other reꢀ
searchers to better evaluate our approach for their own appliꢀ
cations, it is useful to compare the Raman intensity of alkyne
with typical biological molecules. One such molecule is pheꢀ
nylalanine, an amino acid found in cells which exhibits a
strong Raman peak around 1000 cmꢀ1, assigned to the ring
breathing mode (Figure 4b).16b The Raman intensity of phenylꢀ
alanine was compared with that of EdU by measuring the Raꢀ
man spectrum of an equimolar mixture (Supplementary Figure
11). Clearly, the EdU signal was about 5 times stronger than
the phenylalanine signal. Hence, the Raman intensities of varꢀ
ious alkyne tags could be estimated relative to the intensity of
the phenylalanine ring breathing mode peak based on the RIE
values reported here.
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
To evaluate our design strategy for ATRI probes, we perꢀ
formed Raman imaging of a series of CoQ analogues. The
relative Raman intensities of CoQ analogues having various
alkyne tags in live cells were first compared with the RIE valꢀ
ues in DMSO, and the results indicated that RIE in DMSO is a
good index for evaluating potential Raman tags. We then used
the ATRI method to visualize diyneꢀtagged AltQ4 in live cells,
confirming the utility of ATRI for imaging small mobile molꢀ
ecules. Further, the cellular accumulation of CoQ analogues
with different lengths of side chain (diyneꢀtagged CoQ anaꢀ
logues) was estimated semiꢀquantitatively with ATRI, and the
results indicated that uptake of the CoQ analogues is dependꢀ
ent on their ClogP value, as expected. Finally, we successfully
demonstrated simultaneous imaging of EdU and AltQ2 in live
cells, indicating the potential of ATRI for multiꢀcolor imaging.
It is important to note that, although mammalian cells do not
normally contain molecules bearing an alkynyl group, many
alkyneꢀcontaining natural products have been found, e.g., in
plants.20 Also, many pharmaceutical molecules studied in meꢀ
dicinal chemistry contain an alkyne moiety.21Therefore, it
should be possible to perform Raman imaging of such moleꢀ
cules without any modification.
Figure 7. Twoꢀcolor alkyneꢀtag Raman imaging. A: Raman image
obtained from HeLa cells treated with 40 ꢀM EdU for 1 day, 2
ꢀM AltQ2 for 30 min. B: Control HeLa cells. Images at 747,
2122, and 2248 cmꢀ1 are shown at the lower panels. The overlay
images at 747, 2122, and 2248 cmꢀ1 (top panels) were assigned to
the blue, red, and green channels, respectively. The light intensity
at the sample plane was 3 mW/ꢀm2. The exposure time for each
line was 10 sec. The total number of lines of exposure was 170.
The image acquisition time was 38 min.
As an imaging technique, ATRI has several advantages over
conventional fluorescence microscopy. Aside from its main
advantage of using a very small tag, the alkyne tag signal does
not degrade over repeated scans unlike fluorescent probes,
which suffers from photobleaching. Moreover, since a Raman
spectrum is obtained at each pixel, ATRI has the ability to
delineate many other cellular structures in addition to the alꢀ
kyneꢀtagged structures in one scan. On the other hand, the
fluorescence technique can only show stained structures.
However, in terms of sensitivity and imaging speed, ATRI is a
much less sensitive technique than fluorescence imaging. With
the current Raman instrument, the sensitivity of Raman detecꢀ
tion for alkyne tags approaches the subꢀmM range (about 0.1ꢀ
0.2 mM for the more intense diynes) whereas recent fluoresꢀ
cence techniques can detect down to a singleꢀmolecule.22 Furꢀ
thermore, image acquisition for ATRI takes several tens of
minutes while videoꢀrate speed is easily achievable for fluoꢀ
rescence imaging.
DISCUSSION
Our recently developed alkyneꢀtag Raman imaging (ATRI)
method permits visualization of small molecules in living cells
by using very small Ramanꢀactive tags that lack the disadꢀ
vantages of larger fluorescent tags, which may markedly influꢀ
ence the chemical and biological properties of the parent molꢀ
ecule.3 Here, to extend the scope of our method to nonꢀ
immobilized small molecules, such as lipids, we first examꢀ
ined the structureꢀRaman intensity relationships of a series of
alkynes to provide a basis for the molecular design of efficient
ATRI probes. The results clearly showed the efficacy of both
an alkyne moiety conjugated to an aromatic ring and a conjuꢀ
gated diyne. Thus, if the parent compound has an aromatic
moiety, the introduction of the small ethynyl group at an apꢀ
propriate position on the aromatic ring would be an excellent
choice and might have relatively little effect on the biological
properties. In the case of aliphatic molecules, diyne may be a
suitable choice as a tag. Since various synthetic methods for
alkynes conjugated to an aromatic ring19 and conjugated
diynes7 are well established, synthesis of appropriately tagged
All the results presented here were obtained by spontaneous
Raman imaging microscopy, but the concept of ATRI is also
applicable to nonlinear Raman imaging techniques, such as
coherent antiꢀStokes Raman scattering (CARS)23 and stimulatꢀ
ACS Paragon Plus Environment

Page 9 of 11
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
ed Raman scattering (SRS)24 microscopy. These nonlinear
techniques should offer faster image acquisition and greater
detection sensitivity.
(6) (a) Alaune, Z., Talaikyte, Z. Lietuvos TSR Mokslu Akademi-
jos Darbai, Serija B: Chemija, Technika, Fizine Geografija
1967, 55ꢀ66. (b) Alaune, Z., Talaikyte, Z. Lietuvos TSR
Mokslu Akad. Darbai Ser. B 1964, 57ꢀ64. (c) Alaune, Z.,
Mozolis, V. Lietuvos TSR Mokslu Akad. Darbai Ser. B 1963,
101ꢀ105.
CONCLUSIONS
(7) (a) Yin, W., He, C., Chen, M., Zhang, H., Lei, A. Org. Lett.
2009, 11, 709ꢀ712. (b) Siemsen, P., Livingston, R. C.,
Diederich, F. Angew. Chem. Int. Ed. 2000, 39, 2632ꢀ2657.
(c) Shun, A. L. K. S., Tykwinski, R. R. Angew. Chem. Int.
Ed. 2006, 45, 1034ꢀ1057.
In conclusion, we examined the structureꢀRaman
shift/intensity relationship of various alkynes that resulted to
guidelines in designing appropriate alkyne tags. With the aid
of obtained relationships, we have achieved the live cell imagꢀ
ing of alkyneꢀtagged CoQ analogues as examples of mobile
small molecules, and the simultaneous twoꢀcolor imaging of
two small molecules. To our knowledge, this is the first examꢀ
ple of direct multiꢀcolor imaging of small molecules without
fluorescent labels in living cells. We believe there is enormous
potential for application of ATRI in biological research in the
near future.
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
(8) Jones, L. H., Summerhill, N. W., Swain, N. A., Mills, J. E.
Med. Chem. Commun. 2010, 1, 309ꢀ318.
(9) (a) Manen, H.ꢀJ. van, Kraan, Y. M., Roos, D., Otto, C. Proc.
Natl. Acad. Sci. U.S.A. 2005, 102, 10159ꢀ10164. (b) Xie, X.
S., Yu, J., Yang, W. Y. Science 2006, 312, 228ꢀ230. (c)
Manen, H.ꢀJ. van, Lenferink, A., Otto, C. Anal. Chem. 2008,
80, 9576ꢀ9582.
(10) (a) Muldakhmetov, Z. M., Kupriyanov, N. S., Akhmetꢀ
karimov, K., Mai, I. Izvestiya Akademii Nauk Kazakhskoi
SSR, Seriya Khimicheskaya 1972, 22, 38ꢀ72. (b) Krasnoꢀ
molova, L. P., Kushnikov, Y. A. Izvestiya Akademii Nauk
Kazakhskoi SSR, Seriya Khimicheskaya 1967, 17, 31ꢀ36. (c)
Shorygin, P. P., Geiderikh, M. A., Ambrush, T. I. Zhurnal
Fizicheskoi Khimii 1960, 34, 335ꢀ342.
(11) Tanaka, K. A. K., Suzuki, K. G. N., Shirai, Y. M., Shibutani,
S. T., Miyahara, M. S. H., Tsuboi, H., Yahara, M., Yoshiꢀ
mura, A., Mayor, S., Fujiwara, T. K., Kusumi, A. Nat. Meth-
ods 2010, 7, 865ꢀ866.
(12) (a) Rakita, R. M., Michael, B. R., Rosen, H. Biochemistry
1989, 28, 3031ꢀ3036. (b) Bentinger, M., Tekle, M., Dallner,
G. Biochem. Biophys. Res. Commun. 2010, 396, 74ꢀ79.
(13) Armstrong, J. S., Whiteman, M., Rose, P., Jones, D. P. J. Bi-
ol. Chem. 2003, 278, 19079ꢀ49084. (b) Suno, M., Nagaoka,
A. Japan. J. Pharmacol. 1984, 35, 196ꢀ198. (c) Shimamoto,
N., Goto, N., Hirata, M. Nihon yakurigaku zasshi, Folia
Pharnacol. Japon 1982, 80, 137ꢀ145.
(14) Helfenbaum, L., Ngo, A., Ghelli, A, Linnane, A. W., Esposꢀ
ti, M. D. J. Bioenerg. Biomembr. 1997, 29, 71ꢀ80. (b) Esꢀ
posti, M. D., Ngo, A., McMullen, G. L., Ghelli, A., Sparla,
F., Benelli, B., Ratta, M., Linnane, A. W. Biochem. J. 1996,
313, 327ꢀ334.
(15) (a) Ross, M. F., Prime, T. A., Abakumova, I., James, A. M.,
Porteous, C. M., Smith, R. A., Murphy, M. P. Biochem. J.
2008, 411, 633ꢀ645. (b) The CoQ accumulating system in
mitochondria has not been completely elucidated. Niki and
coworkers reported that CoQ10 distribution in PC12 cells
exhibited nonlinear relationship with lipid distribution. In
ASSOCIATED CONTENT
Supporting Information.
Detailed experimental procedures, characterization of all comꢀ
pounds, and additional experimental results. This material is
available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
sodeoka@riken.jp
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENT
We thank Dr. Minoru Kobayashi of Nanophoton Corporation for
helpful discussions. We thank Prof. Yoshiharu Iwabuchi for genꢀ
erously providing AZADO. This work was partly supported by a
GrantꢀinꢀAid for Young Scientist (B) (No. 23710276 to H.Y.)
from the Ministry of Education, Culture, Sports, Science and
Technology, Japan.
REFERENCES
(1) (a) Yamakoshi, H., Ohori, H., Kudo, C., Sato, A., Kanoh, N.,
Ishioka, C., Shibata, H., Iwabuchi, Y. Bioorg. Med. Chem.
2010, 18, 1083ꢀ1092. (b) Thuaud, F., Bernard, Y., Turkeri,
G., Dirr, R., Aubert, G., Cresteil, T., Baguet, A., Tomasetto,
C., Svitlin, Y., Sonenberg, N., Nebigil, C. G., Désaubry, L.
J. Med. Chem. 2009, 52, 5176ꢀ5187.
(2) (a) Puppels, G. J., Greve, J. J. Appl. Spectrosc. 1993, 47,
1256ꢀ1267. (b) Uzunbajakava, N., Otto, C. Opt. Lett. 2003,
28, 2073ꢀ2075. (c) Huang, Y.ꢀS., Karashima, T., Yamamoto,
M., Hamaguchi, H. Biochemistry 2005, 44, 10009ꢀ10019.
(d) Mattäus, C., BoydstonꢀWhite, S., Miljković, M., Romeo,
M., Diem, M. Appl. Spectrosc. 2006, 60, 1ꢀ8. (e) Fujita, K.,
Smith, N. I. Mol. Cells 2008, 26, 530ꢀ535.
contrast, the distribution of lipophilic antioxidant αT was diꢀ
rectly proportional to the lipid distribution. These results
suggest that the accumulation of CoQ in mitochondria is meꢀ
diated by some mitochondrial proteins possessing CoQꢀ
binding properties. See: Saito, Y., Fukuhara, A., Nishio, K.,
Hayakawa, M., Ogawa, Y., Sakamoto, H., Fujii, K., Yoꢀ
shida, Y., Niki, E. J. Nutr. Biochem. 2009, 20, 350ꢀ357.
(16) (a) Okada, M., Smith, N. I., Palonpon, A. F., Endo, H., Kaꢀ
wata, S., Sodeoka, M., Fujita, K. Proc. Natl. Acad. Sci.
U.S.A. 2012, 109, 28ꢀ32. (b) Hamada, K., Fujita, K., Smith,
N. I., Kobayashi, M., Inoue, Y., Kawata, S. J. Biomed. Opt.
2008, 13, No. 044027. (c) Ando, J., Fujita, K. Curr. Pharm.
Biotechno. in press.
(17) Haefeli, R. H., Erb, M., Gemperli, A. C., Robay, D., Fruh, I.
C., Anklin, C., Dallmann, R., Gueven, N. PLoS One 2011, 6,
e17963.
(18) The Raman imaging of EdU distribution reveals higher upꢀ
take of EdU in the nucleolus compared to other regions in the
nucleus. The nucleolus uptake of EdU was still observed
even after fixation and permeabilization of the cell, which is
(3) Yamakoshi, H. Dodo, K., Okada, M., Ando, J., Palonpon,
A., Fujita, K., Kawata, S., Sodeoka, M. J. Am. Chem. Soc.
2011, 133, 6102ꢀ6105.
(4) Larkin, P. J., Gustafson, W. G., Asher, S. A. J. Chem. Phys.
1991, 94, 5324ꢀ5330.
(5) (a) Rea, D. G. J. Mol. Spectrosc. 1960, 4, 507ꢀ517. (b) Biꢀ
net, C., Romanet, R. J. Raman Spectrosc. 1976, 5, 253ꢀ267.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 11
1
2
remarkably opposite from the results obtained using click
chemistry (Supplementary Figure 10). One way to interpret
this discrepancy is that the click reactant could not access the
nucleolus under the examined condition or click reaction did
not proceed. Although it is difficult to draw any conclusions
from these observations, our Raman images show without
doubt that alkyne is present in the nucleolus. This goes to
show that ATRI may get to see the blind spots of click chemꢀ
istry.
3
4
5
6
7
8
9
(19) (a) Chinchilla, R., Nájera, C. Chem. Rev. 2007, 107, 874ꢀ
922. (b) Doucet, H., Hierso, J.ꢀC. Angew. Chem. Int. Ed.
2007, 46, 834ꢀ871.
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
(20) (a) Roman, M., Dobrowolski, J. C., Baranska, M., Baranski,
R. J. Nat. Prod. 2011, 74, 1757ꢀ1763. (b) Dembitsky, V. M.
Lipids 2006, 41, 883ꢀ924.
(21) (a) Christiansen, E., Urban, C., Grundmann, M., Dueꢀ
Hansen, M. E., Hagesaether, E., Schmidt, J., Pardo, L.,
Ullrich, S., Kostenis, E., Kassack, M., Ulven, T. J. Med.
Chem. 2011, 54, 6691ꢀ6703. (b) Crosignani, S., Prêtre, A.,
JorandꢀLebrun, C., Fraboulet, G., Seenisamy, J., Augustine,
J. K., Missotten, M., Humbert, Y., Cleva, C., Abla, N., Daff,
H., Schott, O., Schneider, M., BurgatꢀCharvillon, F., Rivron,
D., Hamering, I., Arrighi, J.ꢀF., Gaudet, M., Zimmerli, S. C.,
Juillard, P., Johnson, Z. J. Med. Chem. 2011, 54, 7299ꢀ7317.
(c) Knaus, E. E., Innocenti, A., Scozzafava, A., Supuran, C.
T. Bioorg. Med. Chem. Lett. 2011, 21, 5892ꢀ5896.
(22) Taniguchi, Y., Choi, P. J., Li, G.ꢀW., Chen, H., Babu, M.,
Hearn, J., Emili, A., Xie, X. S. Science 2010, 329, 533ꢀ538.
(23) (a) Zumbusch, A., Holtom, G. R., Xie, X. S. Phys. Rev. Lett.
1999, 82, 4142ꢀ4145. (b) Okuno, M., Kano, H., Leproux, P.,
Couderc, V., Day, J. P. R., Bonn, M., Hamaguchi, H. Angew.
Chem. Int. Ed. 2010, 49, 6773ꢀ6777.
(24) (a) Freudiger, C. W., Min, W., Saar, B. G., Lu, S., Holtom,
G. R., He, C., Tsai, J. C., Kang, J. X., Xie, X. S. Science
2008, 322, 1857ꢀ1861. (b) Saar, B. G., Freudiger, C. W.,
Reichman, J., Stanley, C. M., Holtom, G. R., Xie, X. S. Sci-
ence 2010, 330, 1368ꢀ1370.
ACS Paragon Plus Environment

Page 11 of 11
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