Synthesis and evaluation of new 18F-labelled acetamidobenzoxazolone-based radioligands for imaging of the translocator protein (18 kDa, TSPO) in the brain

Article abstract of DOI:10.1039/c4ob01933d

The visualization of the activated microglia/TSPO is one of the main aspects of neuroimaging. Here we describe two new ¹⁸F-labelled molecules, 2-[5-(4-[¹⁸F]fluoroethoxyphenyl)- ([¹⁸F]2) and 2-[5-(4-[¹⁸F]fluoropr

Full text of DOI:10.1039/c4ob01933d

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
Synthesis and evaluation of new 18F-labelled
acetamidobenzoxazolone-based radioligands for
imaging of the translocator protein (18 kDa, TSPO)
in the brain†
Cite this: Org. Biomol. Chem., 2014,
12, 9621
Anjani K. Tiwari,a,b Masayuki Fujinaga,a Joji Yui,a Tomoteru Yamasaki,a Lin Xie,a
Katsushi Kumata,a Anil K. Mishra,b Yoko Shimoda,a Akiko Hatori,a Bin Ji,a
Masanao Ogawa,a,c Kazunori Kawamura,a Feng Wangd and Ming-Rong Zhang*a
The visualization of the activated microglia/TSPO is one of the main aspects of neuroimaging. Here we
describe two new 18F-labelled molecules, 2-[5-(4-[18F]fluoroethoxyphenyl)- ([18F]2) and 2-[5-(4-[18F]fluoro-
propyloxyphenyl)- ([18F]3) -2-oxo-1,3-benzoxazol-3(2H)-yl]-N-methyl-N-phenylacetamide as novel PET
ligands for imaging the translocator protein (18 kDa, TSPO) in the brain. The three-D pharmacophore
evaluation and docking studies suggested their high affinity for the TSPO and in vitro binding assays of the
TSPO showed binding affinities 6.6 0.7 nM and 16.7 2.5 nM for 2 and 3, respectively. The radiochemi-
cal yields for [18F]2 and [18F]3 were found to be 22 4% (n = 8) and 5 2% (n = 5), respectively at EOB.
The radiochemical purity for both was found ≥98% and the specific activity was in the range of 98–364
GBq µmol−1 at EOS. In vitro autoradiography with an ischemic rat brain showed significantly increased
binding on the ipsilateral side compared to the contralateral side. The specificity of [18F]2 and [18F]3 for
binding TSPO was confirmed using the TSPO ligands PK11195 and MBMP. The biodistribution patterns of
both PET ligands were evaluated in normal mice by 1 h dynamic PET imaging. In the brain, regional radio-
activity reached the maximum very rapidly within 0–4 min for both ligands, similar to (R)[11C]PK11195. The
metabolite study of [18F]2 also favoured a more favourable profile for quantification in comparison to
(R)[11C]PK11195. In summary, these data indicated that [18F]2 and [18F]3 have good potential to work as
PET ligands, therefore there are merits to use these radioligands for the in vivo evaluation in animal
models to see their efficacy in the living brain.
Received 12th September 2014,
Accepted 1st October 2014
DOI: 10.1039/c4ob01933d
www.rsc.org/obc
neuronal damage and hence, may support accelerated micro-
glial activation for immune responses in such conditions.
Introduction
One of the most focused study for positron emission tomogra-
The first clinical prototype of a PET ligand specific for
phy (PET) imaging is the activation of microglial cells, which is TSPO was [11C]PK11195 (Fig. 1);3 however, its limitations
accompanied by the expression of the translocator protein include high lipophilicity, low in vivo specific binding and an
(18 kDa, TSPO) during acute or chronic neuroinflammation.1,2 unfavourable metabolic profile for receptor quantification.4
The levels of TSPO increase markedly during neuroinflamma- These limitations motivated investigators to develop new TSPO
tion and neurodegeneration. Moreover, it has been demon- PET ligands. Up to date various new radioligands with superior
strated that TSPO functions may be important in modulating imaging characteristics have been developed, and promising
candidates were used in clinical human studies, such as
[11C]DAA1106,5 [18F]FEDAA1106,6 [11C]PBR28,7 [11C]AC-52168
aMolecular Imaging Centre, National Institute of Radiological Sciences, Inage-ku,
and [18F]DPA7149 (Fig. 1). Most of these radioligands bind
Chiba 263-8555, Japan. E-mail: zhang@nirs.go.jp; Fax: +81 43 206 3261;
TSPO with a high affinity in small animals; however, their
Tel: +81 43 382 3709
specific binding is reduced in the human brain.10,11 Moreover,
PBR28 exhibited a mixed affinity with TSPO in “binder” and
“non-binder” human subjects.12–14
bDivision of Cyclotron and Radiopharmaceutical Sciences, Institute of Nuclear
Medicine and Allied Sciences, Brig. S. K. Mazumdar Road, Delhi-110054, India
cSHI Accelerator Service Co. Ltd, Shinagawa-ku, Tokyo 141-8686, Japan
dDepartment of Nuclear Medicine, Nanjing Hospital, Affiliated to Nanjing Medical
To develop a more appropriate PET ligand, recently we deve-
loped a novel 11C-labelled acetamidobenzoxazolone skeleton,
2-[5-(4-[11C]methoxyphenyl)-2-oxo-1,3-benzoxazol-3(2H)-yl]-
University, Nanjing, China
†Electronic supplementary information (ESI) available. See DOI: 10.1039/
c4ob01933d
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 9621–9630 | 9621

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 1 Representative structures of the TSPO ligands used in clinical PET studies.
Fig. 2 Development of 18F-ligands for the TSPO.
N-methyl-N-phenylacetamide ([11C]MBMP, [11C]1; Fig. 2) that venient long-term storage as well as for transportation to
binds TSPO in vitro with a high affinity and specificity in an distant research and clinical facilities. Here, we report the
ischemic rat brain.15 However, approximately 20% of the input radiosynthesis of two new radioligands [18F]2 and [18F]3 and
of [11C]1 was detected as a metabolite in mice brains 60 min the evaluation of their potential in PET imaging of the TSPO
after injection, which hindered its further evaluation and in the brain.
application without further optimization, such as using a
different radioisotope and modifying the linker positions.
To address this issue, we synthesized the 18F-labelled ana-
logues of [11C]1 as putative TSPO ligands for PET as follows: 2-[5-
Results and discussion
Chemistry
(4-[18F]fluoroethoxyphenyl)- ([18F]FEBMP, [18F]2) and 2-[5-(4-[18F]-
fluoropropyloxyphenyl)- ([18F]FPBMP, [18F]3) -2-oxo-1,3-benzoxazol-
3(2H)-yl]-N-methyl-N-phenylacetamide (Fig. 2). Because of the
molecular similarity and bioisosteric property of O-CH2CH2F
and O-CH2CH2CH2F with O-CH3 groups, compounds 2 and 3
may display similar affinities for TSPO with 1. Compounds 1–3
are lipophilic and therefore may pass readily through the blood-
brain barrier (BBB). Moreover, the [18F]fluoroalkoxy substi-
tution may improve the pharmacokinetics and metabolism,16
and the longer half-life (110 min vs. 20 min) of 18F allows time
for complex chemical syntheses and emits lower positron
energy (650 KeV vs. 960 KeV), which is appropriate for con-
Two novel fluoroalkoxy analogues, 2 and 3, were synthesized
by heating precursor 4 15 with the respective tosylates 5 and 6,
which were synthesized by reacting alcohols 7 and 8 with
4-methylbenzenesulfonyl chloride in the presence of K2CO3 at
70 °C for 4–6 h (Scheme 1). Their respective yields ranged from
59% to 65%. The chemical purities of these compounds were
determined using HPLC (ESI Fig. 1†). (R)[11C]PK11195 was syn-
thesized as per previous literature.3 The radiochemical purity
of (R)[11C]PK11195 was 99% and the specific activity was calcu-
lated as 74 8 GBq µmol−1 (n = 3) at EOS.
9622 | Org. Biomol. Chem., 2014, 12, 9621–9630
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Scheme 1 Chemical synthesis of 2 and 3.
Fig. 3 2D/3D TSPO–ligand interactions.
Computational docking of receptor-ligand binding analysis
The ligand–protein interaction through the docking analysis of
molecules 1–3 confirmed short-range interactions with TSPO,
including π–π interactions among the aromatic residues such
as Tyr28, Trp27, Phe19 and Trp99. These molecules were
buried in the hydrophobic pocket in the transmembrane
region (Fig. 3). Compounds 1–3 exhibit similar pharmacophore
features and interactions with the binding pocket of the TSPO.
This was confirmed by the Gscore of docking (ESI Table 1†)
and the same interactions with the amino acid residues in the
diagrams of the 2D and 3D ligand–protein interactions. These
promising data prompted us to analyse these ligands for
in vitro receptor binding.
In vitro binding assay
We used rat brain homogenates to determine the inhibition
constants (Ki) of compounds 1–3 for the TSPO by measuring
the competition for binding with (R)[11C]PK11195, which is a
selective radioligand for TSPO in the concentration range of
0.1 nM to 10 µM (Fig. 4). These data showed that the percen-
tage of specific binding was around 80% for compounds 1–3.
Further Ki was calculated as shown in Table 1, and 2 and 3
Fig. 4 Competitive binding curves for the novel TSPO ligands.
bind TSPO with a high (nanomolar) affinity, thus confirming
the results of the computer modelling study. In particular,
2 bound TSPO with a similar affinity to that of 1 and PK11195,
suggesting that substituting O-CH3 with O-CH2CH2F in the
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 9621–9630 | 9623

View Article Online
Paper
Organic & Biomolecular Chemistry
and the reaction mixture was immediately heated to produce
[18F]9 or [18F]10, which was distilled, dried and trapped in a
solution of 4 and NaOH in DMF. The purification of [18F]9
(boiling point 71.5 °C) or [18F]10 (boiling point 101 °C) by dis-
tillation achieved a radiochemical purity of >95%.20 Moreover,
this method removed all non-volatile impurities such as metal
ions from the cyclotron target, unreacted [18F]F−, and the
phase-transfer catalyst, Kryptofix 222/K2CO3. After trapping
each reagent, the reaction mixture was heated at 90 °C or
120 °C for 10 min. Following HPLC purification, the radio-
chemical yields of [18F]2 and [18F]3 were 22 4% (n = 8) and
Table 1 In vitro binding affinity for the TSPO and the lipophilicity of
compounds 1–3
Lipophilicity
Binding
affinity (Ki)
TSPOa
Experimentally
measured
log Db
Calculated
Compound
c log Dc
2
3
1
6.6 0.7
16.7 2.5
3.9 0.6
4.4 0.4
3.4
3.5
3.4
3.7
3.5
3.7
3.3
5.1
PK11195
a Binding affinities represent the mean standard error of the mean
(SE) of triplicate samples of brain homogenates. b log D values were
measured in octanol/phosphate buffer (pH = 7.4) using the shake flask
method (n = 3; maximum range, 5%). c c log D values were calculated
using the Pallas 3.4 software.
5
2% (n = 5), respectively, and were based on [18F]F−, which
was corrected for physical decay during the reaction times of
64 5 and 61 3 min from the end of bombardment,
respectively.
The identities of [18F]2 and [18F]3 were confirmed by coin-
jection with the corresponding unlabelled 2 and 3 using
reverse phased-analytical HPLC (ESI Fig. 2†). The radiochemi-
cal purities of the final products were >98% and unreacted 4
was not detected in the HPLC chromatograms. Moreover, the
final products did not undergo radiolysis at room temperature
for 120 min. The analytical results were consistent with our in-
house quality specifications of radiopharmaceuticals.
benzene ring did not significantly alter its affinity for the
TSPO. The substitution of O-CH3 with O-CH2CH2CH2F
decreased its affinity. However, this level of binding affinity for
3 was considered appropriate for imaging brain receptors
according to the results of our previous study.17
The computed lipophilicities at pH = 7.4 (c log D) of [18F]2
and [18F]3 were 3.4 and 3.5, respectively, and the corres-
ponding experimental values of 3.5 and 3.7 (log D) were deter-
mined using the shake flask method (Table 1).18 The
lipophilicity values of both ligands were similar to that of
[11C]1 and lower than that of (R)[11C]PK11195, indicating that
they may permeate the BBB as predicted for [11C]1.
In vitro autoradiography of an ischemic rat brain by using
[18F]2 and [18F]3
Because of the low density of the TSPO in a normal brain, a
well-characterized ischemic model was used to visualize the
in vitro binding of [18F]2 and [18F]3 to the TSPO. In this model,
neuroinflammation accompanied by microglial activation
Radiochemistry
Two [18F]fluoroalkoxy ligands [18F]2 and [18F]3 were syn- increased the TSPO density in the ischemic area.21 Fig. 5
thesized by reacting 4 with 1-bromo-2-[18F]fluoroethane ([18F]9) shows the in vitro autoradiograms of [18F]2 and [18F]3 in the
or
1-bromo-3-[18F]fluoropropane
([18F]10),
respectively ischemic rat brain. In the control section (A), radioactivity was
(Scheme 2). These labelling agents were prepared by [18F]fluor- high in the ipsilateral side (right) compared with the contralat-
ination of 2-bromoethyltrifluoromethanesulfonate (11) or eral side (left). However, these differences were abolished by
3-bromopropyltrifluoromethanesulfonate (12) with [18F]F− incubation with an excess of unlabelled 1 (B) or PK11195 (C).
using a synthetic unit developed in our laboratory.19 A solution
Quantification of the autoradiograms shows that the ratios
of 11 or 12 in 1,2-dichlorobenzene was added to dry [18F]F−, of the radioactive signals in the ipsilateral side compared with
Scheme 2 Radiosynthesis of [18F]2 and [18F]3.
9624 | Org. Biomol. Chem., 2014, 12, 9621–9630
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 5 Representative in vitro autoradiograms of [18F]2 and [18F]3 in ischemic rat brains (n = 4). Arrows indicate the ischemic areas. (A) [18F]2 or [18F]3
only; (B) incubation with unlabelled 1; (C) incubation with PK11195.
the contralateral side were 3.1 for [18F]2 and 2.1 for [18F]3, and specific binding of [18F]2 and [18F]3 to the TSPO. Further,
the values for each were approximately 1.0 in the presence of [18F]2 bound TSPO with a higher specificity compared with
unlabelled 1 or PK11195 (Fig. 6). Compared with the controls [18F]3.
([18F]2 or [18F]3 only), competition for the binding of [18F]2 or
[18F]3 with 1 or PK11195 significantly reduced the ipsilateral
Analysis of the dynamic PET images
binding (−91.2% by 1 and −84.6% by PK11195 competing To determine the kinetic parameters related to the distribution
with [18F]2, and −81.2% by 1 and −67.4% by PK11195 compet- and uptake of [18F]2 and [18F]3 in vivo in comparison to
ing with [18F]3), although it was lower for the contralateral (R)[11C]PK11195, dynamic PET imaging was performed. The
binding (−80.3% by 1 and −67.8% by PK11195 competing PET data acquired between 0 and 60 min, 0–4 min, 4–10 min,
with [18F]2, and −50.1% by 1 and −23.7% by PK11195 compet- 10–16 min and 55–60 min were summed, and the images are
ing for [18F]3). These results demonstrate the high in vitro shown in Fig. 7. The reason for taking more summation at the
initial phase, was because both ligands showed a fast uptake
in the brain, within a few minutes after injection.
The PET dynamic scans showed the fast accumulation of
radioactivity in the brain within 0–4 min and similarly in
major TSPO related organs such as the lungs and heart. The
biodistribution patterns of [18F]2 and [18F]3 were found to be
similar to the known PET ligand (R)[11C]PK11195. Both PET
ligands [18F]2 and [18F]3 were able to cross the BBB very
rapidly. Further, the uptake in the peripheral organs decreased
rapidly till 15 min and after that decreased slowly till 60 min.
The uptake value in the kidney reached a maximum and main-
tained a plateau parallel to the x axis. Comparative time-
activity curves (TACs) were also prepared for other organs and
for bones (Fig. 8). The TACs data showed that maximum
uptake in brain was 2.5%–3.0% ID g−1 for [18F]2 and [18F]3,
which decreased rapidly till 15 min and reached a level of
1.0% ID g−1 at 60 min, similar to (R)[11C]PK11195. Bone
uptake was also measured for these ligands as they were
Fig. 6 Quantitation of the autoradiograms of [18F]2 and [18F]3. The con-
labelled with 18F. The accumulation of radioactivity in the
centrations of radioactivity in brain regions are expressed as photo-
bone was found to be 1.4% ID g−1 for [18F]3 as compared to
stimulated luminescence (PSL) mm−2. *P < 0.05, [18F]2: control vs. treat-
1.2% ID g−1 for [18F]2 at 60 min after the injection
(ESI Fig. 3†).
ment with 1 or PK11195, **P < 0.05, [18F]3: control vs. treatment with
1 or PK11195.
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 9621–9630 | 9625

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 7 Dynamic PET images of [18F]2, [18F]3 and (R)[11C]PK11195 at different time intervals (0–60 min). Arrows indicate the brains.
Finally, on the basis of the better in vitro properties (Ki & than [18F]2 and [18F]3. The brain uptake was found similar in
autoradiography) and lower bone uptake in vivo, [18F]2 was the order of [18F]2 > [18F]3 > (R)[11C]PK11195 (ESI Fig. 5†),
analysed for metabolic stability.
which is the main requisite for in vivo imaging. On the other
hand, a previous study indicated the fast metabolism of
[11C]PK11195 in rat plasma along with five radiometabolites.22
The HPLC analysis showed that the retention time of intact
[11C]PK11195 and its radiolabelled non-polar metabolites were
very near, which suggested their similar capabilities for cross-
ing the BBB in rats. Compared to (R)[11C]PK11195, [18F]2 was
not metabolized into non-polar metabolites in the mice
plasma and brain.
Determination of radiolabelled metabolites by HPLC
In the brain [18F]2 showed intactness in the range of 85% to
90% at 15–30 min but in the plasma, the intactness of the
original ligand decreased very rapidly. The HPLC charts of
[18F]2 in mice plasma and brains are shown in Fig. 9. Only one
polar metabolite during 60 min studies was found in both the
plasma and brain. The polarity of this metabolite may not
allow it to cross the BBB and even if so, it did not stay in the
brain due to its hydrophilicity. This may give a good opportu-
nity to evaluate the specific binding to TSPO in the brain by
using the quantification studies of [18F]2.
Conclusion
The comparison of the well-known ligand (R)[11C]PK11195 The rational design of the new 18F-labelled ligands for the PET
with [18F]2 and [18F]3 was performed in similar experimental imaging of TSPO based on our previous study of a 11C-labelled
conditions. The in vitro autoradiography of (R)[11C]PK11195 acetamidobenzoxazolone-based TSPO ligand led to the suc-
(ESI Fig. 4†) showed comparable, or even more specific results cessful synthesis and high purity preparation of [18F]2 and
9626 | Org. Biomol. Chem., 2014, 12, 9621–9630
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
Fig. 8 Time–activity curves of [18F]2, [18F]3 and (R)[11C]PK11195 in organs expressed as radioactivity concentration versus time.
Experimental section
Chemicals and instrumentation
All chemicals were purchased from commercial sources.
Melting points were determined using a Yanaco MP-500P
micromelting point apparatus. The 1H-NMR (300 MHz) spectra
were recorded using a JEOL-AL-300 NMR spectrometer with
tetramethylsilane as an internal standard. The high-resolution
fast atom bombardment mass spectra [HRMS(FAB)] were
acquired using a JEOL JMS-Sx 102A spectrometer. HPLC separ-
ation and analysis were performed using a JASCO HPLC
system. Effluent radioactivity was monitored using a NaI (Tl)
scintillation detector system, and radioactivity was measured
during the synthesis and animal studies using an Aloka Curie-
meter. To measure the in vitro binding affinity for TSPO,
(R)[11C]PK11195 was synthesized as described previously.3
Fig. 9 Representative radio-HPLC chromatograms of mice plasma
sample and brain homogenates obtained at 15 min after injection of
[
18F]2.
Chemical synthesis
2-[5-(4-Fluoroethoxyphenyl)-2-oxo-1,3-benzoxazol-3(2H)-yl]-N-
methyl-N-phenylacetamide (2). 2-Fluoroethyl 4-methylbenze-
nesulfonate18 (5, 56 mg, 0.30 mmol) was added to a solution
of 2-[5-(4-hydroxyphenyl)-2-oxo-1,3-benzoxazol-3(2H)-yl]-N-methyl-
N-phenylacetamide15 (4, 100 mg, 0.27 mmol) and K2CO3
(57 mg, 0.41 mmol) in anhydrous DMF (5 mL). The reaction
mixture was heated for 4 h at 70 °C, quenched with water and
extracted with ethyl acetate. The organic layer was washed with
[18F]3. These radioligands were stable and bound TSPO with a
high affinity and specificity in an ipsilateral ischemic rat brain
in vitro. The PET dynamic studies also validated its distri-
bution and selectivity for TSPO in the brain regions. These
findings provide a compelling argument for the testing of
these novel PET ligands in vivo with a particular focus on the
brain imaging of putative binders and non-binders.
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 9621–9630 | 9627

View Article Online
Paper
Organic & Biomolecular Chemistry
brine, dried over Na2SO4 and evaporated under reduced plete, the reaction mixture was heated at 120 °C for 10 min.
pressure. The residue was purified using silica gel column HPLC purification was completed using a mobile phase of
chromatography with n-hexane–ethyl acetate (1/2, v/v) to give 2 MeCN–H2O–Et3N (65/35/0.01, v/v/v) at a flow rate of 5.0 mL
(58 mg, 52.0% yield) as a colourless solid with the properties min−1 for [18F]3. The tR of [18F]3 was 8.8 min for purification
as follows: mp 170–172 °C; 1H NMR (CDCl3, δ) 3.32 (3H, s), and 8.3 min for analysis on HPLC.
4.27 (2H, dt, J = 4.2, 27.9 Hz), 4.35 (2H, s), 4.79 (2H, dt, J = 4.0,
47.3 Hz), 6.99–7.02 (3H, m), 7.22 (1H, s), 7.23 (2H, d, J = 8.8
Computational analysis
Hz), 7.33–7.36 (2H, m), 7.44–7.54 (5H, m) and HRMS (FAB) The human TSPO (PBR) sequence was retrieved from the NCBI
C24H22O4N2F, calculated 421.1564; found, 421.1516.
database. The modelling of the receptor was performed using
2-[5-(4-Fluoropropyloxyphenyl)-2-oxo-1,3-benzoxazol-3(2H)-yl]- the Prime Homology Modelling tool (Schrödinger). The
N-methyl-N-phenylacetamide (3). 2-Fluoropropyl 4-methylben- docking study was initiated by joining the proteins and
zenesulfonate18 (6; 70 mg, 0.30 mmol) was added to a solution ligands. The pharmacophore modelling studies were per-
of 4 (100 mg, 0.27 mmol) and K2CO3 (57 mg, 0.41 mmol) in formed using PHASE, version 3.4, (Schrödinger, LLC,
anhydrous DMF (5 mL). The reaction mixture was heated for New York, NY 2012).
6 h at 70 °C, quenched with water and extracted with ethyl
acetate. The organic layer was washed with brine, dried over
Measurement and computation of lipophilicity
Na2SO4 and evaporated under reduced pressure. The residue The log D values were measured by mixing [18F]2 or [18F]3
was purified using silica gel column chromatography with (radiochemical purity, 100%; about 200 000 cpm) with
n-hexane–ethyl acetate (1/1, v/v) to give 3 (70 mg, 60.2% yield) n-octanol (3.0 g) and sodium-phosphate buffer (PBS 3.0 g,
as a colourless solid with the properties as follows: mp 0.1 M, pH 7.4) in a test tube, which was vortexed for 3 min at
148–152 °C; 1H-NMR (CDCl3, δ) 2.14–2.27 (2H, m), 3.32 (3H, s), room temperature, followed by centrifugation at 3500 rpm for
4.15 (2H, t, J = 6.0 Hz), 4.35 (2H, s), 4.68 (2H, dt, J = 5.7,47.3 5 min. An aliquot of 1 mL PBS and 1 mL n-octanol was
Hz), 6.96–6.99 (3H, m), 7.22–7.24 (2H, m), 7.35 (2H, t, J = removed, weighed and counted. Samples from the remaining
6.6 Hz), 7.44–7.54 (5H, m) and HRMS (FAB) calculated for organic layer were removed and repartitioned until consistent
C25H24O4N2F, 435.1720; found, 435.1703.
log D values were obtained. The log D value was calculated
from the ratio of cpm g−1 of n-octanol to that of PBS and
expressed as log D = log[cpm g−1 (n-octanol)/cpm g−1 (PBS)].
Radiosynthesis
2-[5-(4-[18F]Fluoroethoxyphenyl)-2-oxo-1,3-benzoxazol-3(2H)-yl]- All assays were performed in triplicate.
N-methyl-N-phenylacetamide ([18F]2). [18O]H2O (95%) was
Animals
used for irradiation. [18F]HF was recovered from the cyclotron
target, separated from [18O]H2O, concentrated on a short QMA Male Sprague-Dawley (SD) rats were purchased from Japan SLC
column and eluted with 400 µL of a solution containing and were maintained in an experimental animal rearing
aqueous K2CO3 (110 mg/8 mL), Kryptofix®222 (330 mg) in room, which was maintained under optimal conditions with a
MeCN (8 mL). Aqueous [18F]KF was transferred to a reaction 12/12 h dark/light cycle and were handled in accordance with
vial and evaporated to remove H2O and MeCN at 110 °C for the recommendations of the National Institute of Health and
15 min. After 11 (7 µL) in 1,2-dichlorobenzene (150 µL) was institutional guidelines of the National Institute of Radiologi-
added to the vial, the reaction mixture was heated at 130 °C to cal Sciences (NIRS). The Animal Ethics Committee of the NIRS
give [18F]9, which was distilled using nitrogen gas (10 mL approved the experiments conducted at the NIRS.
min−1) for 2 min and trapped in a solution of 4 (1.0 mg) and
In vitro assays of TSPO binding
NaOH (5 µL, 0.5 M) in anhydrous DMF (300 µL) at −15 to
−20 °C. After the radioactivity plateaued, the reaction mixture The TSPO binding assay was performed as per the previous
was heated at 90 °C for 10 min. HPLC purification was com- literature.21 Briefly, three rats were sacrificed by cervical dislo-
pleted using a mobile phase of MeCN–H2O–Et3N (6.0/4.0/0.01, cation with 5% (v/v) isoflurane anaesthesia. The brains were
v/v/v) at a flow of 5.0 mL min−1. The radioactive fraction corres- removed quickly and homogenized in ice-cold Tris-HCl buffer
ponding to the desired product was collected in a sterile flask, (50 mM, pH 7.4) containing 120 μM NaCl. The homogenate
evaporated to dryness in vacuo, redissolved in 3 mL of sterile was centrifuged at 40 000g for 15 min at 4 °C. Subsequently
normal saline and passed through a 0.22 μm Millipore filter. the supernatant was discarded and then the pellet was resus-
The tR of [18F]2 was 9.7 min for purification and 9.4 min for pended, homogenized, and centrifuged under the same con-
analysis on HPLC.
ditions. This procedure was repeated twice. The pellet was
2-[5-(4-[18F]Fluoropropyloxyphenyl)-2-oxo-1,3-benzoxazol-3(2H)- resuspended in 50 mM Tris-HCl buffer (pH 7.4) at a concen-
yl]-N-methyl-N-phenylacetamide ([18F]3). 12 (10 µL) in 1,2- tration of 100 mg of the original wet tissue per mL and used
dichlorobenzene (150 µL) was added to a dried [18F]KF-con- for the binding assays.
taining vial. The reaction mixture was heated at 150 °C for
The crude mitochondrial preparations (100 µL) were incu-
2 min to produce [18F]10, which was distilled and trapped in a bated with (R)[11C]PK11195 (final concentration 2.5 nM) and
solution of 4 (1.0 mg) and NaOH (5 µL, 0.5 M) in anhydrous various concentrations (0.3 nM–3 µM) of ligands 1–3 and
DMF (300 µL) at −15 to −20 °C. After the trapping was com- PK11195 in a total volume of 1 mL Tris-HCl buffer at room
9628 | Org. Biomol. Chem., 2014, 12, 9621–9630
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Paper
temperature for 30 min, and the reaction was terminated by time intervals after the injection of [18F]2 (5.18 0.29 MBq/
rapid filtration through Whatman GF/C glass fibre filters pre- 0.02–0.04 nmol), [18F]3 (4.85 0.17 MBq/0.018–0.034 nmol)
treated with 0.3% polyethylenimine, using an M-24 cell har- and (R)[11C]PK11195 (37.0 MBq
2.0 MBq/0.02–0.04 nmol)
vester (Brandel). The filters were washed three times with 5 mL through the tail vein. All image frames were summed and the
of ice-cold Tris-HCl buffer (50 mM), and the filter-bound radio- regions of interest (ROI) were drawn over the brain, heart,
activity was counted using an auto-gamma scintillation liver, spleen, kidneys, stomach and bone of each mouse within
counter. The nonspecific binding was determined using a four time intervals of 0–4 min, 4–10 min, 10–16 min and
saturated concentration (10 µM) of PK11195. The Ki for the 55–60 min. The time activity curves (TACs) for the brain, heart,
TSPO of each ligand was determined according to the follow- blood, kidney, lung, liver and bone were generated from the
ing equation:
dynamic PET data in order to parameterize the radioactivity
uptake, clearance and distribution in healthy mice. The radio-
activity (% ID g−1) was estimated as the ratio of the regional
activity concentration normalized by the injected dose and the
weight of the animal to give the PET-generated biodistribution
pattern over various critical organs.
The data modelling for the PET scans was performed into
three-dimensional (3D) sinograms, which were changed into
two-dimensional (2D) sinograms by Fourier rebinning.
Dynamic image reconstruction was done by filtered back-pro-
jection using the Hanning’s filter with a Nyquist cut-off fre-
quency of 0.5 cycle per pixel. The PET images were analyzed
using ASIPro VM™ (Analysis Tools and System Setup/Diagnos-
tics Tool; Siemens Medical Solutions) with reference to the MR
imaging template.
IC50
Ki ¼
½Lꢀ
1 þ
Kd
where [L] is the concentration of (R)[11C]PK11195 (nM). The
dissociation constant (Kd) of (R)[11C]PK11195 for the TSPO was
obtained using the Scatchard plot analysis.
In vitro autoradiography of an ischemic rat brain
Mild focal ischemia in rat brains was produced by intralumi-
nal occlusion of the middle cerebral artery for 30 min using an
intraluminal thread model.23,24 Briefly, a Sprague Dawley rat
(8–9 weeks of age, 240–330 g) was anesthetized with 4% (v/v)
isoflurane and maintained with 1.8% isoflurane, and the right
internal carotid artery was ligated. A 4.0-monofilament nylon
suture coated with silicon was inserted (16–18 mm) into the
internal carotid artery to the middle cerebral artery branches.
The neck incision was closed with a silk suture. After regaining
consciousness from the anaesthesia, the rats were anaesthe-
tized again after 30 min, and the filament was carefully
removed for reperfusion. Throughout the surgery, the body
temperature was monitored and maintained at an optimal
level. All rats (n = 4) were employed for experiments 7 days
after the surgery.
Radiolabelled metabolite analysis
The ddY mice were intravenously injected with [18F]2 (7.4 MBq
per mouse) through the tail vein. The animals were sacrificed
by cervical dislocation at 5, 15, 30 or 60 min (n = 3 for each
point). Blood (0.7–1.0 mL) and entire brain samples were
obtained and treated as reported previously.15 The supernatant
of the plasma and brain homogenate was analyzed according
to the following conditions: Capcell Pack UG80 C18 column,
4.6 mm i.d. × 250 mm; MeCN–H2O–Et3N, 7/3/0.01 (v/v/v); flow
rate, 1.0 mL min−1. The percentages of the unchanged [18F]2
were calculated, and at the same time, the radioactivity frac-
tions in the HPLC waste solution were measured using a 1480
Wizard 3” auto-gamma counter.
The coronal sections (10 μm) were prepared from frozen rat
brains using a cryostat (HM560, Carl Zeiss, Germany). Brain
sections were preincubated for 20 min in 50 mM Tris-HCl
buffer (pH 7.4), and the sections were then incubated for
30 min at room temperature in fresh buffer containing [18F]2
or [18F]3 (1.5 MBq, 0.08 nM). For the inhibition studies,
unlabelled 1 (10 µM) or PK11195 (10 µM) was incubated with
[18F]2 or [18F]3. After incubation, the brain sections were
washed (3 × 2 min) with cold buffer, dipped in cold distilled
water and dried with cold air. The sections were placed in
contact with imaging plates (BAS-MS2025, Fujifilm), and the
autoradiograms were analysed using a bio-imaging system
(BAS5000, Fujifilm). The concentrations of radioactivity deter-
mined in the brain sections were expressed as photo-stimu-
Acknowledgements
The authors thank the staff of the National Institute of Radio-
logical Sciences for their support for radiosynthesis and
animal experiments. In addition the first author would like to
thank the Japan Society for the Promotion of Science for
support.
lated luminescence (PSL) mm−2
.
PET study and image analysis
References
A small-animal PET scanner from Siemens Medical Solutions,
was used for imaging. Normal mice were anesthetized during
the scan and the animals’ body temperature was maintained
with a 40 °C water circulation system (T/Pump TP401, Gaymar
Industries). The emission scans were acquired at different
1 V. Papadopoulos, M. Baraldi, T. R. Guilarte, T. B. Knudsen,
J. J. Lacapère, P. Lindemann, M. D. Norenberg, D. Nutt,
A. Weizman, M. R. Zhang and M. Gavish, Trends Pharmacol.
Sci., 2006, 27, 402–409.
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 9621–9630 | 9629

View Article Online
Paper
Organic & Biomolecular Chemistry
2 P. Casellas, S. Galiegue and A. S. Basile, Neurochem. Int., 14 D. R. Owen, O. W. Howell, S. P. Tang, L. A. Wells,
2002, 40, 475–486.
I. Bennacef, M. Bergstrom, R. N. Gunn, E. A. Rabiner,
M. R. Wilkins, R. Reynolds, P. M. Matthews and
C. A. Parker, J. Cereb. Blood Flow Metab., 2010, 30,
1608–1618.
3 R. Camsonne, C. Crouzel, D. Comar, M. Maziere,
C. Prenant, J. Sastre, M. A. Moulin and A. Syrota, J. Labelled
Compd. Radiopharm., 1984, 21, 985–991.
4 F. Chauveau, H. Boutin, N. Van Camp, F. Dollé and 15 A. K. Tiwari, J. Yui, M. Fujinaga, K. Kumata, Y. Shimoda,
B. Tavitian, Eur. J. Nucl. Med. Mol. Imaging, 2008, 35,
2304–2319.
T. Yamasaki, L. Xie, A. Hatori, J. Maeda, N. Nengaki and
M. R. Zhang, J. Neurochem., 2014, 129, 712–720.
5 M. R. Zhang, T. Kida, J. Noguchi, K. Furutsuka, J. Maeda, 16 A. Varrone, C. Steiger, M. Schou, A. Takano, S. J. Finnema,
T. Suhara and K. Suzuki, Nucl. Med. Biol., 2003, 30,
513–519.
D. Guilloteau, B. Gulyás and C. Halldin, Synapse, 2009, 63,
871–880.
6 M. R. Zhang, J. Maeda, M. Ogawa, J. Noguchi, T. Ito, 17 M. Fujinaga, T. Yamasaki, J. Maeda, J. Yui, L. Xie, Y. Nagai,
Y. Yoshida, T. Okauchi, S. Obayashi, T. Suhara and
N. Nengaki, A. Hatori, K. Kumata, K. Kawamura and
K. Suzuki, J. Med. Chem., 2004, 47, 2228–2235.
M. R. Zhang, J. Med. Chem., 2012, 55, 11042–11051.
7 E. Briard, S. S. Zoghbi, M. Imaizumi, J. P. Gourley, 18 M. Fujinaga, T. Yamasaki, J. Yui, A. Hatori, L. Xie,
H. U. Shetty, J. Hong, V. Cropley, M. Fujita, R. B. Innis and
K. Kawamura, C. Asagawa, K. Kumata, T. Fukumura and
V. W. Pike, J. Med. Chem., 2008, 51, 17–30.
M. R. Zhang, J. Med. Chem., 2012, 55, 2342–2352.
8 M. R. Zhang, K. Kumata, J. Maeda, K. Yanamoto, A. Hatori, 19 M. R. Zhang, A. Tsuchiyama, T. Haradahira, Y. Yoshida,
M. Okada, M. Higuchi, S. Obayashi, T. Suhara and
K. Suzuki, J. Nucl. Med., 2007, 48, 1853–1861.
K. Furutsuka and K. Suzuki, Appl. Radiat. Isot., 2002, 57,
335–342.
9 M. L. James, R. R. Fulton, J. Vercoullie, D. J. Henderson, 20 A. A. Wilson, J. N. Dasilva and S. S. Houle, Appl. Radiat.
L. Garreau, S. Chalon, F. Dollé, S. Selleri, D. Guilloteau and
M. Kassiou, J. Nucl. Med., 2008, 49, 814–822.
10 N. Arlicot, J. Vercouillie, M. J. Ribeiro, C. Tauber, Y. Venel,
J. L. Baulieu, S. Maia, P. Corcia, M. G. Stabin, A. Reynolds,
M. Kassiou and D. Guilloteau, Nucl. Med. Biol., 2012, 39,
570–578.
11 M. Fujita, M. Imaizumi, S. S. Zoghbi, Y. Fujimura,
A. G. Farris, T. Suhara, J. Hong, V. W. Pike and R. B. Innis,
Neuroimage, 2008, 40, 43–52.
12 F. Dollé, C. Luus, A. Reynolds and M. Kassiou, Curr. Med.
Chem., 2009, 16, 2899–2923.
Isot., 1995, 46, 765–770.
21 K. Kumata, J. Yui, A. Hatori, M. Fujinaga, K. Yanamoto,
T. Yamasaki, K. Kawamura, H. Wakizaka, N. Nengaki,
Y. Yoshida, M. Ogawa, T. Fukumura and M. R. Zhang,
J. Med. Chem., 2011, 54, 6040–6049.
22 A. Roivainen, K. Någren, J. Hirvonen, V. Oikonen, P. Virsu,
T. Tolvanen and J. O. Rinne, Eur. J. Nucl. Med. Mol.
Imaging, 2009, 36, 671–682.
23 J. Yui, J. Maeda, K. Kumata, K. Kawamura, K. Yanamoto,
A. Hatori, T. Yamasak, N. Nengaki, M. Higuchi and
M. R. Zhang, J. Nucl. Med., 2010, 51, 1301–1309.
13 D. R. Owen, R. N. Gunn, E. A. Rabiner, I. Bennacef, 24 J. Yui, A. Hatori, K. Kawamura, K. Yanamoto, T. Yamasaki,
M. Fujita, W. C. Kreisl, R. B. Innis, V. W. Pike, R. Reynolds,
P. M. Matthews and C. A. Parker, J. Nucl. Med., 2011, 52,
24–32.
M. Ogawa, Y. Yoshida, K. Kumata, M. Fujinaga,
N. Nengaki, T. Fukumura, K. Suzuki and M. R. Zhang,
NeuroImage, 2011, 54, 123–130.
9630 | Org. Biomol. Chem., 2014, 12, 9621–9630
This journal is © The Royal Society of Chemistry 2014