Development of a hypoxia-selective near-infrared fluorescent probe for non-invasive tumor imaging

Article abstract of DOI:10.1248/cpb.60.402

A near-infrared fluorochrome, GPU-311, was designed, synthesized and evaluated for its application in non-invasive imaging of tumor hypoxia. Efficient synthesis was achieved by nucleophilic substitution and click chemistry ring using the bifunctional tetr

Full text of DOI:10.1248/cpb.60.402

—
402
Note
Chem. Pharm. Bull. 60(3) 402 407 (2012)
Vol. 60, No. 3
Development of a Hypoxia-Selective Near-Infrared Fluorescent Probe for
Non-invasive Tumor Imaging
Bahaa Gamal Mohamed Youssif,a,b Kensuke Okuda,a Tetsuya Kadonosono,c
Ola Ibrahim Abdel Razek Salem,b Alaa Arafat Mohamed Hayallah,b Mostafa Ahmed Hussein,b
,a
Shinae Kizaka-Kondoh,c and Hideko Nagasawa*
a Laboratory of Medicinal and Pharmaceutical Chemistry, Gifu Pharmaceutical University; Gifu 501–1196, Japan:
b Pharmaceutical Organic Chemistry Department, Faculty of Pharmacy, Assiut University; Assiut 71526, Egypt: and
c Division of Biofunctional Engineering, School of Bioscience and Biotechnology, Tokyo Institute of Technology;
Midori-ku, Yokohama 226–8503, Japan. Received September 30, 2011; accepted December 31, 2011
A near-infrared fluorochrome, GPU-311, was designed, synthesized and evaluated for its application in
non-invasive imaging of tumor hypoxia. Efficient synthesis was achieved by nucleophilic substitution and
click chemistry ring using the bifunctional tetraethylene glycol linker 2 containing thiol and azide groups for
the conjugation of the propargylated nitroimidazole 1 and the heptamethine cyanine dye 3 bearing a 2-chlo-
ro-1-cyclohexenyl ring. GPU-311 exhibited long excitation and emission wavelength (Ex/Em=785/802nm) and
a decent quantum yield (0.05). The water solubility and hydrophilicity of GPU-311 increased. After in vitro
treatment of SUIT-2/HRE-Luc pancreatic cancer cells with GPU-311, a higher level of fluorescence was ob-
served selectively in hypoxia than in normoxia. However, in vivo fluorescence imaging of a mouse xenograft
model after GPU-311 administration revealed inadequate accumulation of GPU-311 in tumors due to its
rapid elimination through the liver.
Key words near-infrared fluorescence; tumor hypoxia; 2-nitroimidazole; in vivo imaging
Tumor hypoxia is now known as one of the most prominent NIR fluorescence with significant tumor-to-background (T/B)
features of malignant neoplasias, distinguishing them from ratio in a mouse xenograft model. GPU-167 maintained a good
normal tissues.1) The hypoxic status of various solid tumors T/B ratio (>2.5), and it was retained for a relatively longer pe-
can markedly influence the therapeutic response of malignant riod (>48h) in the body. However, the considerably high liver
tumors to conventional irradiation, chemotherapy and other distribution of GPU-167 was similar to that observed in the tu-
nonsurgical treatment modalities.2,3) Therefore, there has been mor. In this study, to improve the pharmacokinetic properties
a growing impetus to develop non-invasive imaging methods and synthetic yield of GPU-167, we designed and synthesized
to detect and assess tumor hypoxia.4,5)
In the development of imaging probes for tumor hy- tumor hypoxia.
a new NIR fluorescent probe, GPU-311 for in vivo imaging of
poxia, nitroimidazoles have received particular attention
As shown in Fig. 2, the molecular design involved conju-
as exogenous markers because of their unique behavior in gation of a 2-nitroimidazole derivative as a hypoxia marker
hypoxic environments owing to their high electron affinity.4) and a heptamethine cyanine dye as an NIR fluorescent dye
Nitroimidazole compounds undergo selective bioreduction in through a polyether linker. To synthesize GPU-311, we em-
hypoxic cells to form reactive species that irreversibly bind ployed the chloro-substituted heptamethine cyanine dye
to cellular macromolecules.6) Currently, positron emission to- 311) as an intermediate fluorophore, which has a 2-chloro-
mography using nitroimidazole tracers is the most advanced 1-cyclohexenyl ring moiety providing a chemical reactive site
technique to achieve non-invasive in vivo mapping of tumor for nucleophilic substitution at the center methine carbon and
hypoxia with anatomical resolution.7) However, radionuclide increased photostability in comparison to cyanine dyes with
imaging requires radioactive compounds which have an in- an open polymethine chain.12) Thus we can easily connect a
trinsically limited half-life and expose the patient and practi- hypoxia target moiety to the fluorophore through a suitable
tioner to ionizing radiation, and are therefore subject to a vari- linker. In addition, GPU-311 has two sulfonate groups which
ety of stringent safety regulations that limit their repeated use.
In contrast, optical imaging has comparable sensitivity to
“
”
radionuclide imaging, and it can be targeted if the emitting
fluorophore is conjugated to a targeting ligand.8) That is safer
and easier to perform than nuclear imaging. Optical imaging
—
in the near-infrared (NIR) region (700 900nm) has a low
absorption by intrinsic photoactive biomolecules and allows
light to penetrate several centimeters into the tissue, a depth
sufficient for imaging practically all small animal models.9)
Therefore, NIR fluorescence imaging, a less expensive and
non-invasive technique, may be a powerful tool for imag-
ing tumor hypoxia. Recently, we developed an NIR hypoxia
probe, GPU-167 (Fig. 1), comprising 2-nitroimidazole and
tricarbocyanine dye, for in vivo imaging of tumor hypoxia.10)
This is the first chemical probe to visualize tumor hypoxia by
Fig. 1. Structure of GPU-167
© 2012 The Pharmaceutical Society of Japan
*To whom correspondence should be addressed. e-mail: hnagasawa@gifu-pu.ac.jp

March 2012
403
Fig. 2. Molecular Design of GPU-311
Chart 1. Synthesis of Hypoxia Probe, GPU-311
increase water solubility and prevent dye aggregation.13) Then, linker 2, and NIR dye 3 are the key intermediates for develop-
the bifunctional tetraethylene glycol linker 2 containing thiol ing the probe.
and azide groups was employed to increase the hydrophilicity
of the probe compared to GPU-167 and to facilitate crosslink- cedure mentioned in previous studies.11,16) Then compound 9,
ing between the NIR dye and 2-nitroimidazole moiety.14)
which consisted of 1 and 2, was prepared from the known
Compounds 1 and 3 were prepared according to the pro-
A
triazole linker was selected owing to its metabolic stability S-11-hydroxy-3,6,9-trioxaundecyl ethanethioate (4)17) in six
and ease of introduction from azide- and alkyne-functional- steps (Chart 1). First, tosylation of 4 followed by azide sub-
ized precursors using click chemistry.15) Good nucleophilicity stitution gave 614) (71% yield, two steps), and then a Cu(I)-
of thiol is suitable for substitution at the central vinylogous catalyzed 1,3-dipolar cycloaddition reaction with 1 yielded 7
halide carbon of the chloro-substituted dye 3 to connect the at a 91% yield. An attempt to deprotect the acetyl group of
linker moiety, and the triazole-forming click reaction is a reli- 7 caused accompanying disulfide formation of the product
able method to link molecules; thus we designed GPU-311 as by either acid (method A) or alkaline treatment (method B)
a probe candidate (Fig. 2). Propargylated nitroimidazole 1,16) to produce 8 (78% or 82% yield, respectively). Reduction of

404
Vol. 60, No. 3
Fig. 3. LC Spectra of GPU-311
Fig. 4. (A) Accumulation of GPU-311 to Hypoxic Cells in Vitro, (B)
Representative Fluorescence Image, (C) Cellular Uptake of GPU-311
(A) The SUIT-2/HRE-Luc cells (2×105 cells) cultured under normoxic (N) or
hypoxic (H) conditions were treated with GPU-311 and analyzed by measurement
of fluorescent intensity with an Ex/Em wavelength of 740nm/780nm (*p<0.05,
8 by triphenylphosphine produced 9 (74% yield). Finally, the
target molecule GPU-311 was obtained in 66% yield by the
reaction of 3 with excess amount of 9. It was purified by C18 n=3). The amount of GPU-311 in the cells was calculated by using intensity–con-
reversed-phase flash chromatography. Purity and identity were
determined by NMR and LC-MS, which yielded single peaks
at both 300nm and 630nm absorption (Fig. 3) and a high-res-
centration curve. (B) NIR imaging was performed using an IVIS spectrum (Ex/
Em=710nm/800nm), (C) Cellular uptake was evaluated from fluorescent intensity
immediately after the probe treatment at the concentration of 1μM, while intracel-
lular residual binding was examined after the probe treatment and post-incubation
in fresh medium for 1h.
olution mass spectrum with m/z 1077.4260 for C52H69N8O11S3
([M−Na+2H]+). The evaluation of the photophysical proper-
ties of GPU-311 revealed the absorption and emission peaks one third. These results suggested that GPU-311 may penetrate
at 785 and 802nm, respectively, a large extinction coefficient and bind to intracellular substances irreversibly in hypoxic
(1.2×104 M−1 cm−1 at 785nm) and a decent quantum yield cells significantly more than in aerobic cells.
(0.05). The hydrophobicity index (Rm) of GPU-311 and GPU-
Next, GPU-311 was evaluated by in vivo experiments using
167 was examined by a reversed-phase thin-layer chromato- xenografts of SUIT-2/HRE-Luc cells according to the proce-
graphic system (mobile phase: 10% 5mM sodium phosphate dure described in previous studies.19) The SUIT-2/HRE-Luc
buffer, pH 7.4 and 90% methanol). GPU-311 was believed to cells stably carry 5HRE-luciferase, a luciferase reporter gene
be more hydrophilic than GPU-167 due to their Rm values under the control of the hypoxia-inducible factor (HIF)-
(−0.43 vs. −0.26).
dependent promoter, and thus HIF-1-activity in xenografts can
To evaluate the hypoxia selectivity of GPU-311, an in vitro be monitored by bioluminescence imaging. When GPU-311
fluorescence assay was performed using SUIT-2/HRE-Luc was injected into tumor-bearing mice, GPU-311 could reveal
pancreatic cancer cells.18) In the assay, after treatment with the tumors 3h after injection with the highest T/B ratio of
GPU-311 under normoxic or hypoxic conditions for 30min, 2.52 (Fig. 5). In SUIT-2/HRE-Luc cells, HIF-1α expression
the cells were incubated for 1h in fresh medium under nor- is undetectable under normoxic conditions but significantly
moxic condition and washed with phosphate buffered saline increased under hypoxic conditions because HIF-1α stability is
(PBS) to remove unbound dye. During post-incubation in strictly regulated by oxygen concentration.18) Then, to confirm
fresh medium, unbound probe should be released outside the existence of hypoxic tumor cells, bioluminescence emit-
the cells. We compared intensity of cellular fluorescence of ted from HIF-1-active hypoxic cells was also examined 24h
GPU-311 treated in hypoxic or aerobic condition before or after injection of GPU-311 (Fig. 5A, right panel). Fluorescence
after post-incubation, to distinguish between nonspecific cel- signals of GPU-311 were observed in the forefoot and the ab-
lular uptake and irreversible intracellular binding. When the dominal regions after 3h (Fig. 5A, left panel). The signal in
cells were cultured in hypoxia (0.1% O2) and normoxia (21% the forefoot seemed to be correlated with the bioluminescent
O2) with GPU-311 at concentrations of 0.2, 1 and 5μM, and signal of the tumor, suggesting that the fluorescence owing to
then exposed to post-incubation for 1h in fresh medium, GPU-311 in the tumor was related to hypoxia. However, fluo-
significantly stronger fluorescence signals were detected by rescence signals of GPU-311 in the tumor rapidly disappeared
probe treatment in hypoxia than in normoxia (Figs. 4A,B). As 6h after its injection, suggesting that the fluorescence signal
shown in Fig. 4C, fluorescence signals of hypoxic cells were of GPU-311 was unstable in hypoxic tumors. The fluorescence
higher than that of normoxic cells even before post-incubation. signal of GPU-311 was also detected in the abdominal region
After post-incubation, fluorescence signals of hypoxic cells until 24h. In comparison with GPU-167, the clearance of
decreased by half, while those of aerobic cells decreased by GPU-311 became much faster, while tumor accumulation was

March 2012
405
Fig. 5. (A) In Vivo Optical Imaging of Hypoxic Region in Subcutaneous Cancers with GPU-311, (B) The Relative Fluorescence Intensity of the
Tumor to the Muscle (T/B Ratio)
(A) Representative in vivo image of hypoxic tumors after 10nmol of GPU-311 administration. Nude mouse carrying SUIT-2/HRE-Luc xenografts in both forefeet were
imaged at the indicated time after GPU-311 injection. The right panel (HIF) shows bioluminescent image of HIF-1-active hypoxic tumor at 24h. The minimum and maxi-
mum photons/s/cm2/steradian of each image was indicated in each figure by a color bar scale. (B) Fluorescence intensities of the SUIT-2/HRE-Luc xenografts and the
muscle of the hind foot were measured at the indicated time after GPU-311 administration.
declined significantly. These data suggest that the hypoxia sphere. 1H-NMR spectra were obtained with JEOL JNM-
probe requires further optimization of the hydrophilic–hydro- ECA500 spectrometer at 500MHz frequency or JNM-AL400
phobic balance.
spectrometer at 400MHz frequency in CDCl3 or DMSO-d6
In summary, we developed an NIR fluorescent hypoxia with tetramethylsilane as an internal standard. Chemical shifts
probe, GPU-311, which was synthesized using simple proce- are reported in ppm. Coupling constants are reported in Hz.
dures to produce moderate or good yields and was easily puri- The multiplicity is defined by s (singlet), d (doublet), t (triplet),
fied. It exhibited good optical properties, high hydrophilicity dd (doublet of doublets), br (broad), and m (multiplet). The
and significant in vitro hypoxia selectivity. According to in direct analysis in real time (DART)-MS measurements were
vivo imaging, although GPU-311 was obviously detected with carried out on a JEOL JMS-T100TD. The electrospray ioniza-
an adequate T/B ratio (2.52) in tumors 3h after injection, its tion (ESI)-MS measurements were carried out on a Shimadzu
fluorescence intensity was not strong, and most of the probe LCMS-IT-time-of-flight (TOF). UV-Vis spectra were obtained
was thought to be rapidly eliminated though the liver due to by Beckman DU-600 spectrophotometer. The fluorescence ex-
its considerable hydrophilicity. Consequently, the design strat- citation spectra and emission spectra were obtained by JASCO
egy of nitroimidazole–NIR cyanine dye conjugates appears FP-6600 Fluorescence spectrophotometer. The slit width was
promising for developing hypoxia probes for in vivo tumor im- 5nm for excitation and 10nm for emission. 2-Nitro-1-(prop-
aging. However, further molecular modifications are necessary 2-ynyl)-1H-imidazole (1),16) tricarobycyanine dye 3,11) S-11-
to improve the pharmacokinetics of the nitroimidazole–NIR hydroxy-3,6,9-trioxaundecyl ethanethioate (4)17) were prepared
dye conjugates to be applicable for in vivo hypoxia imaging.
according to the literature procedures.
S-11-(Tosyloxy)-3,6,9-trioxaundecyl Ethanethioate (5)
To an ice-cooled solution of 417) (500mg, 2.0mmol) in CH2Cl2
Experimental
General Information Unless otherwise stated, all re- (5mL), p-toluene sulfonyl chloride (TsCl) (570mg, 3.0mmol)
agents and solvents were obtained from Sigma-Aldrich, Wako in CH2Cl2 (2mL) and triethylamine (400mg, 4.0mmol) were
(Japan), Nacalai (Japan), or TCI without further purification. added. The mixture was stirred at room temperature for 20h.
Normal-phase TLC was carried out on Silica gel 60 F254 Water (5mL) was added to the mixture, and then extracted
(Merck, 1.05715.0009) using reagent graded solvents. Normal- with CHCl3 (3×10mL). The organic phase was washed with
phase column chromatography was performed on silica gel water and brine, dried over anhydrous Na2SO4 and concentrat-
(AP-300, Dai koh Trading Co., Ltd.). Reverse-phase TLC was ed. The crude product was purified by column chromatogra-
performed on RP-18 F254S (Merck, 1.15389). All moisture or phy using diethyl ether as an eluent to give 5 (670mg, 83%) as
1
air-sensitive reactions were carried out under nitrogen atmo- a yellow oil. H-NMR (CDCl3, 500MHz) δ: 2.32 (3H, s), 2.44

406
Vol. 60, No. 3
—
(3H, s), 3.07 (2H, t, J=6.6Hz), 3.56 3.59 (6H, m), 3.59 (4H, 82%). The spectroscopic data was identical to the product by
s), 3.68 (2H, t, J=4.7Hz), 4.16 (2H, t, J=4.8Hz), 7.35 (2H, acid hydrolysis described above.
d, J=8.2Hz), 7.79 (2H, d, J=8.3Hz). High resolution (HR)-
DART-MS m/z: 407.1195 ([M+H]+, Calcd for C17H27O7S2: azol-1-yl}-3,6,9-trioxaundecan-1-thiol (9) To a solution of
407.1198). 8 (100mg, 0.13mmol) in a mixture of dimethoxyethane–H2O
S-11-Azido-3,6,9-trioxaundecyl Ethanethioate (6)14) To (2:1 (v/v), 9mL), triphenyl phosphine (40mg, 0.15mmol)
11-{4-[(2-Nitro-1H-imidazol-1-yl)methyl]-1H-1,2,3-tri-
°
a solution of 5 (600mg, 1.5mmol) in dry N,N-dimethyl- was added. The resulting mixture was stirred at 40 C for
formamide (DMF) (10mL), sodium azide (480mg, 7.4mmol) 24h. The mixture was extracted with CHCl3 (3×10mL), the
°
was added. The reaction mixture was heated at 60 C while organic phase was washed with water, dried over anhydrous
stirring for 13h. After the solvent was removed, the residue MgSO4 and concentrated. The crude product was purified by
was poured into water and extracted with CH2Cl2 (3×10mL). column chromatography using CHCl3–methanol (9:1, v/v) as
1
The organic phase was washed with water and brine, dried an eluent to give 9 (74mg, 74%) as a waxy solid. H-NMR
over anhydrous Na2SO4, filtered and concentrated. the crude (CDCl3, 500MHz) δ: 1.60 (1H, t, J=8.1Hz, D2O exchange-
product was purified by column chromatography using able), 2.68 (2H, q, J=6.9Hz), 3.62 (6H, s), 3.63 (4H, s), 3.87
CHCl3–methanol (95:5, v/v) as an eluent to afford 6 as yellow (2H, t, J=4.9Hz), 4.55 (2H, t, J=5.0Hz), 5.72 (2H, s), 7.14
1
oil; yield (350mg, 86%). H-NMR (CDCl3, 500MHz) δ: 2.34 (1H, s), 7.39 (1H, s), 7.94 (1H, s). HR-DART-MS m/z: 387.1464
(3H, s), 3.10 (2H, t, J=6.6Hz), 3.40 (2H, t, J=5.1Hz), 3.61 ([M+H]+, Calcd for C14H23N6O5S: 387.1451).
(2H, t, J=6.4Hz), 3.62–3.67 (4H, m), 3.67 (4H, s), 3.69 (2H, t,
J=4.8Hz). HR-DART-MS m/z: 278.1174 ([M+H]+, Calcd for 0.047mmol) in DMF (4mL), 9 (60mg, 0.16mmol) was added.
C10H20N3O4S: 278.1175). The mixture was stirred for 18h at room temperature. After
Tricarbocyanine GPU-311 To a solution of 3 (35mg,
S-11-{4-[(2-Nitro-1H-imidazol-1-yl)methyl]-1H-1,2,3-tri- the solvent was removed, the residue was dissolved in metha-
azol-1-yl}-3,6,9-trioxaundecyl Ethanethioate (7) To a solu- nol (2mL) and then precipitated by diethyl ether (25mL).
tion of 116) (200mg, 1.3mmol) in a mixture of tert-butanol– The solid was filtered and further purified by Yamazen
H2O (1:1 (v/v), 5mL), 6 (440mg, 1.6mmol), copper(II) acetate flash liquid chromatography W-Prep 2XY system equipped
(24mg, 0.13mmol) and sodium ascorbate (52mg, 0.26mmol) with Ultra Pack ODS-SM-50B column (26mm inner diam-
were added. The resulting mixture was stirred at room tem- eter×300mm length) using 0.1M triethylammonium acetate
perature for 20h, and the residue was extracted with CHCl3 buffer (pH 7.3)/acetonitrile (30–60% in 70min) as an eluent
1
(3×10mL). The organic phase was washed with water, dried to give GPU-311 (34mg, 66%) as a dark green solid. H-NMR
—
over anhydrous Na2SO4 and concentrated. The crude product (DMSO-d6, 400MHz) δ: 1.69 (12H, s), 1.77 1.80 (10H, m),
was purified by column chromatography using CHCl3–metha- 2.54 (4H, t, J=7.1Hz), 2.66 (4H, br), 2.97 (2H, t, J=6.1Hz),
nol (19:1, v/v) as an eluent to give 7 (515mg, 91%) as a waxy 3.44 (6H, br), 3.46 (2H, t, J=5.1Hz), 3.56 (2H, t, J=6.2Hz),
1
solid. H-NMR (CDCl3, 500MHz): δ: 2.32 (3H, s), 3.07 (2H, 3.76 (2H, t, J=5.2Hz), 4.20 (4H, br), 4.47 (2H, t, J=5.2Hz),
t, J=6.5Hz), 3.59 (2H, t, J=6.6Hz), 3.61 (4H, s), 3.62 (4H, s), 5.69 (2H, s), 6.36 (2H, d, J=14.3Hz), 7.19 (1H, s), 7.25 (2H,
3.87 (2H, t, J=5.0Hz), 4.56 (2H, t, J=4.9Hz), 5.72 (2H, s), 7.14 t, J=7.4Hz), 7.41 (2H, t, J=7.7Hz), 7.46 (2H, d, J=8.0Hz),
(1H, d, J=1.1Hz), 7.39 (1H, d, J=1.2Hz), 7.95 (1H, s). HR- 7.60 (2H, d, J=7.4Hz), 7.74 (1H, s), 8.09 (1H, s), 8.76 (2H, d,
DART-MS m/z: 429.1550 ([M+H]+, Calcd for C16H25N6O6S: J=14.1Hz). HR-ESI-MS m/z: 1077.4260 ([M+2H−Na]+, Calcd
429.1556).
Bis(11-{4-[(2-nitro-1H-imidazol-1-yl)methyl]-1H-1,2,3-
for C52H69N8O11S3: 1077.4242).
Analytical HPLC was performed on a reverse phase column
triazol-1-yl}-3,6,9-trioxaundecan-1-yl) Disulfide (8) To a (Waters symmetry C18 (3.5μm particle size, 4.6mm inner
solution of 7 (240mg, 0.56mmol) in methanol (3mL), ace- diameter×75mm length), using eluent A (acetonitrile, 10%
tyl chloride (0.57mL, 8.0mmol) was added. The resulting (0min) to 90% (15min)) and eluent B (10mM ammonium for-
°
mixture was stirred at room temperature for 30h, CH2Cl2 mate buffer (pH 7.3)) at flow rate 1.0mL/min at 40 C, fitted on
and H2O (5mL each) was added to the mixture. A saturated Shimadzu LC-20AD pumps, SPD-M20A detectors, CTO-20A
NaHCO3 solution was added until pH 7. The organic layer column oven, and LCMS solution system.
was washed with water (5mL) and concentrated. The obtained
Optical Properties and Quantum Efficiency of Fluores-
crude product was purified by column chromatography using cence Optical properties of dyes were examined in methanol
CHCl3–methanol (9:1, v/v) as an eluent to give 8 as a light using a Beckmann DU 640 UV-VIS spectrophotometer and
1
yellow waxy solid (168mg, 78%). H-NMR (CDCl3, 500MHz) JASCO FP-6600 fluorescence spectrophotometer. The slit
δ: 2.86 (4H, t, J=6.5Hz), 3.60 (8H, s), 3.62 (8H, s), 3.71 (4H, t, width was 5nm for excitation and 10nm for emission. The
°
J=6.5Hz), 3.86 (4H, t, J=5.0Hz), 4.55 (4H, t, J=4.9Hz), 5.71 fluorescence measurement was performed at 25 C.
(4H, s), 7.125 (1H, s), 7.127 (1H, s), 7.386 (1H, s), 7.388 (1H, s),
7.95 (2H, s).
For determination of the quantum efficiency of fluorescence
Φ
(
Φ
fl), indocyanine green (ICG) in methanol ( =0.04) was
fl
The another procedure was as follows. To a solution of 7 used as a standard.20) Values were calculated according to the
(100mg, 0.23mmol) in dry methanol (2mL), K2CO3 (120mg, following equation.
0.87mmol) was added. The resulting mixture was stirred at
Φx /Φst =[Ast / Ax ][Dx / Dst ]
°
50 C for 2h. A saturated solution of NH4Cl (5mL) was added
and the mixture was extracted with EtOAc (3×10mL). The st: standard, x: sample. A: absorbance at the excitation wave-
organic phase was washed with water, dried over anhydrous length. D: area under the fluorescence spectra on an energy
Na2SO4 and concentrated. The obtained crude product was scale.
purified by column chromatography using CHCl3–methanol
Hydrophobicity Index (Rm) Value Rm values were de-
(9:1, v/v) as an eluent to give 8 as waxy solid; yield (74mg, termined according to the literature procedure.21) In brief,

March 2012
407
solution of test compounds were applied to the reversed-phase relative fluorescence intensity of the tumor to the muscle (T/B
thin layer chromatography plate and developed in saturated ratio) was calculated by dividing the normalized values in tu-
chambers with mobile phase (10% 5mM sodium phosphate mors by ones in the muscle.
buffer, pH 7.4 and 90% methanol). The separated spots were
HIF-1-active hypoxic tumor was detected by biolumi-
identified under visible light. The retention factor (Rf) values nescent bioimaging. For in vivo photon counting to assess
were obtained from the RP-TLC system and Rm values were bioluminescence, tumor-bearing mice were intraperitoneally
calculated by the expression: Rm=Log (1/Rf−1).
injected with 200μL of D-luciferin solution (10mg/mL in PBS,
In Vitro Fluorescence Assay of SUIT-2/HRE-Luc Cells Promega, Madison, WI, U.S.A.) and examined for biolumi-
Incubated under Different Oxygen Conditions SUIT-2/ nescence images at 20min after the injection of D-luciferin by
HRE-Luc cells18) were maintained at 37 C in 5% FCS- IVIS®-SPECTRUM.
°
’
’
Dulbecco s-modified Eagle s medium (Nacalai Tesque, Kyoto,
Japan) supplemented with penicillin (100units/mL) and strep-
Acknowledgements This research was supported in part
tomycin (100μg/mL). Hypoxic cell cultures were performed by a Grant from The Saijiro Endo Memorial Foundation for
in 5% CO2/0.1% O2 in a INVIVO2 400 hypoxia workstation Science & Technology (to K.O.), a Grant-in-Aid for Young
(Ruskinn Technology, Ltd., Bridgend, U.K.).
Scientists (B) (22790042 to K.O.), a Grant-in-Aid for Scientific
SUIT-2/HRE-Luc cells (2×105 cells/well) were seeded in a Research on Priority Areas (20011009 to H.N.) from the Japan
six-well plate and preincubated under normoxic (21% O2) or Society for the Promotion of Science, and a Grant-in-Aid for
“
hypoxic (0.1% O2) conditions for 4h, followed by incubation Scientific Research on Innovative Areas Integrative Research
”
with 0.2, 1, or 5μM of GPU-311 and 0.1% dimethyl sulfoxide on Cancer Microenvironment Network (22112009 to S.K.K.)
(DMSO) for 30min. Then, after the medium was changed, the from the Ministry of Education, Culture, Sports, Science and
cells were exposed to post-incubation for 1h in fresh medium Technology of Japan. We also thank Mr. Ryosuke Sakai, Mr.
for releasing the unbound GPU-311 from the cells, washed Takahiro Ueno, and Ms. Emi Inaba for their technical assis-
with PBS, and suspended in 200μL of radio-immunoprecipita- tance.
tion assay (RIPA) buffer. To evaluate cellular uptake including
unbound probe, the cells were washed and lysed immediately
after the probe treatment without post-incubation. Fluorescent
intensity and NIR imaging were obtained using Infinite®
F500 microplate reader (Tecan, Männedorf, Switzerland) and
IVIS®-SPECTRUM (Caliper Life Sciences, Alameda, CA,
U.S.A.), respectively. The amount of GPU-311 in the cells was
calculated by using intensity–concentration curve. The excita-
tion/emission wavelengths of filters were 740nm/780nm for
the measurement of fluorescent intensity and 710nm/800nm
for the fluorescent imaging, respectively. Statistical analysis
References
—
1) Wilson W. R., Hay M. P., Nat. Rev. Cancer, 11, 393 410 (2011).
—
2) Vaupel P., Exp. Oncol., 32, 125 127 (2010).
3) Osinsky S., Zavelevich M., Vaupel P., Exp. Oncol., 31, 80 86 (2009).
4) Kizaka-Kondoh S., Konse-Nagasawa H., Cancer Sci., 100, 1366
—
—
1373 (2009).
5) Sun X., Niu G., Chan N., Shen B., Chen X., Mol. Imaging Biol., 13,
—
399 410 (2011).
6) Chitneni S. K., Palmer G. M., Zalutsky M. R., Dewhirst M. W., J.
Nucl. Med., 52, 165 168 (2011).
7) Astner S. T., Shi K., Vaupel P., Molls M., Exp. Oncol., 32, 149 152
—
—
’
was carried out with a Student s t-test. Values of p<0.05 were
(2010).
—
8) Pham W., Medarova Z., Moore A., Bioconjugate Chem., 16, 735
considered statistically significant.
740 (2005).
In Vivo Fluorescence Imaging of SUIT-2/HRE-Luc
Xenografts by GPU-311 and Analysis of T/B Ratio All
animal experiments were performed with the approval of
Tokyo Institute of Technology and in accordance with the
Ethical Guidelines for Animal Experimentation of Tokyo
Institute of Technology. Male Balb/c nu/nu were purchased
from Oriental Yeast Co., Ltd. All mice underwent experiments
9) Lewis J. S., Achilefu S., Garbow J. R., Laforest R., Welch M. J.,
Eur. J. Cancer, 38, 2173 2188 (2002).
10) Okuda K., Yuossif B. G. M., Ueno T., Kadonosono T., Kizaka-
Kondoh S., Nagasawa H., The 2010 World Molecular Imaging
Congress, Kyoto, Japan, September 8 11, 2010.
11) Flanagan Jr. J. H., Khan S. H., Menchen S., Soper S. A., Hammer
—
“
”
—
—
R. P., Bioconjugate Chem., 8, 751 756 (1997).
—
at 6 10 weeks of age. Low fluorescence background diet was
—
12) Lee H., Mason J. C., Achilefu S., J. Org. Chem., 71, 7862 7865 (2006).
13) Song F., Peng X., Lu E., Zhang R., Chen X., Song B., J. Photochem.
Photobiol., A, 168, 53–57 (2004).
14) Zhao N., Roesler S., Kissel T., Int. J. Pharm., 411, 197 205 (2011).
15) Kolb H. C., Sharpless K. B., Drug Discov. Today, 8, 1128 1137
fed to experimental mice. SUIT-2/HRE-Luc cells suspended
in PBS (1.0×106 cells/20μL) were mixed with an equal volume
of Geltrex (Invitrogen, Carlsbad, CA, U.S.A.) and injected
into both forelegs of 7-week-old male nude mice. Mice with
—
—
(2003).
—
subcutaneous tumors of 5 15mm in diameter were used for
16) Yap S., Woodman O. L., Crack P. J., Williams S. J., Bioorg. Med.
Chem. Lett., 21, 5102 5106 (2011).
17) Zhang L., Sun L., Cui Z., Gottlieb R. L., Zhang B., Bioconjugate
Chem., 12, 939 948 (2001).
18) Kizaka-Kondoh S., Itasaka S., Zeng L., Tanaka S., Zhao T.,
Takahashi Y., Shibuya K., Hirota K., Semenza G. L., Hiraoka M.,
Clin. Cancer Res., 15, 3433 3441 (2009).
19) Kuchimaru T., Kadonosono T., Tanaka S., Ushiki T., Hiraoka M.,
experiments.
—
According to the method described previously,19) we per-
formed in vivo imaging and analysis of T/B ratio. In brief,
10nmol of GPU-311 in 100μL of PBS (pH 8.0) containing
0.2% DMSO were injected into tumor-bearing mice intrave-
nously. In vivo fluorescence images were acquired at the indi-
cated times with IVIS®-SPECTRUM (excitation filter: 710nm,
emission filter: 800nm). Photon flux/sec in the same area
(region of interest: ROI) of the tumors and the muscle in the
lower limbs were counted at the indicated time after GPU-311
injection. Each photon flux/sec was normalized by subtracting
photon flux/sec in the corresponding ROI of background. The
—
—
Kizaka-Kondoh S., PLoS ONE, 5, e15736 (2010).
20) Soper S. A., Mattingly Q. L., J. Am. Chem. Soc., 116, 3744 3752 (1994).
21) Cardoso M. G., Nelson D. L., Tavares do Amaral A., Donizete dos
—
Santos C., Pereira A. A., Barenche de Oliveira A. C., Int. J. Mol.
Sci., 3, 755 763 (2002).
—