Development of luminescent coelenterazine derivatives activatable by β-galactosidase for monitoring dual gene expression

Article abstract of DOI:10.1002/chem.201302002

Two bioluminogenic caged coelenterazine derivatives (bGalCoel and bGalNoCoel) were designed and synthesized to detect β-galactosidase activity and expression by means of bioluminescence imaging. Our approach addresses the instability of coelenterazine by introducing β-galactose caging groups to block the auto-oxidation of coelenterazine. Both probes contain β-galactosidase cleavable caging groups at the carbonyl group of the imidazo-pyrazinone moiety. One of the probes in particular, bGalNoCoel, displayed a fast cleavage profile, high stability, and high specificity for β-galactosidase over other glycoside hydrolases. bGalN-oCoel could detect β-galactosidase activity in living HEK-293T cell cultures that expressed a mutant Gaussia luciferase. It was determined that coelenterazine readily diffuses in and out of cells after uncaging by β-galactosidase. We showed that this new caged coelenterazine derivative, bGalNoCoel, could function as a dual-enzyme substrate and detect enzyme activity across two separate cell populations. Copyright

Full text of DOI:10.1002/chem.201302002

DOI: 10.1002/chem.201302002
Development of Luminescent Coelenterazine Derivatives Activatable by
b-Galactosidase for Monitoring Dual Gene Expression
Eric Lindberg,[a] Shin Mizukami,[a, b] Keiji Ibata,[c, d] Atsushi Miyawaki,[d, e] and
Kazuya Kikuchi*[a, b]
Abstract: Two bioluminogenic caged
coelenterazine derivatives (bGalCoel
and bGalNoCoel) were designed and
synthesized to detect b-galactosidase
activity and expression by means of
bioluminescence imaging. Our ap-
proach addresses the instability of coe-
lenterazine by introducing b-galactose
caging groups to block the auto-oxida-
tion of coelenterazine. Both probes
the imidazo–pyrazinone moiety. One of
the probes in particular, bGalNoCoel,
displayed a fast cleavage profile, high
stability, and high specificity for b-gal-
actosidase over other glycoside hydro-
lases. bGalN-oCoel could detect b-gal-
actosidase activity in living HEK-293T
cell cultures that expressed a mutant
Gaussia luciferase. It was determined
that coelenterazine readily diffuses in
and out of cells after uncaging by b-
galactosidase. We showed that this new
caged coelenterazine derivative, bGal-
NoCoel, could function as
a dual-
enzyme substrate and detect enzyme
activity across two separate cell popu-
lations.
Keywords: biosensors · cleavage re-
actions · enzymes · galactosidase ·
luminescence
contain
b-galactosidase
cleavable
caging groups at the carbonyl group of
Introduction
emission spectrum, stability, and nontoxicity of its biolumi-
nogenic substrate d-luciferin. However, cofactors such as
adenosine triphosphate (ATP) are required, thereby limiting
its function to intracellular environments and making it de-
pendent on the metabolic state of cells, which can perturb
the cells in which it is expressed.[7,8] ATP also plays a role in
neuronal transmission by activating purinergic neurons,[9]
which makes the use of ATP-dependent luciferases highly
unsuitable for such assays.
Another bioluminogenic substrate, coelenterazine and its
derivatives, are commonly used in conjunction with lucifer-
ases derived from marine organisms such as Renilla, Gaus-
sia, and Oplophorus luciferases. However, BLI with coelen-
terate substrates suffers from several drawbacks, including a
blueshifted emission spectrum that is unfavorable for whole-
Bioluminescence imaging (BLI) has emerged as a useful
tool for the study and monitoring of biological phenomena.
BLI has found wide applications in areas including monitor-
ing of gene expression,[1] protein–protein interactions,[2] dis-
ease progression,[3] stem cells and organs,[4] cancer, and neu-
rological disorders,[5] and several others.[6] Bioluminescence
imaging and analysis have several advantages over fluores-
cence, including low background, high signal-to-noise (S/N)
ratio, wider dynamic range of signals, and applicability in
imaging of small animal models. Firefly luciferase is the
most commonly utilized BLI system owing to its favorable
[a] E. Lindberg, Dr. S. Mizukami, Prof. K. Kikuchi
Graduate School of Engineering, Osaka University
Osaka, 565-0871 (Japan)
Fax : (+81)6-6879-7925
E-mail: kkikuchi@mls.eng.osaka-u.ac.jp
[b] Dr. S. Mizukami, Prof. K. Kikuchi
Immunology Frontier Research Center, Osaka University
Osaka, 565-0871 (Japan)
[c] Dr. K. Ibata
Department of Physiology, School of Medicine
Keio University, Tokyo 160-8582 (Japan)
[d] Dr. K. Ibata, Dr. A. Miyawaki
RIKEN Brain Science Institute
Saitama, 351-0198 (Japan)
[e] Dr. A. Miyawaki
Life, Function and Dynamics, Exploratory Research for
Advanced Technology (Japan), Science and Technology Agency
Saitama, 351-0198 (Japan)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201302002.
14970
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 14970 – 14976

FULL PAPER
animal imaging owing to light attenuation by tissues. In ad-
dition it has been shown that MDR1 P-glycoprotein can ac-
tively transport coelenterazine out of cells.[10] Coelenterazine
is also very unstable,[11] with a half-life of 15 min at 378C.[12]
Its utility is further compromised by its proneness to bind to
serum albumin and other intrinsic proteins, which have been
shown to catalyze the oxidation of coelenterazine, thus
giving rise to background noise in BLI with cellular
assays.[13]
tive, Lugal.[19] Herein we report two new coelenterazine–b-
galactose conjugates as new potential tools for a wide array
of applications including monitoring of dual gene expres-
sion, enzyme activity, and high-throughput screening. It has
also been suggested that SRL could enable applications in
toxicology, pharmacology, and in organ and tissue physiolo-
gy and pathology.[20]
It has been shown that the substituent at the C2 position
of coelenterazine plays a major role in its stability and auto-
Results and Discussion
oxidation.[14,15] Introduction of CF3 at the para position of
Design, synthesis, and properties of bGalCoel: The initial
probe design strategy was based on the previously reported
natural Diaphus Preluciferin (3-enol glucuronide from
fish)[21] by conjugating a b-galactose molecule to coelentera-
zine at the 3-position. bGalCoel was synthesized in four
steps from coelenterazine (Scheme S1 in the Supporting In-
formation). Following peracetylation of coelenterazine, se-
lective deacetylation at the 3-position of the imidazo pyrazi-
none ring was accomplished by the addition of 1% NH3 in
MeOH to 2 in CH2Cl2 at 08C.[22] Subsequently, 2,3,4,6-tetra-
O-acetyl-a-d-galactopyranosyl bromide was added to 3 in
DMF to give 4 as the main product. After standard deacety-
lation with sodium methoxide in MeOH, bGalCoel was pu-
rified by means of reverse-phase HPLC.
À
the C2-benzyl group increased the stability of coelenterazine
against auto-oxidation, but also resulted in a significant de-
crease in bioluminescence activity.[14] Coelenterazine can
also be stabilized by introducing protecting groups at the 3-
carbonyl position of the imidazpyrazinone ring of coelenter-
azine.[16,17] The protecting groups are then cleaved by intra-
cellular esterases and lipases to result in a longer half-life
and a lower rate of auto-oxidation. However, it has been
suggested that different cell lines might give different results
because the type and level of esterases vary with each cell
line, thus leading to a lower or higher light output.[17] In ad-
dition, because cellular esterases are ubiquitous in most
cells, undesired nonselective cleavage of the ester groups in
nontargeted cells could lead to background noise.
Next, the luminescent properties of bGalCoel were evalu-
ated in comparison with native coelenterazine. bGalCoel
displayed glow-type kinetics in the presence of b-Gal and
GLuc (Figure 1a) with an average signal half-life of 77 min
(Figure S5 in the Supporting Information). GLuc is known
to exhibit flash-type kinetics with coelenterazine, with a
half-life of only a few seconds.[23] The luminescent signal of
bGalCoel did not return to basal levels even after 3 h, there-
by making a direct comparison with native coelenterazine
very difficult in terms of relative signal output. Relative bio-
luminescence activity of bGalCoel was 11% of that of coe-
lenterazine after a photon count of approximately 10 min
using a luminometer (Figure 1c). The observed glow-type lu-
minescence of bGalCoel suggested slow cleavage kinetics of
b-galactosidase. HPLC studies further showed that the
cleavage reaction of b-galactosidase with bGalCoel was very
slow, with most of the substrate remaining intact even after
8 h (Figure S3a in the Supporting Information).
To capitalize on the advantages of coelenterate luciferases
while addressing their limitations, we adopted sequential re-
porter-enzyme luminescence (SRL) imaging technology.[18]
SRL allows imaging of another enzyme by making lumines-
cence dependent on the activity of that enzyme through the
use of a caged luciferin substrate (Scheme 1). SRL has been
used to image b-galactosidase (b-Gal) activity, a widely used
reporter enzyme, by using the d-luciferin-b-galactose deriva-
Design, synthesis, and properties of bGalNoCoel: Owing to
slow kinetics in the cleavage of bGalCoel with b-Gal, we re-
designed our substrate by incorporating a self-immolative
nitrobenzyl linker between the coelenterazine and b-galac-
tose moiety as previously described (Scheme 2).[24] We
named this new substrate bGalNoCoel. bGalNoCoel was
synthesized in six steps (Scheme S2 in the Supporting Infor-
mation). Initially, compound 8 was synthesized from 4-hy-
droxy-3-nitrobenzaldehyde and 2,3,4,6-tetra-O-acetyl-a-d-
galactopyranosyl bromide (Scheme S2 in the Supporting In-
formation). The primary alcohol was then treated with thi-
onyl chloride to give 9, followed by alkylation with 3 to give
the per-O-acetylated galactonitrobenzylcoelenterazine deriv-
Scheme 1. The concept of sequential reporter-enzyme luminescence
(SRL) applied to coelenterazine. SRL allows one to monitor the activity
of a separate enzyme by making the luminescence generated from the
oxidation of coelenterazine by its luciferase dependent on the activity of
that enzyme. By installing a caging group on coelenterazine it cannot
react with its luciferase, but in the presence of separate enzyme, the
caging group can be cleaved and free coelenterazine is generated, which
is subsequently oxidized by its luciferase to generate light.
Chem. Eur. J. 2013, 19, 14970 – 14976
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
14971

K. Kikuchi et al.
Scheme 2. Cleavage of bGalNoCoel by b-galactosidase generates free
coelenterazine that is oxidized by Gaussia luciferase in a light-producing
reaction.
ative 10. Deacetylation of 10 with sodium methoxide in
methanol followed by purification by means of reverse-
phase HPLC gave bGalNoCoel.
Evaluation of the luminescent properties of bGalNoCoel
showed a much faster response than bGalCoel (Figure 1b)
and comparable luminescence output to that of native coe-
lenterazine, 78% (Figure 1c). Unlike bGalCoel, bGalNo-
Coel displayed flash kinetics with a signal half-life of less
than 30 s (Figure 1b). Monitoring by means of HPLC indi-
cated a very fast cleavage reaction with the majority of the
substrate cleaved after 5 min (Figure S3b,c in the Support-
ing Information). No intermediates were detected in the
HPLC cleavage reaction. This suggested that the cleavage of
the b-galactoside bond was the slowest step in the enzyme-
catalyzed hydrolysis cascade. The stability of bGalNoCoel
and bGalCoel in nontransfected HEK293T cells was evalu-
ated in comparison to coelenterazine by the addition of
bGalNoCoel, bGalCoel, or coelenterazine to HEK293T cell
cultures, followed by measurement of autoluminescence for
10 min (Figure 2a). Both substrates showed very low levels
of autoluminescence under these conditions. It has also been
reported that coelenterazine is extremely unstable in the
presence of serum albumin by acting as a mono-oxygenase
that catalyzes the oxidation of coelenterazine, thereby
giving rise to background chemiluminescence. Thus we eval-
uated the relative stability of bGalNoCoel and bGalCoel
against coelenterazine by measuring the autoluminescence
levels for 2 min after addition of bGalNoCoel, bGalCoel, or
coelenterazine to solutions that contained bovine serum al-
bumin (BSA; Figure 2b). Both bGalNoCoel and bGalCoel
were found to be highly stable in the presence of BSA,
whereas coelenterazine displayed high levels of autolumi-
nescence, which agrees with previous publications.[13]
Figure 1. Kinetic profiles and relative bioluminescence activity of
a) bGalCoel and b) bGalNoCoel cleavage by b-galactosidase measured
by light output of Gaussia luciferase; reaction conditions: 5 mm bGal-
Coel/bGalNoCoel, 2 U b-galactosidase, 3 mm MgCl2, Gaussia luciferase,
0.1m PBS (pH 7.4). Total light output measured using a luminometer
(12 sꢁ100). c) Relative light output of bGalCoel and bGalNoCoel versus
native coelenterazine (CTZ).
Relative cleavage was measured as total luminescence over
30 min in the presence of GLuc. bGalNoCoel and bGalCoel
showed over 20000- and 1000-fold higher selectivity towards
b-Gal, respectively (Figure 2c).
The relative substrate specificity of bGalNoCoel and
bGalCoel to b-Gal was compared to two other glycoside hy-
drolases, b-glucosidase (b-Glc) and b-glucuronidase (b-Glu).
14972
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 14970 – 14976

Coelenterazine Derivatives
FULL PAPER
Figure 2. Stability and enzyme specificity of bGalCoel and bGalNoCoel.
a) Substrate stability in 2ꢁ104 nontransfected HEK293T cell cultures.
The stability was evaluated by directly adding substrate (15 mm), and the
resulting light output was measured for 10 min at 378C. Experiments
were performed in triplicate. b) Substrate stability in the presence of
BSA. Substrate (final concentration 20 mm) was added to wells containing
BSA (final concentration 2% w/v) in 50 mm HEPES buffer (pH 7.5).
c) Relative enzyme specificity of bGalCoel and bGalNoCoel with three
different glycoside hydrolases. Reaction conditions: bGalCoel or bGal-
NoCoel (final concentration 5 mm) was added to a well containing Gaus-
sia luciferase and b-galactosidase, b-glucosidase, or b-glucuronidase (2 U)
in 0.1m PBS (pH 7.4), and the light output was measured for 30 min.
Each experiment was repeated three times.
Figure 3. Bioluminescence assays of HEK293T cells transfected with
GLucM23 with bGalCoel and bGalNoCoel. a) Relative bioluminescence
of bGalCoel and bGalNoCoel versus native coelenterazine (CTZ) in ap-
proximately 2ꢁ104 HEK293T cells transfected with GLucM23-Venus-
KDEL and in the presence (+) and in the absence (À) of b-Gal. b) Bio-
luminescence output dependence on cell number with bGalNoCoel in
&
^
the presence ( ) and in the absence ( ) of b-Gal and GLucM23ER. c) De-
pendence of contrast (S/N) on cell number with bGalNoCoel. Reaction
conditions: 25 mm substrates, 378C; total light output measured using a
luminometer every 10 s for 100 repeats.
Next, we evaluated the relative bioluminescence activity
in 2ꢁ104 living HEK293T cells that expressed a mutant
GLuc (GLucM23) with a KDEL endoplasmic reticulum lo-
Chem. Eur. J. 2013, 19, 14970 – 14976
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
14973

K. Kikuchi et al.
calization peptide sequence (GLucM23er). GLucM23 was
generated through random mutagenesis and quick-change
standard protocols (see the Supporting Information for
more details). Unlike the wild-type GLuc, GLucM23 does
not display the same flash-type kinetics, but rather a more
stable glow-type signal. In addition, GLucM23 has roughly
10 times higher luminescence output over the wild-type luci-
ferase in the context of BLI with living cells (Figure S1 in
the Supporting Information). bGalCoel or bGalNoCoel was
added to cells that expressed GLucM23er and b-Gal or just
the luciferase (Figure 3a). The maximum luminescence
signal intensity was obtained approximately 2 min after the
addition of bGalNoCoel. bGalNoCoel showed a high signal-
to-noise ratio (S/N=524; Figure 3a). The bioluminescent
emission from cells that expressed b-Gal and GLucM23er
was proportionally dependent on the cell numbers, and as
few as approximately 300 intact cells were detected in the
presence of 25 mm bGalNoCoel (Figure 3b). The observed
contrast (signal-to-noise ratio) was also proportionally de-
pendent on the cell numbers (Figure 3c). These results dem-
onstrate that bGalNoCoel is cell-permeable and able to
detect the b-galactosidase activity in living mammalian cells.
bGalCoel, on the other hand, showed very poor relative bio-
luminescence activity relative to coelenterazine, with a 100-
fold weaker output. Initially this was attributed to poor cell-
membrane permeability of the substrate. However, signifi-
cant luminescence was also observed in the absence of b-
Gal, which suggested that free coelenterazine was generated
from bGalCoel independently of the expression of b-Gal.
The relative signal-to-noise ratio (S/N) was just above 1. In
a separate experiment, we evaluated the relative biolumi-
nescent activity of bGalCoel with HEK293T cells that ex-
pressed Renilla luciferase, another commonly used coelen-
terate luciferase. In this case the signal-to-noise ratio was
even lower (data not shown). Hence the instability of bGal-
Coel seemed to be related to the expression of GLucM23
itself. However, as bGalCoel was shown to be stable in the
presence of secreted GLuc under cell-free assay conditions,
with a signal-to-noise ratio of over 2300, the nonspecific lu-
minescence of bGalCoel does not seem to be related to the
luciferase itself. Interestingly, it was also shown that bGal-
Coel was highly stable in nontransfected HEK293T cell cul-
tures (Figure 2a). In addition, we also investigated whether
the nature of the transfection agent could have any influ-
ence on the stability of bGalCoel. Luminescence was meas-
ured after the addition of bGalCoel to solutions that con-
tained GLuc and one of the following transfection agents:
Lipofectamine 2000, Lipofectamine LTX, FuGene, and
Xtreme Gene. No significant difference could be observed
between the experimental or control groups (data not
shown). Therefore, we are currently uncertain about the
cause of the observed instability of bGalCoel in luciferase-
expressing cells. From the available data, it seems to suggest
that the poor bioluminescence response of bGalCoel ob-
served in living cells is the result of several factors including
slow cleavage kinetics, poor cell-membrane permeability,
and auto-cleavage due to instability of the substrate.
Dual gene-reporter assay: Finally we investigated whether
coelenterazine, generated after cleavage of bGalNoCoel by
b-Gal-expressing cells, would readily diffuse out of the cells
or be retained inside. We excluded bGalCoel on the premise
that it was earlier shown to have poor S/N in living cells
(Figure 3a). In addition to the lacZ gene, we also utilized
two different GLucM23-expressing plasmid vectors, thereby
localizing GLucM23 to the cell-membrane surface (mem) or
endoplasmic reticulum (er), respectively (Figure 4). The lu-
minescence profiles of HEK293T cells that expressed the lu-
Figure 4. Showing the concept of dual gene reporter assay of GLucM23 and b-Gal with bGalNoCoel in living cells. bGalNoCoel diffuses into cells and is
cleaved by b-Gal. The generated coelenterazine will readily diffuse and become oxidized by GLucM23 to generate luminescence: a) cells coexpressing
GLucM2ER and b-Gal; b) mixture of two cell populations expressing GLucM2ER and b-Gal separately; c) cells coexpressing GLucM23MEM and b-Gal; a
mixture of two cell populations expressing GLucM2MEM and b-Gal separately. Orange broken arrows indicate extracellular diffusion of coelenterazine.
14974
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 14970 – 14976

Coelenterazine Derivatives
FULL PAPER
cleaved by b-galactosidase to generate free coelenterazine,
which can then go on to react with Gaussia luciferase to
generate a bioluminescent signal. Our aim was to create bio-
luminogenic coelenterate substrates that could detect b-gal-
actose activity. Both substrates showed very low autolumi-
nescence in living cell cultures and high specificity for b-gal-
actosidase over other glycoside hydrolases. However, bGal-
Coel showed a very slow cleavage reaction profile with b-
galactosidase, and a poor signal-to-noise ratio when measur-
ing b-galactosidase activity in living cell cultures. We found
that incorporation of a 4-hydroxy-3-nitrobenzyl linker be-
tween the b-galactose and the coelenterazine moiety (bGal-
NoCoel) greatly improved the kinetic profile, overall stabili-
ty, luminescence output, and signal-to-noise ratio when de-
tecting b-galactosidase activity in living cells relative to our
first-generation substrate bGalCoel.
Figure 5. Representative mean TLMax values of bGalNoCoel (25 mm) with
2ꢁ104 HEK293T cells cotransfected (/) or independently transfected (+)
with GLucM23 (mem=surface membrane bound; er=endoplasmic reti-
By using bGalNoCoel, we could also show that coelenter-
azine, generated from b-Gal-expressing cells, could readily
diffuse to other cells after uncaging by b-galactosidase. Thus
bGalNoCoel could be used as a dual reporter in two differ-
ent cell-populations in which the “producer” cells, express-
ing b-galactosidase, generate free coelenterazine by cleavage
of bGalNoCoel; the coelenterazine can then diffuse to
neighboring “detector” cells, which express GLucM23,
thereby generating a luminescent response.
Existing caged coelenterazine derivatives such as Endu-
Ren and ViviRen were conceived in part to lower the auto-
luminescence background signal of the natural substrate and
increase the signal-to-noise ratio. However, it has been
shown that both of these substrates display similar back-
ground signal levels to that of native coelenterazine in
vivo.[11,17] It is known that reporter gene assays such as fire-
fly luciferase, which is used in screening applications, can
suffer from artifacts due to inhibition by small-molecular-
weight compounds, many of which are found in typical
screening libraries.[25] In a recent study, many commonly
used reporter gene assays were evaluated against a library
of compounds.[26] Reporters such as firefly luciferase, Renilla
luciferase, and Nanoluc all displayed an increase in biolumi-
nescence output caused by inhibitor-based enzyme stabiliza-
tion. Interestingly, few inhibitors were indentified for b-gal-
actosidase and Gaussia-Dura (a mutated form of Gaussia lu-
ciferase). Hence our bGalNoCoel substrate could be a suita-
ble dual reporter for screening applications. We also hope to
further evaluate our dual-reporter substrate for in vivo
imaging applications in the near future.
ACHTUNGTRENNUNGculum localization) and b-Gal. Luminescence was measured using a lu-
minometer at 378C every 10 s for 100 repeats. Statistical analyses were
performed with a two-tailed Studentꢂs t-test; *p<0.01 (n=4), error bars
indicate standard error of mean, and p refers to the p-value, or signifi-
cance value.
ciferase together with b-Gal were compared against the luci-
ferase and b-Gal being independently expressed and com-
bined before the addition of bGalNoCoel. We combined
two sets of 1ꢁ104 cells in which GlucM23 and b-Gal ex-
pressed separately, thus giving a cell ratio of 1:1. If coelen-
terazine could freely diffuse out of cells, we hypothesized
that the average time to reach maximum luminescence
(TLmax) should be related to the relative distance between
the sites of expression of the luciferase and b-Gal, that is,
the shortest distance would be the set of cells that coexpress
GLucM23er/b-Gal, whereas the longest distance would be
between the set of cells that independently express
GLucM23er+b-Gal. Therefore, the following trend in terms
of TLmax would be expected: GLucM23er/b-Gal<GLucM23-
mem/b-Gal <GLucM23mem+b-Gal<GLucM23er+b-Gal.
However, statistical analysis of the observed data indicated
the following trend: GLucM23er/b-Gal<GLucM23mem/b-
Gal !GLucM23mem+b-Gal=GLucM23er+b-Gal
(Fig-
ACHTUNGTRENNUNGure 5). We could not observe a significant difference be-
tween the means of GLucM23mem+b-Gal and GLucM23-
er+b-Gal. Yet we could show from our data that after the
b-galactose moiety of bGalNoCoel is cleaved by b-Gal, the
resulting coelenterazine diffused from the cytosol to the ex-
tracellular environment. Hence bGalNoCoel could poten-
tially function as a dual gene reporter in two different cell
populations, whereby the b-Gal-expressing cells could be
thought of as “producers” of coelenterazine from bGalNo-
Coel, whereas the GLucM23-expressing cells could be la-
beled as “detectors”.
Experimental Section
Measurement of bioluminescence activity: Stock solutions of biolumino-
genic substrates (10 mm; bGalCoel, bGalNoCoel, or coelenterazine) in
phosphate-buffered saline (PBS) (0.1m, pH 7.4, 0.02% tween) were
added using an auto-injector to wells that contained a solution of b-galac-
tosidase (2 U), MgCl2 (6 mm), and Gaussia luciferase in PBS buffer
(10 mL; 0.1m, pH 7.4). Photon counting took place every 6 s for 100 re-
peats. Each experiment was repeated three times.
Conclusion
We have designed and synthesized two caged coelenterazine
derivatives, bGalCoel and bGalNoCoel. The probes were
Chem. Eur. J. 2013, 19, 14970 – 14976
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
14975

K. Kikuchi et al.
Autoluminescence of bioluminogenic probes in cultured live cells: Non-
transfected HEK293T cells were washed in PBS, trypsinized, and seeded
into wells of a multiplate (2ꢁ104 cells/well) that contained Leibovitzꢂs L-
15 medium, and the plate was incubated at 378C. A solution of bGal-
Coel, bGalNoCoel, or coelenterazine (final concentration 15 mm) in PBS
buffer (0.1m, pH 7.4) was added to each well. Autoluminescence was
measured using a luminometer every minute for 10 min. Each experi-
ment was repeated four times.
national frontier biotechnology program. We would like to thank Dr.
Yutaro Kumagai and Prof. Shizuo Akira for providing us with the plas-
mid for expressing Renilla luciferase.
[1] Y. Nakajima, Y. Ohmiya, Expert Opin. Drug Discovery 2010, 5,
835–849.
[2] A. Kaihara, Y. Umezawa, Chem. Asian J. 2008, 3, 38–45.
[3] D. M. Close, T. T. Xu, G. S. Sayler, S. Ripp, Sensors 2010, 11, 180–
206.
[4] N. F. Huang, J. Okogbaa, A. Babakhanyan, J. P. Cooke, Theranostics
2012, 2, 346–354.
[5] K. Hochgrafe, E.-M. Mandelkow, Mol. Neurobiol. 2013, 47, 868–
882.
Autoluminescence in BSA:
A stock solution of BSA (4% w/v) in
HEPES buffer (50 mm, pH 7.5) (final concentration 2% w/v) was added
to wells that contained bGalCoel, bGalNoCoel, or coelenterazine in 4-(2-
hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) buffer (50 mm,
pH 7.5) (final concentration 20 mm). The plate was incubated at 258C. Lu-
minescence was detected by a luminometer every 6 s for 20 repeats. Con-
trol was taken as the total photon count when no substrate was added.
Each experiment was repeated three times.
[6] T. Ozawa, H. Yoshimura, S. Bae Kim, Anal. Chem. 2013, 85, 590–
609.
[7] O. Gandelman, I. Allue, K. Bowers, P. Cobbold, J. Biolumin. Chemi-
lumin. 1994, 9, 363–371.
Measurement of probe enzyme specificity: b-Galactosidase, b-glycosi-
dase, or b-Glucuronidase in PBS (2 U; 0.1m, pH 7.4) was added to wells
that contained a solution of Gaussia luciferase (10 mL), bGalCoel, or
bGalNoCoel (final concentration 5 mm) in PBS buffer (0.1m, pH 7.4) in-
cubated at 258C. Photon count was measured using a luminometer every
18 s for 100 repeats (total 1800 s).
[8] J. K. Rꢃty, T. Liimatainen, M. Unelma Kaikkonen, O. Grohn, K. J.
Airenne, S. Yla-Herttuala, Mol. Ther. 2007, 15, 1579–1586.
[9] B. S. Khakh, R. A. North, Neuron 2012, 76, 51–69.
[10] A. Pichler, J. L. Prior, D. Piwnica-Worms, Proc. Natl. Acad. Sci.
USA 2004, 101, 1702–1707.
[11] a) T. Kimura, K. Hiraoka, N. Kasahara, C. R. Logg, J. Gene Med.
2010, 12, 528; b) M. K. Wendt, J. Molter, C. A. Flask, W. P. Schie-
mann, J. Vis. Exp. 2011, e3245.
[12] E. M. Hawkins, in Bioluminescence and Chemiluminescence: Prog-
ress and Current Applications (Eds.: P. E. Stanley, L. J. Kricka)
World Scientific, Singapore, 2002, p. 149.
[13] a) N. Vassel, C. D. Cox, R. Naseem, V. Morse, R. T. Evans, R. L.
Power, A. Brancale, K. T. Wann, A. K. Campbell, Luminescence
2012, 27, 234; b) M. Keyaerts, C. Heneweer, L. O. Tchouate Gain-
kam, V. Caveliers, B. J. Beattie, G. A. Martens, C. Vanhove, A. Bos-
suyt, R. G. Blasberg, T. Lahoutte, Mol. Imaging Biol. 2011, 13, 59–
66.
Measurement of b-galactosidase activity in HEK293T cells: HEK293T
cells grown in Dulbecco’s modified eagle medium (DMEM; Invitrogen)
that contained fetal bovine serum (FBS; 10%) at 378C under CO2 (5%)
were transfected with plasmids coding for GLucM23-Venus-KDEL or/
and cytoplasmic LacZ using lipofectamine 2000 (Invitrogen), and the
cells were incubated at 378C for 24 h. The cells were then washed three
times with PBS, trypsinized, and seeded into 96-microplate wells (2ꢁ104
cells per well) containing Leibovitzꢂs L-15 medium, and incubated at
378C. To each well was added bGalCoel, bGalNoCoel, or coelenterazine
(final concentration 25 mm) in PBS (0.1m, pH 7.4). Luminescence was
measured for each well every 10 s for 100 repeats (ꢀ17 min).
Determination of relative diffusion of coelenterazine: HEK293T cells
transfected with four different sets of plasmids: 1) pDisp-mKO-
[14] S. Inouye, Y. Sahara-Miura, J.-I. Sato, R. Iimori, S. Yoshida, T.
Hosoya, Protein Expres. Purif. 2013, 88, 150.
GLucM23/pcDNA4/TO/myc-His/lacZ;
2) pDisp-mKO-GLucM23+
pcDNA4/TO/myc-His/lacZ;
pcDNA4/TO/myc-His/lacZ;
3) pcDNA3-GLucM23-Venus-KDEL/
4) pcDNA3-GLucM23-Venus-KDEL+
[15] M. P. Hall, J. Unch, B. F. Binkowski, M. P. Valley, B. L. Butler, M. G.
Wood, P. Otto, K. Zimmerman, G. Vidugiris, T. Machleidt, M. B.
Robers, H. A. Benink, C. T. Eggers, M. R. Slater, P. L. Meisenheim-
er, D. H. Klaubert, F. Fan, L. P. Encell, K. V. Wood, ACS Chem.
Biol. 2012, 7, 1848–1857.
pcDNA4/TO/myc-His/lacZ ((/): co-transfected; (+): independently trans-
fected). After transfection, cells were incubated for 24 h at 378C under
CO2 (5%). The cells were washed with PBS three times, trypsinized,
washed with Leibovitzꢂs L-15 medium, transferred, and re-suspended in
black 96-well optiplates (Perkin–Elmer); a total volume of 100 mL Leibo-
vitzꢂs L-15 medium was incubated at 378C in the luminometer until addi-
tion of bGalNoCoel (100 mL, 50 mm in 0.1m PBS pH 7.4). Note: For ex-
periments with cotransfected cells (/), 2ꢁ104 cells were resuspended in
each well; for experiments with independently expressed cells (+), 1ꢁ104
cells from cells that expressed bGal were re-suspended with 1ꢁ104 cells
that expressed GLucMem or GLucER. Luminescence was measured for
each well every 10 s for 100 repeats (ꢀ17 min). The values of TLMax were
taken as the data point with the highest signal intensity. Each experiment
was repeated four times.
[16] J. Levi, A. De, Z. Cheng, S. Gambhir, J. Am. Chem. Soc. 2007, 129,
11900.
[17] M. Otto-Duessel, V. Khankaldyyan, I. Gonzalez-Gomez, M. C.
Jensen, W. E. Laug, M. Rosol, Mol. Imaging 2006, 5, 57–64.
[18] a) T. S. Wehrman, G. von Degenfeld, P. O. Krutzik, G. P. Nolan,
H. M. Blau, Nat. Methods 2006, 3, 295–301; b) J. Li, L. Chen, L. Du,
M. Li, Chem. Soc. Rev. 2013, 42, 662–676.
[19] R. Geiger, E. Schneider, K. Wallenfels, W. Miska, Biol. Chem.
Hoppe-Seyler 1992, 373, 1187–1191.
[20] G. von Degenfeld, T. S. Wehrman, H. M. Blau, Methods Mol. Biol.
2009, 574, 249–259.
[21] S. Inoue, K. Okada, H. Tanino, H. Kakoi, Chem. Lett. 1987, 417–
418.
[22] S. Inoue, S. Sugiura, H. Kakoi, K. Hashizume, T. Goto, H. Iio,
Chem. Lett. 1975, 141–144.
[23] B. A. Tannous, D.-E. Kim, J. L. Fernandez, R. Weissleder, X. O.
Breakefield, Mol. Ther. 2005, 11, 435–443.
[24] N.-H. Ho, R. Weissleder, C.-H. Tung, ChemBioChem 2007, 8, 560.
[25] D. S. Auld, N. T. Southall, A. Jadhav, R. L. Johnson, D. J. Diller, A.
Simeonov, C. P. Austin, J. Inglese, J. Med. Chem. 2008, 51, 2372–
2386.
[26] P.-i. Ho, K. Yue, P. Pandey, L. Breault, F. Harbinski, A. J. McBride,
B. Webb, J. Narahari, N. Karassina, K. V. Wood, A. Hill, D. S. Auld,
ACS Chem. Biol. 2013, 8, 1009–1017.
Acknowledgements
This work was supported in part by the Cabinet Office, Government of
Japan and through the Funding Program for World-Leading Innovative
R&D on Science and Technology (FIRST Program) from the JSPS, by
the Grant-in-Aid for Scientific Research from MEXT of Japan (nos.
20675004, 22108519, 25620133, 25242072, and 24108724 for K.K.;
24685028, 24115513, and 24651259 for S.M.; 23700387 for K.I.) by
CREST from JST, by the Grant-in-Aid from MHLW of Japan, by the
Naito Foundation, by the Takeda Science Foundation, by the Inamori
Foundation, by the Sumitomo Foundation, and by the Asahi Glass Foun-
dation. The authors are also grateful to Japanese government scholarship
(Monbukagakusho) for E.L. to perform his PhD thesis work in the inter-
Received: May 24, 2013
Revised: July 8, 2013
Published online: September 17, 2013
14976
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 14970 – 14976