Stoichiometry-focused 18F-labeling of alkyne-substituted oligodeoxynucleotides using azido([18F]fluoromethyl)benzenes by Cu-catalyzed Huisgen reaction

Article abstract of DOI:10.1016/j.bmc.2010.11.033

A novel method for ¹⁸F-radiolabeling of oligodeoxynucleotides (ODNs) by a Cu-catalyzed Huisgen reaction has been developed by using the lowest possible amount of the precursor biomolecule for the realization of stoichiometry-oriented PET (posit

Full text of DOI:10.1016/j.bmc.2010.11.033

Bioorganic & Medicinal Chemistry 19 (2011) 249–255
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Stoichiometry-focused 18F-labeling of alkyne-substituted
oligodeoxynucleotides using azido([18F]fluoromethyl)benzenes
by Cu-catalyzed Huisgen reaction
Takeshi Kuboyama a,b, Motoi Nakahara a, Masafumi Yoshino c, Yilong Cui b, Takeo Sako b, Yasuhiro Wada b,
b,
b,
b
Takeshi Imanishi a, Satoshi Obika a, , Yasuyoshi Watanabe , Masaaki Suzuki , Hisashi Doi
⇑
⇑
⇑
a Graduate School of Pharmaceutical Sciences, Osaka University, 1-6 Yamadaoka, Suita, Osaka 565-0871, Japan
b RIKEN Center for Molecular Imaging Science, 6-7-3 Minatojima-minamimachi, Chuo-ku, Kobe, Hyogo 650-0047, Japan
c GeneDesign, Inc., 7-7-20 Saitoasagi, Ibaraki, Osaka 567-0085, Japan
a r t i c l e i n f o
a b s t r a c t
A novel method for 18F-radiolabeling of oligodeoxynucleotides (ODNs) by a Cu-catalyzed Huisgen reaction
has been developed by using the lowest possible amount of the precursor biomolecule for the realization of
stoichiometry-oriented PET (positron emission tomography) chemistry. Under the optimized cyclization
conditions of p- or m-azido([18F]fluoromethyl)benzene and alkyne-substituted ODN (20 nmol) at 40 °C
for 15 min in the presence of CuSO4, TBTA [tris((1-benzyl-1H-1,2,3-triazol-4-yl)methyl)amine], and
sodium ascorbate (2:1:2), the synthesis of 18F-labeled ODNs with sufficiently high radioactivities of
Article history:
Received 29 September 2010
Revised 9 November 2010
Accepted 9 November 2010
Available online 19 November 2010
Keywords:
Click chemistry
Isotopic labeling
Oligonucleotides
2.1–2.5 GBq and specific radioactivities of 1800–2400 GBq/
and human PET studies.
lmol have been accomplished for use in animal
Ó 2010 Elsevier Ltd. All rights reserved.
PET (positron emission tomography)
Radiochemistry
1. Introduction
groups, have been reported.2b 18F radionuclides have a moderate
half-life of 109.8 min, which make them suitable for use in pharma-
cokinetic studies, and relatively high specific radioactivity; further,
they form chemically stable covalent bonds at the labeling position.
For these reasons, 18F radionuclide is widely used in most PET insti-
tutes and clinics.
Positron emission tomography (PET) is a powerful noninvasive
molecular imaging technique for the in vivo investigation of the dis-
tribution and dynamics of bioactive molecules labeled with posi-
tron-emitting radionuclides in the brain, heart, and other organs.
PET has been extensively used for clinical diagnoses and drug devel-
opment.1 Positron-emitting radionuclides such as 18F, 68Ga, and 64Cu
are often utilized for radiolabeling of biomolecules such as peptides,
oligonucleotides, and antibodies. Recently, 68Ga and 64Cu have
gained importance because of their ability to form chelate com-
plexes with 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic
acid (DOTA) which is conjugated to target biomolecules.2a On the
other hand, many practical 18F-labeling methods that involve the
use of sophisticated 18F-labeling agents, the so-called 18F-prosthetic
However, in the research field of 18F-labeling, the excessive use
of a precursor biomolecule (ca. 30–5000 nmol)3 for enhancing the
labeling reaction rate remains an outstanding problem, compared
with 68Ga or 64Cu-chelation labeling accomplished by the use of
several nmol of a biomolecule, because the available quantities of
biomolecules are often very tiny due to the expensive cost and
time-consuming process for the synthesis. This problem has lim-
ited the usefulness and expansion of biomolecular 18F-labeling.
Hence, we focused on the realization of more efficient 18F-labeling
with the use of the lowest possible amount, that is, ca. 20 nmol, of
our original oligodeoxynucleotide (ODN). This figure is related to
the level of ca. 25 nmol of an 18F-prosthetic group, which can be
quantitatively converted from an [18F]fluoride ion whose radioac-
Abbreviations: PET, positron emission tomography; ODN, oligodeoxynucleotide;
SPE, solid phase extraction; TBTA, tris((1-benzyl-1H-1,2,3-triazol-4-yl)methyl)-
amine; DCY, decay-corrected radiochemical yield; DOTA, 1,4,7,10-tetraazacyclod-
odecane-1,4,7,10-tetraacetic acid; TEAA, triethylammonium acetate.
tivity and specific radioactivity are 50 GBq and 2000 GBq/lmol,
respectively.4 In this paper, we describe a novel and effective
method for 18F-labeling of our original alkyne-substituted ODNs5
with the aim of achieving stoichiometric PET chemistry.
⇑
Corresponding authors. Tel.: +81 6 6879 8200; fax: +81 6 6879 8204 (S.O.).
E-mail addresses:
obika@phs.osaka-u.ac.jp
(S. Obika), yywata@riken.jp
(Y. Watanabe), suzuki.masaaki@riken.jp (M. Suzuki).
0968-0896/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2010.11.033

250
T. Kuboyama et al. / Bioorg. Med. Chem. 19 (2011) 249–255
conventional HPLC UV detectors, (3) low volatility and small molec-
2. Results and discussion
ular size as a labeling unit, and (4) design in consideration of the use
of our Pd(0)-mediated rapid C-[18F]fluoromethylation.11 By remote
control of the entire radiolabeling process, tetrabutylammonium
A Cu-catalyzed Huisgen reaction, so-called ‘click chemistry’,6
which connects two chemically stable moieties of azide and alkyne
with high chemoselectivity in mild biocompatible conditions, is
frequently used in bioconjugate chemistry. Actually, many groups
have applied this reaction to the 18F-labeling of biomolecules such
as ODNs,7 oligopeptides, azidothymidine, and sugar analogs, etc.8
According to the previous reports, more than 300 nmol of precur-
sor biomolecules have generally been used. In order to decrease
the substrate amount toward stoichiometric PET chemistry, two
main problems must be solved. One is that of obtaining a highly
pure 18F-prosthetic group because the purity is a crucial factor
for achieving ‘PET stoichiometric reaction’. Solid phase extraction
(SPE) and distillation are usually preferred as the methods of puri-
fication in a PET radiolabeling system because they are fast and
convenient. However, we could not completely remove large ex-
cesses of unreacted precursor, remaining reagents, and byproducts
using these methods. Consequently, we opted to use HPLC purifica-
tion to overcome this problem. The second is that a Cu-catalyzed
Huisgen reaction must be further accelerated because previous
conditions cannot satisfy our demands for two substrates with
amounts as low as ꢀ20 nmol being coupled in a short time. These
problems prompted us to explore more effective conditions that
would be applicable for PET tracer synthesis.
[
18F]fluoride, which was converted from 50 GBq of [18F]fluoride
ion with tetrabutylammonium hydrogen carbonate,12 reacted with
large excesses of tosylate 1 to produce [18F]3 in 97% HPLC analytical
yield (Scheme 1). Sequential HPLC isolation of [18F]3 from the reac-
tion mixture resulted in [18F]3 with ꢀ95% chemical purity (HPLC
analysis by UV 254 nm) (Fig. 1). The operation time of the reaction
and purification process was approximately 40 min. In the repeated
test syntheses using a relatively small radioactivity of 3.0 GBq of
[
[
18F]fluoride ion, HPLC analytical yield and the DCY of isolated
18F]3 based on [18F]fluoride ion were 97% and 30 13% (n = 3),
respectively.
2.3. Optimization of Cu-catalyzed Huisgen reaction toward
stoichiometric PET chemistry
We then concentrated on the acceleration of the Cu-catalyzed
Huisgen reaction in order to match time-limited 18F-labeling. In
our earlier study,5 we found that the combination of CuSO4, TBTA
[tris((1-benzyl-1H-1,2,3-triazol-4-yl)methyl)amine],13 and sodium
ascorbate as a catalyst system effectively reacted alkyne-substi-
tuted ODNs with various azides. In light of the development of this
research, we pursued further optimization of this catalyst system
2.1. For discussion of the synthetic efficiency of 18F-labeling
with unlabeled
3 and N-propargylbenzamide 4 as a model
Before a study, we here briefly mention the synthetic efficiency
of our 18F-labeling method by two parameters. One is HPLC analyt-
ical yield showing the reaction efficiency of 18F-labeling, which is
calculated by the peak area ratio of the desired [18F]fluorinated
product by radio-HPLC analysis of the reaction mixture. Another
decay-corrected radiochemical yield (DCY) shows the production
efficiency of the 18F-labeled compound based on decay-corrected
radioactivity of isolated [18F]fluorinated product. DCY is usually
lower than HPLC analytical yield mainly because DCY involves
the radioactivity loss owing to the absorption of 18F in glass ves-
sels9 and mechanical losses10 in PET tracer synthesizer.
2.2. Design and synthesis of 18F-prosthetic group [18F]3
Initially, we designed 1-azido-4-([18F]fluoromethyl)benzene
([18F]3) as a novel 18F-prosthetic group. [18F]3 has several advanta-
ges: (1) one-step conversion from tetrabutylammonium [18F]fluo-
ride, (2) good UV absorbance providing low detection limit using
18
[
F]TBAF
X
OTs
TBAHCO
3
Figure 1. Top: HPLC chromatogram of the reaction mixture in [18F] fluorination
with 1 by UV absorbance (254 nm). Middle and bottom: HPLC chromatograms of
N
N
3
3
5 min
purified [18F]3 by UV absorbance (254 nm) and
c-ray detector. Retention time of
3: para, X=F
F]3: para, X=
6: meta, X= F
1: para
2: meta
18
18
18
18F]3: 5.5 min.
[
[
F
[
18
F]6: meta, X=
F
O
HO
X
O
3
N
H
4
N
N
N
O CCGGTTGCT-
CTGAGACAT-3'
Cu-catalyzed
Huisgen reaction
HO
O
7 (20 nmol)
O
F
O CCGGTTGCT-
CuSO , TBTA,
4
CTGAGACAT-3'
N
H
N
N N
sodium ascorbate
15 min
8: para, X=F
18
18
18
[
F]8: para, X=
9: meta, X= F
F
5
18
[
F]9: meta, X= F
Scheme 2. Cu-catalyzed Huisgen reaction using model compounds azide 3 and
alkyne 4.
Scheme 1. Synthesis of 18F-labeled ODNs, [18F]8 and [18F]9.

T. Kuboyama et al. / Bioorg. Med. Chem. 19 (2011) 249–255
251
Table 1
Solvent effects observed in Cu-catalyzed Huisgen reactiona
Entry
Organic solvent
Yield of 5b (%)
1
2
3
4
5
6c
CH3CN
DMF
Dioxane
t-BuOH
DMSO
DMSO
1
82
94
95
97
8
a
Reaction of 3 (100 nmol) and 4 (400 nmol) was conducted in the presence of
CuSO4 (100 nmol), TBTA (100 nmol), and sodium ascorbate (200 nmol) in 100
a 70:30 mixture of organic solvent and H2O for 5 min at rt.
lL of
b
Yield was determined by HPLC analysis.
c
Reaction was conducted in the absence of TBTA.
Figure 2. Preparative HPLC profile of the reaction mixture in Huisgen type
18F-labeling with [18F]3 and 7.14 Top: UV chromatogram analyzed at 260 nm.
Bottom: radioactivity chromatogram obtained with a
c
A
-ray detector. Retention
time: 7, 8.9 min;
[
18F]8, 14.8 min; 18F]3, 29.0 min.
[
portion of
[
18F]8 was
substrate so that it can be used for PET reactions under highly di-
luted conditions (Scheme 2).
contained in the first peak of DMSO with the retention time of 3.5 min without
affinity for HPLC column.
We first investigated the solvent effect by reacting millimolar
concentrations of 3 and 4 in a mixture of an organic solvent and
H2O (70:30) for a fixed reaction time of 5 min at room temperature.
As shown in Table 1, CH3CN was not effective as an organic solvent
(entry 1), whereas the desired product 5 was obtained in excellent
yields when using DMSO, dioxane, and t-BuOH (entries 2–5). How-
ever, dioxane was known to be carcinogenic, and the solubility of
TBTA in t-BuOH was relatively low. Therefore, we chose DMSO as
the best organic solvent for our subsequent reactions. Table 2 sum-
marizes the effect of the ratio of the additives, reaction tempera-
ture, and the mixed solvent of DMSO-buffer on the yield when
reactions. It was also found in our experiments that the product
yield decreased slightly when the pH of sodium phosphate buffer
was 6.5 or 7.5. As optimized in Table 2, when using the pH 7.0 buf-
fer, the reaction efficiency was well satisfying, and the ODNs in the
solution were stabilized.
2.4. 18F-labeling of ODN under optimized conditions
Thus, we concluded that such optimized conditions (entry 9 in
Table 2) would have high potential for conducting the stoichiome-
try-focused Huisgen reaction for the 18F-labeling of our target ODN
(Scheme 1). In the actual PET radiolabeling, 20 nmol of ethynyl
low concentrations of 3 (50 lM) and 4 (100 lM) were used. The di-
lute reaction conditions mentioned here are similar to those
adopted in PET radiolabeling experiments. The yield obtained
when using a 2:1 mixture of CuSO4 and TBTA in a mixture of DMSO
and pH 7.0 sodium phosphate buffer (20:80) was better than that
obtained when using a 1:1 mixture of CuSO4 and TBTA in the same
solvent system (entries 1 and 2). Decreasing the amount of Cu cat-
alyst resulted in lower yields (entries 3–5). Interestingly, with a
slight increase of the temperature to 40 °C, the reaction smoothly
proceeded, giving the desired 5 in excellent yields of 87–94% (en-
tries 6 and 7). After extensive experiments, we found that the reac-
tion was carried out effectively when a 2:1:2 mixture of CuSO4,
TBTA, and sodium ascorbate was used in a 30:70 mixture of DMSO
and buffer (entries 8 and 9). In particular, when the reaction
temperature was 40 °C, the yield of 5 was the highest, that is,
96% (entry 9). Even with a 10-fold decrease in the amounts of 3
ODN 714 (0.50 mM aqueous solution, 40
l
l
L) was mixed with a
L). Then, a mixture of
solution of [18F]3 in DMSO–H2O (180–290
TBTA (50 mM in DMSO, 6
12 L, 600 nmol), sodium ascorbate (50 mM, 12
and sodium phosphate buffer (pH 7.0, 100 mM, 60
The mixture with a total volume of 600 L was heated for 15 min
l
L, 300 nmol), CuSO4 (50 mM in H2O,
L, 600 nmol),
L) was added.
l
l
l
l
at 40 °C. The reaction mixture was directly analyzed by radio-HPLC
and the analytical yield of [18F]8 was calculated to be up to 92%
(typically ꢀ60%).15 Subsequent HPLC purification of the reaction
mixture (Fig. 2) gave the desired 18F-labeled ODN [18F]8 with sat-
isfactory radioactivity (ca. 2.5 GBq) and fairly high specific radioac-
tivity (2400 GBq/lmol). The entire synthesizing time, including the
preparation of [18F]3 with [18F]fluoride ion, was 84 min and the
DCY was up to 8.6%,12,15 calculated on the basis of 50 GBq [18F]fluo-
ride ion in a single cyclotron bombardment. The DCY of the desired
and 4 (the concentrations of 3 and 4: 5 lM and 10 lM, respec-
[
18F]8 based on the 18F-prosthetic group [18F]3 was ca. 30%. The
tively), the reaction proceeded effectively to afford 5, although in
a relatively low yield (61%; entry 10). It was encouraging that
the substrates underwent click cyclization even when their con-
centration was lower than that in the conventional 18F-labeling
chemical and radiochemical purities of [18F]8 were 95% and 87%,
respectively. [18F]8 was identified by HPLC analysis using the unla-
beled authentic compound 8 as the reference.16
Table 2
Optimizations of Cu-catalyzed Huisgen reactiona
Entry
CuSO4, M (
l
M)
TBTA, M (
l
M)
Sodium ascorbate, M (
lM)
T
Solvent DMSO:buffer
Yield of 5b (%)
1
2
3
4
5
6
7
8
1000
1000
500
200
100
1000
500
1000
1000
1000
1000
500
250
100
50
500
250
500
500
500
1000
1000
500
200
100
1000
500
1000
1000
1000
rt
rt
rt
rt
20:80
20:80
20:80
20:80
20:80
20:80
20:80
30:70
30:70
30:70
71
79
81
67
41
94
87
92
96
61
rt
40 °C
40 °C
rt
40 °C
40 °C
9
10c
a
b
c
Reaction of 3 (5 nmol) and 4 (10 nmol) was conducted in 100
Yield was determined by HPLC analysis.
Reaction of 3 (0.5 nmol) and 4 (1 nmol) was conducted in 100
l
L of a mixture of DMSO and sodium phosphate buffer (10 mM, pH 7.0) for 15 min.
lL of a mixture of DMSO and buffer.

252
T. Kuboyama et al. / Bioorg. Med. Chem. 19 (2011) 249–255
2.5. Synthesis of 18F-labeled meta-isomer
by use of the integrated microfluid devices for multistep 18F-radio-
synthesis,19c which are making remarkable progress in PET chem-
istry. In addition, it is worth mentioning that since a compact
aryltriazole-substituted deoxyribose unit is structurally similar to
a nucleoside, it can be easily incorporated into an ODN as a dummy
nucleoside. Thus, with the proposed method, site-specific 18F-
labeling can be carried out at any position in the ODN without
causing any drastic change in the intrinsic biological activities of
the resulting oligonucleotides.16 The applications of the proposed
labeling method and further biological studies will be reported in
due course.20
By following the procedure for synthesizing [18F]8, we suc-
ceeded in synthesizing [18F]9, a meta structural isomer of [18F]8
in the labeling position,17 under the same reaction conditions
using 7 and a meta prosthetic group 1-azido-3-([18F]fluorometh-
yl)benzene ([18F]6) which was synthesized from tosylate 2.9 The
radioactivity and specific radioactivity of this compound were
2.1 GBq and ca. 1800 GBq/lmol, respectively. The chemical and
radiochemical purities of [18F]9 were 99% and 93%, respectively.
The DCY was approximately 7.2%, calculated on the basis of
50 GBq of [18F]fluoride ion in a single cyclotron bombardment
(see Section 4).
4. Experimental section
4.1. General
2.6. PET study of 18F-labeled ODNs using rats
PET imaging of [18F]8 was conducted on male rats under isoflu-
rane anesthesia (n = 4). After an intravenous injection of [18F]8, the
radioactivity was first detected in the inferior vena cava for a few
seconds. Subsequently, the radioactivity was detected in the kid-
neys, liver, urinary bladder, and intestines. Afterwards, the radioac-
tivity accumulated in the bones, such as the vertebrae, costae,
femur, etc. (Fig. 3). It is likely that [18F]8 was gradually metabolized
into [18F]fluoride ion, which accumulated in the bones.18 Mean-
while, the in vivo behavior of the meta-labeled ODN [18F]9 was
substantially different from that of [18F]8. These PET studies are
currently still in progress.
Melting points were measured with a Yanagimoto micro melt-
ing point apparatus. IR spectra were recorded on a Shimadzu IR
Prestige-21 Fourier Transform Infrared Spectrophotometer. 1H
and 13C NMR spectra were recorded on a JEOL JNM-AL400. EI and
FAB mass spectra were recorded on a JEOL JMS-D300 or JMS-600
mass spectrometer. MALDI-TOF mass spectra were recorded on
an Applied Biosystems VoyagerÒ-DE. HPLC analysis was performed
on two JASCO PU-980 Intelligent HPLC pumps, an MD-910 multi-
wavelength detector, and a CU-965 column oven. Radio-HPLC
analysis was performed on a Shimadzu Prominence CBM-20A com-
munication bus module, an LC-20AB liquid chromatograph, and a
CTO-20AC column oven; an SPD-20A UV/VIS detector and an
RLC-700 Aloka radio analyzer. The amount of a radioacitve com-
pound was determined by UV254 or UV260 calibration curve pro-
vided by radioinactive authentic samples. The specific activity of
a radioactive compound was calculated as the ratio of the dose
measured on a Biodex Mecical Systems ATOMLAB 300 by the
amount of the compound. ODNs were synthesized utilizing an Ap-
plied Biosystems 394 DNA/RNA synthesizer. An anion exchange re-
sin cartridge SAIKA-SPE SAX-30 (AiSTI SCIENCE), Sep-Pak Plus C18
(Waters), and other chemicals (Sigma–Aldrich or Nacalai Tesque)
were used for 18F-labeling reactions and for synthesizing the
ODN precursors, model compounds, and authentic compounds.
18O-enriched water (Taiyo-Nissan) was irradiated by 12 MeV pro-
tons in a Sumitomo CYPRIS HM-12S cyclotron (Sumitomo Heavy
Industries) to generate [18F]fluoride ion. Literature procedures
were used for the syntheses of 4-azidobenzyl alcohol21a and 3-azi-
dobenzyl alcohol.21b 18F-labeling reaction was conducted using an
automated radiolabeling system at RIKEN CMIS.
3. Conclusion
In summary, we succeeded in developing a method for the site-
specific 18F-labeling of ODNs using novel 18F-prosthetic groups,
azido([18F]fluoromethyl)benzenes ([18F]3 and [18F]6), via a Cu-cat-
alyzed Huisgen reaction. The labeling was conducted using a mere
20 nmol of alkyne-substituted ODN, an amount similar to that of
18F-prosthetic groups, under mild conditions at 40 °C for 15 min
to provide the desired 18F-labeled ODNs with an adequate radioac-
tivity of ꢀ2.5 GBq applicable for animal and human PET studies.
This 18F-labeling procedure provides the firm chemical basis for
stoichiometry-oriented PET labeling reactions, potentially useful
for the PET tracer syntheses of middle molecular weight biomole-
cules, such as ODNs, and would be especially valuable as requisites
for the promotion of recently attracted 18F-labelings with micro-
scale reactors.19 That is, the advantage of our optimized conditions
for stoichiometry-focused 18F-labeling will be preferably insured
Figure 3. Maximum-intensity-projection PET image of [18F]8 in the rat abdomen.

T. Kuboyama et al. / Bioorg. Med. Chem. 19 (2011) 249–255
253
4.2. Product analyses
0 min to 20 min, flow rate: 4.0 mL minꢂ1
,
UV wavelength:
260 nm, column temperature: 50 °C, retention time of [18F]8:
14–15 min. The fraction containing the desired 18F-labeled com-
pound was collected, and CH3CN was evaporated under reduced
pressure to afford a solution of [18F]8 in TEAA buffer (ꢀ5 mL).
HPLC analysis of [18F]3 and [18F]6; column: COSMOSIL 5C18-MS-
II 4.6 ꢁ 150 mm (Nacalai Tesque, Inc.), eluent: 50:50 (v/v) mixture
of CH3CN and phosphate buffer (10 mM NaH2PO4 + 10 mM
Na2HPO4), flow rate: 1.5 mL minꢂ1, UV wavelength: 240 and
254 nm, column temperature: 35 °C, retention time of [18F]3:
5.5 min (see Fig. 1), retention time of [18F]6: 5.8 min.
[
18F]8 was identified by HPLC analysis using the unlabeled
authentic compound 8 as the reference. The total time taken for
the synthesis, from the generation of [18F]fluoride ion in the cyclo-
tron to the preparation for the TEAA buffer solution of [18F]8, was
84 min. The isolated radioactivity and specific radioactivity at
HPLC analysis of [18F]8 and [18F]9; column: COSMOSIL 5C18-MS-
II 4.6 ꢁ 100 mm (Nacalai Tesque, Inc.), linear gradient elution:
10–24% CH3CN in 0.1 M TEAA buffer (tetraethylammonium
acetate) from 0 min to 20 min, and 24–80% CH3CN in 0.1 M TEAA
buffer from 20 min to 30 min, flow rate: 1.0 mL minꢂ1, UV wave-
length: 260 nm, column temperature: 50 °C, retention time of
the end of synthesis: 2.53 GBq and 2366 GBq/lmol, respectively.
The chemical purity determined at 260 nm and radiochemical
purity: 95% and 87%, respectively. The isolated yield of the
product was calculated to be 5.1% on the basis of 50 GBq of the
[
18F]8: 12.1 min, retention time of [18F]9: 12.1 min.
[
18F]fluoride ion in a single cyclotron bombardment, and the DCY
HPLC analysis of 5 in the model reaction; column: COSMOSIL
was 8.6%.
5C18-MS-II 4.6 ꢁ 150 mm (Nacalai Tesque, Inc.), eluent: 40:60 (v/v)
mixture of CH3CN and phosphate buffer (10 mM NaH2PO4 + 10 mM
Na2HPO4), flow rate: 1.5 mL minꢂ1, UV wavelength: 240 nm, column
temperature: 35 °C, retention time of 5: 4.0 min. Yield of 5 was deter-
mined by using n-butyl 4-hydorxybenzoate as an internal standard.
4.3.2. Syntheses of the prosthetic group [18F]6 and the ODN PET
tracer [18F]9
[
18F]6 and [18F]9 were synthesized from 2 by following the
same procedure described above. The HPLC analytical yields of
[
18F]6 and [18F]9 were 96% and 90%, respectively. [18F]9 was iden-
4.3. 18F-Labeling of ODNs and preparation of the corresponding
precursors and unlabeled authentic compounds
tified by HPLC analysis using the unlabeled authentic compound 9
as the reference. The total time taken for the synthesis, starting
from the generation of
83 min. The isolated radioactivity and specific radioactivity of the
product at the end of synthesis: 2.12 GBq and 1809 GBq/ mol,
[
18F]fluoride ions by the cyclotron,
4.3.1. Syntheses of prosthetic group [18F]3 and ODN PET tracer
[
18F]8
[
18F]fluoride ion with a radioactivity of approximately 50 GBq in
l
2 mL of [18O]water was transferred to an automated radiolabeling
system, and [18F]fluoride ion was trapped by a short anion-ex-
change resin cartridge SAIKA-SPE SAX-30 which had been precon-
ditioned with aqueous sodium hydrogen carbonate (1.0 M, 5 mL)
and water (10 mL). The trapped [18F]fluoride ion was eluted into
the first reaction vessel with tetrabutylammonium hydrogen car-
bonate (0.6 mL of 0.025 M solution in 80% CH3CN–water mixture)
and rinsed with CH3CN (0.6 mL). CH3CN and water were evapo-
rated under reduced pressure at 110 °C under a He gas stream,
and the residue was azeotropically dried using an additional
1 mL of CH3CN.
respectively. Chemical purity (based on UV260) and radiochemical
purity: 99% and 93%, respectively. The isolated yield was calculated
to be 4.2% on the basis of 50 GBq of the [18F]fluoride ion in single
cyclotron bombardment, and the DCY was 7.2%.
4.3.3. Desalting of the solution [18F]8 or [18F]9 in TEAA buffer
If necessary, the solution of [18F]8 or [18F]9 in the TEAA buffer
can be desalted by passing it through a Sep-Pak Plus C18 (precon-
ditioned with 40 mL of EtOH and 40 mL of H2O), washed twice
with 5 mL H2O, dried under N2 flow for 1 min, and eluted using
1 mL EtOH. Then, EtOH was evaporated under a N2 gas stream
and the residue was dissolved in saline to be used as the injection
solution in the animal PET studies. The total process time was
30 min and 90% (DCY) of the radioactivity was recovered.
A solution of 1 (6.0 mg) in CH3CN (1 mL) was added to the res-
idue. The resulting mixture was heated at 85 °C for 5 min. A sam-
pling solution (ꢀ5
lL) withdrawn from the reaction mixture was
directly analyzed by HPLC (HPLC analytical yield of [18F]3: 97%).
Then, the reaction mixture was diluted with 0.3 mL of H2O and in-
jected into preparative HPLC. The purification conditions are as fol-
lows: COSMOSIL 5C18-MS-II 10 ꢁ 250 mm (Nacalai Tesque, Inc.),
eluent: 40:60 (v/v) CH3CN/H2O mixture from 0 min to 6 min, and
then 50:50 CH3CN/H2O mixture from 6 min to 14 min, flow rate:
4.0 mL minꢂ1, UV wavelength: 254 nm, column temperature: rt,
retention time of [18F]3: 15–16 min.
4.3.4. Synthesis of 4-azidobenzyl 4-methylbenzenesulfonate (1)
Pyridine (0.162 mL, 2.00 mmol) and p-toluenesulfonic anhy-
dride (343 mg, 1.05 mmol) were added to a stirred solution of
4-azidobenzyl alcohol (149 mg, 1.00 mmol) in dichloromethane
(5.0 mL) at 0 °C. After 30 min, the resulting mixture was parti-
tioned between H2O and dichloromethane. The organic phase
was washed with 1.0 M hydrochloric acid, saturated aqueous so-
dium bicarbonate, and brine, then, it was dried over anhydrous
magnesium sulfate, filtered, and concentrated under reduced pres-
sure. The residue was suspended with diethyl ether (3 mL) and fil-
tered through a cotton plug. The filtrate was diluted with hexane
(5 mL) and cooled in an ice bath. After 30 min, the supernatant
fluid was pipetted off and the crystals were washed twice with
hexane and dried under reduced pressure to afford 1 (218 mg,
0.719 mmol, 72% yield; colorless crystals). Since 1 was found to
decompose rapidly at room temperature, it was stored in a freezer
and kept away from moisture.
The fraction containing [18F]3 was collected and transferred to
the second reaction vessel that was charged with 0.18 mL of DMSO.
CH3CN was gently evaporated under reduced pressure at 40 °C
with a He gas stream. To the resulting DMSO–H2O (0.18 mL—ca.
0.29 mL) solution of [18F]3, sodium phosphate buffer (60
l
L,
100 mM, pH 7.0), the ethynyl precursor of our ODN 7 (0.50 mM
in H2O, 40 L, 20 nmol), CuSO4 (50 mM in H2O, 12 L, 600 nmol),
TBTA (50 mM in DMSO, 6 L, 300 nmol), and sodium ascorbate
(50 mM in H2O, 12 L, 600 nmol) were added. The mixture was
heated at 40 °C for 15 min. A sampling solution (ꢀ5 L) withdrawn
l
l
l
l
l
from the reaction mixture was diluted with 0.1 M TEAA buffer and
analyzed by HPLC (HPLC analytical yield of [18F]8: 92%). Then, the
reaction mixture was diluted with 0.3 mL of TEAA buffer and in-
jected into HPLC. The purification conditions are as follows: COS-
MOSIL 5C18-AR-II 10 ꢁ 250 mm (Nacalai Tesque, Inc.), linear
gradient elution: 10–20% of CH3CN in 0.1 M TEAA buffer from
1H NMR (400 MHz, CDCl3) d: 2.45 (3H, s), 5.02 (2H, s), 6.97 (2H,
d, J = 8.3 Hz), 7.24 (2H, d, J = 8.3 Hz), 7.34 (2H, d, J = 8.3 Hz), 7.79
(2H, d, J = 8.3 Hz). 13C NMR (100 MHz, CDCl3) d: 21.7, 71.3, 119.2,
127.9, 129.9, 129.9, 130.3, 133.3, 141.0, 144.9. HRMS (EI) calcd
for C14H13N3O3S (M+) 303.0677, found 303.0650.

254
T. Kuboyama et al. / Bioorg. Med. Chem. 19 (2011) 249–255
4.3.5. Synthesis of 3-azidobenzyl 4-methylbenzenesulfonate (2)
We synthesized 2 from 3-azidobenzyl alcohol by following the
same procedure reported for 1 (yield of 2: 52%). Compound 2
decomposes gradually at room temperature and is stored in a free-
zer and kept away from moisture.
4 °C before Tm measurements. The samples were heated at a rate
of 0.5 °C/min. The melting temperatures (Tm calves) were deter-
mined by plotting the first derivative of the absorbance (260 nm)
versus the temperature curve (n = 3). Tm values of 50-CCGGTTGCT
CTGAGACAT-30, 8, and 9 were 65 °C, 66 °C, and 66 °C, respectively.
IR (film, KBr): 2114, 1593, 1489, 1452, 1360, 1292, 1177, 945,
835, 814, 781, 665 cmꢂ1
.
1H NMR (400 MHz, CDCl3) d: 2.45 (3H,
4.4.1. Synthesis of N-((1-(4-(fluoromethyl)phenyl)-1H-1,2,3-
triazol-4-yl)methyl)benzamide (5) for model Cu-catalyzed
Huisgen reaction
s), 5.03 (2H, s), 6.85 (1H, s), 6.97 (1H, d, J = 7.8 Hz), 7.03 (1H, d,
J = 7.8 Hz), 7.30 (1H, t, J = 7.8 Hz), 7.33 (2H, d, J = 8.5 Hz), 7.79
(2H, d, J = 8.5 Hz). 13C NMR (100 MHz, CDCl3) d: 21.6, 71.0, 118.8,
119.5, 124.7, 128.0, 129.9, 130.1, 133.1, 135.3, 140.5, 145.0. MS
(EI): m/z 303 (M+). Anal. Calcd for C14H13N3O3S: H, 4.32; C, 55.43;
N, 13.85. Found: H, 4.32; C, 55.59; N, 13.89.
To a stirred solution of 2.5 mg (0.010 mmol) of copper (II) sul-
fate pentahydrate and 7.9 mg (0.040 mmol) of sodium ascorbate in
0.25 mL of H2O were added a DMF solution (1.0 mL) of N-propa-
rgylbenzamide 4 (31.8 mg, 0.200 mmol), and a DMF solution
(0.5 mL) of 3 (30.2 mg, 0.200 mmol). The mixture was heated to
80 °C and stirred for 15 min. Then the mixture was cooled to room
temperature and partitioned between 3% aqueous ammonia and
AcOEt. The organic layer was washed with H2O and brine, dried
over anhydrous sodium sulfate, filtered, and concentrated under
reduced pressure. The residue was crystallized with AcOEt, fil-
tered, washed with cooled AcOEt, and then dried under reduced
pressure to give 5 (49.4 mg, 0.159 mmol, yield: 80%) as a white
powder.
4.3.6. Synthesis of 1-azido-4-(fluoromethyl)benzene (3)
A solution of 1 (348 mg, 1.15 mmol) in CH3CN (5 mL) was added
to a stirred suspension of potassium fluoride (267 mg, 4.59 mmol)
and 18-crown-6 (303 mg, 1.15 mmol) in CH3CN (5 mL). The mix-
ture was heated to 50 °C and stirred for 30 h. Then the mixture
was cooled to room temperature and partitioned between H2O
and AcOEt. The aqueous layer was extracted once with AcOEt.
The combined organic extracts were dried over anhydrous sodium
sulfate, filtered, and concentrated under reduced pressure. The res-
idue was purified by silica gel column chromatography (eluent: 5%
AcOEt in hexane) to give 3 (123 mg, 0.814 mmol, yield: 71%) as a
colorless oil.
Mp: 190–192 °C. IR (film on KBr): 3319, 1634, 1553, 698 cmꢂ1. 1H
NMR (400 MHz, DMSO-d6) d: 4.61 (2H, d, J = 5.6 Hz), 5.49 (2H, d,
J = 47.6 Hz), 7.44–7.55 (3H, m), 7.62 (2H, dd, J = 8.2, 1.6 Hz),
7.88–7.91 (2H, m), 7.96 (2H, d, J = 8.2 Hz), 8.72 (1H, s), 9.08 (1H, t,
J = 5.6 Hz). 13C NMR (100 MHz, DMSO-d6) d: 34.8, 83.5 (d,
J = 166.4 Hz), 120.0, 121.2, 127.3, 128.3, 129.2 (d, J = 5.8 Hz), 131.3,
134.1, 136.4 (d, J = 16.6 Hz), 136.7 (d, J = 3.3 Hz), 146.4, 166.2. HRMS
(FAB) calcd for C17H16FN4O (M+H)+ 311.1308, found 311.1313. Anal.
Calcd for C17H15FN4O: H, 4.87; C, 65.80; N, 18.05. Found: H, 4.87; C,
65.70; N, 18.02.
IR (film on KBr): 2920, 2112, 1284, 1016, 912, 742 cmꢂ1 1H
.
NMR (400 MHz, CDCl3) d: 5.34 (2H, d, J = 47.8 Hz), 7.05 (2H, d,
J = 8.4 Hz), 7.38 (2H, dd, J = 8.4, 2.0 Hz). 13C NMR (100 MHz, CDCl3)
d: 84.0 (d, J = 165.5 Hz), 119.2 (d, J = 1.7 Hz), 129.4 (d, J = 5.8 Hz),
132.8 (d, J = 17.3 Hz), 140.7 (d, J = 3.3 Hz). HRMS (EI) calcd for
C7H6FN3 (M+H)+ 151.0546, found 151.0555.
4.3.7. Synthesis of 1-azido-3-(fluoromethyl)benzene (6)
We synthesized 6 from 2 by the following same procedure re-
ported for the synthesis of 3 (yield of 6: 66%).
4.5. PET studies using rats
Male Sprague–Dawley rats (SLC, Hamamatsu, Shizuoka, Japan)
weighing approximately 250 g were used. The rats (n = 4) were
anesthetized and maintained with a mixture of 1.5% isoflurane
and nitrous oxide/oxygen (7:3) and positioned in the gantry of a
PET scanner (microPET Focus 220, Siemens Co., Ltd, Knoxville,
TN, USA). After positioning and fixation, a 30-min transmission
scan with a rotating 68Ge–68Ga pin source was performed. Then a
120-min emission scan of the abdomen was performed with
400–650 keV as the energy window and 6 ns as the coincidence
time window. Emission data were acquired in the list mode, after
an intravenous bolus injection of [18F]8 (approximately 50 MBq
per animal) via a tail vein. The acquired data were sorted into dy-
namic sinograms (6 ꢁ 10 s, 6 ꢁ 30 s, 11 ꢁ 60 s, 15 ꢁ 180 s, 6 ꢁ 600
s; a total of 44 frames). The data were reconstructed by a statistical
maximum a posteriori probability algorithm (MAP) of 12 iterations
with point spread function (PSF) effect. During the experiment,
body temperature was maintained at 37 °C with a heating blanket.
The radioactivity concentrations were normalized with cylinder
phantom data, and were expressed as standardized uptake values
(SUV). All experimental protocols were approved by the RIKEN’s
Ethics Committee on Animal Care and Use and were performed
in accordance with the Principles of Laboratory Animal Care (NIH
publication No. 85–23, revised 1985).
IR (film on KBr): 2114, 1589, 1485, 1450, 1295, 990, 786 cmꢂ1
.
1H NMR (400 MHz, CDCl3) d: 5.37 (2H, d, J = 47.6 Hz), 7.02 (1H, d,
J = 7.8 Hz), 7.04 (1H, s), 7.13 (1H, d, J = 7.8 Hz), 7.37 (1H, d,
J = 7.8 Hz). 13C NMR (100 MHz, CDCl3) d: 83.8 (d, J = 168.0 Hz),
117.6 (d, J = 6.6 Hz), 119.2 (d, J = 3.3 Hz), 123.5 (d, J = 6.6 Hz),
130.0, 138.2 (d, J = 17.4 Hz), 140.5. MS (EI): m/z 151 (M+H)+. Anal.
Calcd for C7H6FN3: H, 4.00; C, 55.63; N, 27.80. Found: H, 4.19; C,
55.74; N, 27.95.
4.3.8. Synthesis of ODN 7, a precursor for 18F-labeling
ODN 7 was synthesized according to our previously reported
procedure (see Refs. 5,14).
MALDI-TOF MS m/z 7 [M+H]+ calcd 5693.52, found 5692.07.
4.3.9. Synthesis of unlabeled ODNs 8 and 9
We synthesized 8 from 3 and 7 according to our previously re-
ported procedure (see Refs. 5,14). MALDI-TOF MS m/z 8 [M+H]+
calcd 5844.66, found 5844.39.
We also prepared 9 from 6 and 7. MALDI-TOF MS m/z 9 [M+H]+
calcd 5844.66, found 5844.39.
4.4. DNA melting experiments
The thermal stabilities of the duplexes were determined using a
spectrophotometer equipped with a thermoregulator (SHIMADZU
Acknowledgments
UV-1650PC). Hybridization mixtures of 130
using a medium salt buffer (10 mM sodium phosphate buffer [pH
l
L were prepared
This work was supported in part by a consignment expense for
the Molecular Imaging Program ‘Research Base for Exploring New
Drugs’ from the Ministry of Education, Culture, Sports, Science
and Technology (MEXT) of Japan. The authors acknowledge the
technical support extended by Mr. Masahiro Kurahashi in the
7.0] and 100 mM NaCl) and an equimolar amount (2.0 lmol/L) of
ODNs and complementary RNA (50-AUGUCUCAGAGCAACCGG-30).
The mixtures were incubated at 85 °C, then cooled back slowly to

T. Kuboyama et al. / Bioorg. Med. Chem. 19 (2011) 249–255
255
the reaction mixture into preparative-HPLC, and incomplete isolation of the
18F-labeled product by preparative-HPLC.
cyclotron experiments. T.K. acknowledges the support extended by
Kyowa Hakko Kirin Co., Ltd.
11. We have also intended the syntheses of azido[18F]fluorobenzenes [18F]3 and
[
18F]6 by our Pd(0)-mediated rapid C-[18F]fluoromethylation using arylborons
and [18F]fluoromethyl halide, see, (a) Doi, H.; Ban, I.; Nonoyama, A.; Sumi, K.;
Kuang, C.; Hosoya, T.; Tsukada, H.; Suzuki, M. Chem. Eur. J. 2009, 15, 4165; (b)
Doi, H.; Goto, M.; Suzuki, M. ‘Pd0-mediated rapid C-[18F]fluoromethylation by
the reaction of a [18F]fluoromethyl halide and an organoboron compound’,
15TH European Symposium On Radiopharmacy And Radiopharmaceuticals,
Edinburgh, Scotland (UK), April 8–11, 2010, Abstract, Q. J. Nucl. Med. Mol.
Imaging, April 2010, 54, Suppl. 1 to No. 2, pp 18–19.
Supplementary data
Supplementary data (13C NMR spectra of 1 and 3) associated
with this article can be found, in the online version, at
doi:10.1016/j.bmc.2010.11.033.
12. Culbert, P. A.; Adam, M. J.; Hurtado, E. T.; Huser, J. M. A.; Jivan, S.; Lu, J.; Ruth, T.
J.; Zeisler, S. K. Appl. Radiat. Isot. 1995, 46, 887.
13. Chan, T. R.; Hilgraf, R.; Sharpless, K. B.; Fokin, V. V. Org. Lett. 2004, 6, 2853–
2855.
14. The sequence of the oligonucleotide used in this study is complementary to the
135–152 region containing AUG start codon of bcl-xL mRNA. Roongjang, S.;
Takahashi, K.; Obika, S.; Imanishi, T. Nucleic Acids Symp. Ser. 2007, 51,
113.
References and notes
1. (a) Phelps, M. E. PET: Molecular Imaging and Its Biological Applications; Springer:
New York, Berlin, Heidelberg, 2004; (b) Positron Emission Tomography; Bailey,
D. L., Townsend, D. W., Valk, P. E., Maisey, M. N., Eds.; Basic Science: Springer,
New York, Berlin, Heidelberg, 2005; (c) Fowler, J. S.; Wolf, A. P. Acc. Chem. Res.
1997, 30, 181; (d) Ametamey, S. M.; Honer, M.; Schubiger, P. A. Chem. Rev. 2008,
108, 1501.
15. In addition to the mechanical loss of radioactivity observed in the synthesis of
[
18F]3, we found that when the reaction mixture containing ODN [18F]8 was
2. For a recent review of metal nuclides conjugates, see: (a) Tanaka, K.; Fukase, K.
Org. Biomol. Chem. 2008, 6, 815–828; For non-metal nuclides, see: (b) Miller, P.
W.; Long, N. J.; Vlar, R.; Gee, A. D. Angew. Chem. 2008, 120, 9136; Angew. Chem.,
Int. Ed. 2008, 47, 8998–9033.
purified by semipreparative HPLC, as shown in Figure 2, a portion of [18F]8 was
eluted off with DMSO with the retention time of 3.5 min without affinity for
HPLC column.
16. Binding of the unlabeled prosthetic group 3 to the original sequence of ODNs
had no effect on Tm value (see Section 4).
3. Flavell et al. reported a coupling reaction between aldehyde [18F]FBA and a
nearly stoichiometric amount (30 nmol) of aminooxy peptide. They used the
17. We designed and synthesized para and meta compounds of [18F]8 and [18F]9
for the purpose of analyzing the stabilities of the labeling positions in in vivo
biological process. In addition, by comparison of the chemical properties, a
para tosylate 1 is so unstable that should be stored at ꢂ80 °C, but a meta
structural isomer 2 is more stable and can be stored in a normal freezer
(ꢂ20 °C).
‘expressed protein ligation’ technique to introduce
a chemically-unstable
aminooxy group into leptin. Flavell, R. R.; Kothari, P.; Bar-Dagan, M.; Synan, M.;
Vallabhajosula, S.; Friedman, J. M.; Muir, T. W.; Ceccarini, G. J. Am. Chem. Soc.
2008, 130, 9106.
4. In our laboratory, the specific radioactivity of [18F]fluoride was estimated to be
approximately 500–2500 GBq/
obtained in conventional single cyclotron bombardment would be
lmol. Therefore, [
18F]fluoride with 50 GBq
18. For an example of the rapid degradation of 18F-labeled DNA in baboons, see:
Tavitian, B.; Terrazzino, S.; Kühnast, B.; Marzabal, S.; Stettler, O.; Dollé, F.;
Deverre, J.-R.; Jobert, A.; Hinnen, F.; Bendriem, B.; Crouzel, C.; Giamberardino,
L. D. Nat. Med. 1998, 4, 467.
a
approximately 20–100 nmol.
5. Nakahara, M.; Kuboyama, T.; Izawa, A.; Hari, Y.; Imanishi, T.; Obika, S. Bioorg.
Med. Chem. Lett. 2009, 19, 3316.
6. (a) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. Angew. Chem., Int.
Ed. 2002, 41, 2596; Review: (b) Kolb, H. C.; Sharpless, K. B. Drug Discovery Today
2003, 8, 1128–1137; Review in oligonucleotide chemistry: (c) Amblard, F.; Cho,
J. H.; Schinazi, R. F. Chem. Rev. 2009, 109, 4207.
7. Inkster, J. A. H.; Adam, M. J.; Storr, T.; Ruth, T. J. Nucleosides Nucleotides Nucleic
Acids 2009, 28, 1131.
8. For review of click labeling in PET chemistry, see: Glaser, M.; Robins, E. G. J.
Labelled Compd. Radiopharm. 2009, 52, 407. and see references cited therein.
9. Brodack, J. W.; Kilbourn, M. R.; Welch, M. J.; Katzenellenbogen, J. A. Appl. Radiat.
Isot. 1986, 37, 217.
10. The mechanical losses of synthesized PET tracer were due to incomplete
transfer of the reaction mixture from vessel to vessel, insufficient injection of
19. (a) Lu, S. Y.; Pike, V. W. In PET Chemistry, The Driving Force in Molecular Imaging;
Schubiger, P. A., Lehmann, L., Friebe, M., Eds.; Springer: New York, Berlin,
Heidelberg, 2007; pp 271–287; (b) Cai, L.; Lu, S.; Pike, V. W. Eur. J. Org. Chem.
2008, 2853; (c) Elizarov, A. M.; van Dam, R. M.; Shin, Y. S.; Kolb, H. C.; Padgett,
H. C.; Stout, D.; Shu, J.; Huang, J.; Daridon, A.; Heath, J. R. J. Nucl. Med. 2010, 51,
282.
20. This 18F-labeling methodology have been demonstrated by the syntheses of 20
types of 18F-labeled ODNs focusing on nucleotide sequences and backbone
modifications, thus far.
21. (a) Andersen, J.; Madsen, U.; Björkling, F.; Liang, X. Synlett 2005, 2209; (b)
Georg, G. I.; Boge, T. C.; Park, H.; Himes, R. H. Bioorg. Med. Chem. Lett. 1995, 5,
615.