Copper-mediated aromatic radiofluorination revisited: Efficient production of PET tracers on a preparative scale

Article abstract of DOI:10.1002/chem.201405586

Two novel methods for copper-mediated aromatic nucleophilic radiofluorination were recently reported. Evaluation of these methods reveals that, although both are efficient in small-scale experiments, they are inoperative for the production of positron emission tomography (PET) tracers. Since high base content turned out to be responsible for low radiochemical conversions, a "low base" protocol has been developed which affords ¹⁸F-labeled arenes from diaryliodonium salts and aryl pinacol boronates in reasonable yields. Furthermore, implementation of our "minimalist" approach to the copper-mediated [¹⁸F]-fluorination of (mesityl)(aryl)iodonium salts allows the preparation of ¹⁸F-labeled arenes in excellent RCCs. The novel radiofluorination method circumvents time-consuming azeotropic drying and avoids the utilization of base and other additives, such as cryptands. Furthermore, this procedure enables the production of clinically relevant PET tracers; [¹⁸F]FDA, 4-[¹⁸F]FPhe, and [¹⁸F]DAA1106 are obtained in good isolated radiochemical yields. Additionally, [¹⁸F]DAA1106 has been evaluated in a rat stroke model and demonstrates excellent potential for visualization of translocator protein 18 kDa overexpression associated with neuroinflammation after ischemic stroke.

Full text of DOI:10.1002/chem.201405586

DOI: 10.1002/chem.201405586
Full Paper
&
Radiopharmaceuticals
Copper-Mediated Aromatic Radiofluorination Revisited:
Efficient Production of PET Tracers on a Preparative Scale
[
a, b]
[a, b]
[a, b]
[a, b]
Boris D. Zlatopolskiy,[ Johannes Zischler,
Philipp Krapf,
Fadi Zarrad,
b, c]
[a]
[a]
[a, b]
Elizaveta A. Urusova,
Elena Kordys, Heike Endepols, and Bernd Neumaier*
1
8
Abstract: Two novel methods for copper-mediated aromatic
nucleophilic radiofluorination were recently reported. Evalua-
tion of these methods reveals that, although both are effi-
cient in small-scale experiments, they are inoperative for the
production of positron emission tomography (PET) tracers.
Since high base content turned out to be responsible for
low radiochemical conversions, a “low base” protocol has
l)(aryl)iodonium salts allows the preparation of F-labeled
arenes in excellent RCCs. The novel radiofluorination
method circumvents time-consuming azeotropic drying and
avoids the utilization of base and other additives, such as
cryptands. Furthermore, this procedure enables the produc-
18
18
tion of clinically relevant PET tracers; [ F]FDA, 4-[ F]FPhe,
18
and [ F]DAA1106 are obtained in good isolated radiochemi-
18
18
been developed which affords F-labeled arenes from diary-
liodonium salts and aryl pinacol boronates in reasonable
yields. Furthermore, implementation of our “minimalist” ap-
cal yields. Additionally, [ F]DAA1106 has been evaluated in
a rat stroke model and demonstrates excellent potential for
visualization of translocator protein 18 kDa overexpression
associated with neuroinflammation after ischemic stroke.
18
proach to the copper-mediated [ F]-fluorination of (mesity-
Introduction
mainly due to the relatively short half-life and the tiny
18
amounts of F used for tracer preparation. Moreover, during
18
Positron emission tomography (PET) is one of the leading
imaging modalities in clinical diagnostics. This technique
allows the visualization of different physiological and patholog-
ical processes at the molecular level with high resolution and
radiosyntheses with F radiation, safety issues have to be con-
sidered. Despite this, in recent years several novel fluorination
techniques have been successfully applied for the preparation
[1]
of radiofluorinated compounds. Pioneering approaches in-
+
18
[2]
extraordinary sensitivity, using probes labeled with suitable b
volving transition metal-mediated [ F]-fluorination were used
1
8
18
-emitting radioisotopes. Among these radionuclides, F holds
for the efficient preparation of F-labeled arenes and heteroar-
18
a dominant position, owing to its almost ideal decay proper-
ties with respect to half-life (109.8 min) and decay energy
enes, including electron-rich ones, from nucleophilic [ F]-fluo-
ride. The original procedure reported by Lee et al.
[3]
IV
(630 keV). The efficacy of PET examinations is strongly depen-
(Scheme 1A) comprises the reaction of a radiofluorinated Pd
II
dent on the development of PET tracers that selectively ad-
dress targets of high relevance. For this purpose, development
of novel high-yielding and simple radiofluorination techniques
is essential. However, in comparison to fluorination methods,
these techniques are still rather rare in radiochemistry. This is
complex prepared in advance with an appropriate Pd –aryl
18
compound, yielding the corresponding F-labeled arene after
[3,4]
reductive elimination (Scheme 1A).
A further development
of this method involves radiolabeling of an arylnickel(II) com-
plex with a radiofluorination agent generated in situ from a hy-
pervalent iodine oxidant [bis(onio)-substituted aryliodine(III)]
18
[5]
+
+
and [ F]-fluoride (Scheme 1B). These works are of fundamen-
tal importance for the development of radiochemistry.
[a] Dr. B. D. Zlatopolskiy, J. Zischler, P. Krapf, F. Zarrad, Dr. E. Kordys,
Priv.-Doz. Dr. H. Endepols, Prof. Dr. B. Neumaier
Institute of Radiochemistry and Experimental Molecular Imaging,
University Clinic Cologne, Kerpener Str. 62, 50937 Cologne (Germany)
Fax: (+49)221-47886851
Both methods, however, are rather impractical and difficult
to perform in automated synthesis modules due to the high
moisture sensitivity of the involved reaction components (radi-
E-mail: bernd.neumaier@uk-koeln.de
IV
+
+
ofluorinated Pd complex and hypervalent iodine oxidant).
[
b] Dr. B. D. Zlatopolskiy, J. Zischler, P. Krapf, F. Zarrad, E. A. Urusova,
Prof. Dr. B. Neumaier
Max Planck Institute for Metabolism Research
Gleueler Str. 50, 50931 Cologne (Germany)
These difficulties have so far precluded widespread application
of either procedure.
18
Recently, two methods for the F-labeling of arenes based
[
c] E. A. Urusova
Clinic of Nuclear Medicine, RWTH Aachen University
Pauwelsstr. 30, 52074 Aachen (Germany)
on copper-mediated aromatic nucleophilic radiofluorination
were reported. The first, developed by Sanford, Scott, and co-
[6]
+
workers (Scheme 2A), enables the preparation of diverse radi-
[
] Both authors contributed equally to this work
1
8
ofluorinated arenes via sterically-controlled Cu-mediated [ F]-
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201405586.
1
8
fluorination of (mesityl)(aryl)iodonium salts using [ F]KF/18-
Chem. Eur. J. 2015, 21, 5972 – 5979
5972
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
and radiofluorinations were carried out under air. The com-
plexes used as copper sources ((MeCN) CuOTf and
4
Cu(OTf) (py) for the Sanford/Scott and Gouverneur protocols,
2
4
respectively) have the advantage of being bench stable and
commercially available.
Our studies revealed that the pioneering methods for the
18
copper-catalyzed preparation of [ F]arenes, though efficient in
small-scale experiments, were not well-suited for PET tracer
production on a practical scale. The reason for the failure of
the two methods on preparative-scale application was eluci-
dated and prompted us to develop novel radiofluorination
procedures to overcome this shortcoming.
Results and Discussion
Initially, we studied the Cu(OTf) -mediated fluorination of mesi-
2
18
tyl-substituted diaryliodonium salts with [ F]KF/18-crown-6 for
1
8
[10]
F-labeling.
(Mesityl)(aryl)iodonium precursors of 3-
1
8
18
18
[
F]fluorobenzene, [ F]fluorobenzaldehyde (3-[ F]FBA), and 4-
18
18
Scheme 1. Pd- and Ni-mediated preparation of [ F]fluoroarenes according
[ F]fluoroanisole were selected as models of electron-neutral,
[
3]
[5a]
to Ritter and co-workers (A and B, respectively).
-
rich, and -deficient arene substrates, respectively (Table 1).
After some experimental refinements, the following procedure
18
À
was applied. F was trapped on an anion-exchange resin and
then eluted with a methanolic solution of K CO into a reaction
2
3
vial containing 18-crown-6. A small amount of K CO (0.5 mg
2
3
instead of usually used 2.8–3.5 mg) was sufficient to recover
18
À
F
quantitatively (>98%). Low-boiling methanol was com-
pletely removed at 70–808C within 2–3 min without the need
for azeotropic drying. Radioactivity loss during the evaporation
step was negligible (<2%). The residue was taken up in a solu-
tion of the corresponding iodonium precursor and Cu(OTf) in
2
DMF and heated for 20 min at 858C min under Ar. Afterwards,
an excess of water was added to quench the reaction and
18
À
completely solubilize surface-adsorbed F . Subsequently, ra-
18
Scheme 2. Copper-mediated F-labeling of diaryliodonium salts (A) and pi-
nacolyl arylboronates (B).
[1b]
diochemical conversions (RCCs) were determined via radio-
18
18
18
HPLC. [ F]Fluorobenzene, 4-[ F]fluoroanisole and 3-[ F]FBA,
were successfully prepared from the respective (mesityl)(ary-
l)iodonium tetrafluoroborate precursors in moderate RCCs of
41, 29, and 50%, respectively (Table 1, entries 1, 7, and 10). The
selectivity of radiofluorination was excellent in all experiments.
crown-6. For uncatalyzed fluorination of diaryliodonium salts,
ortho substituents direct the addition of fluoride to the more
[
7]
sterically crowded ring. In contrast, if copper catalysts are
used, the mesityl group directs oxidative addition and fluorina-
tion to the sterically less crowded aryl group on iodine, regard-
less of its electronic properties. Most importantly, this method
allows preparation of fluoroarenes with electron-donating sub-
stituents in high yields and with excellent selectivity. These
compounds are difficult to access via nucleophilic fluorination
reactions under commonly used non-catalytic conditions.
In the second method, proposed by Gouverneur and co-
18
18
[ F]Fluoride and traces of [ F]fluoromesitylene (<1%) were
the only radioactive impurities. A prolonged reaction time of
30 min resulted in a higher RCC of 64% in the case of
18
18
[ F]fluorobenzene but not for 4-[ F]fluoroanisole (Table 1, en-
tries 2 and 8, respectively). Additionally, (2,4,6-triisopropylphe-
nyl)(aryl)iodonium salt precursors also afforded radiofluorinat-
ed arenes, although in lower RCCs (Table 1, entries 5, 6, and 9).
In addition to tetrafluoroborates, other (mesityl)(aryl)iodonium
salts such as perchlorates, tosylates, bromides and triflates
were tested. Perchlorate and triflate salts were found to be
suitable substrates for radiofluorination (Table 1, entries 3 and
18
workers (Scheme 2B), [ F]arenes are obtained from pinacolyl
1
8
arylboronates (arylBPin) upon treatment with [ F]KF/K and
222
[
8]
Cu(OTf) (py) . They adapted the method of Cu(OTf) -mediat-
2
4
2
[11]
ed fluorination of aryltrifluoroborates and arylBPin with KF,
originally reported by Ye et al. for the preparation of
6). Tosylate or bromide, in contrast, afforded only low RCCs
(Table 1, entries 4 and 9).
18
[9]
[
F]arenes.
Both procedures were used for the radiofluorination of vari-
ous substrates containing electron-donating, -withdrawing and
neutral substituents. No special precautions were necessary
While our own studies were in progress, the radiolabeling
[6]
[8]
protocols of Sanford/Scott and Gouverneur were published.
We therefore carried out experiments to compare our proce-
dure (Table 1) with the published ones (Table 2 and Table 3).
-
Chem. Eur. J. 2015, 21, 5972 – 5979
www.chemeurj.org
5973
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Table 1. Cu(OTf)
donium salts using K
2
-mediated one-pot radiofluorination of (mesityl)(aryl)io-
Table 2. (MeCN)
nium salts: “High base” vs. “low base” protocol.
4
CuOTf-mediated radiofluorination of (mesityl)(aryl)iodo-
18
À
[a]
[a]
2
CO
3
in MeOH for elution of F .
[b]
Entry
Product
Aliquot
One-Pot
Entry
1
Precursor
Product
RCC [%]
41
RCC [%]
RCC [%]
“
High
“Low
base”
[c,d]
[e]
base”
1
2
3
56
84
36
1 (5)
<0.5 (3)
5 (1)
28
[
b]
2
3
4
64
35
0
56
26
[a] All syntheses were carried out manually. RCCs were determined by
radio-HPLC. Each experiment was carried out at least in triplicate. The
standard deviation of RCC did not exceed 10% of its mean value. [b] Syn-
theses of the radiolabeled compounds were carried out according to the
literature. Thereafter, the reaction mixture was cooled down to ambient
2
temperature and diluted with H O (2 mL). [c] Syntheses of the radiola-
beled compounds were carried out according to the literature. After
that, the reaction mixture was cooled down to ambient temperature and
2
diluted with H O (2 mL). [d] RCCs of [ F]fluoromesitylene are given in pa-
rentheses. [e]
5
6
36
35
[13]
[14]
18
1
8
À
F
(100–750 MBq) was eluted from an anion exchange
resin with K CO
2
3
(0.11 mg, 0.8 mmol) in water (200 mL). The resin was
washed with a solution of 18-crown-6 (0.47 mg, 1.8 mmol) in MeCN
1 mL). The combined fractions were concentrated under reduced pres-
sure at 958C and the residue was azeotropically dried with MeCN (2”
mL) under reduced pressure and air. The residue was taken up in a solu-
tion of (MeCN) CuOTf (4.2 mg, 12 mmol) in DMF (300 mL) and heated at
(
7
8
9
29
29
13
1
4
[b]
858C for 20 min under air. The reaction mixture was cooled to room tem-
perature and diluted with water (2 mL).
RCCs of 56, 84 and 36%, respectively. However, unacceptably
18
high losses of F activity due to adsorption onto vessel walls
were already observed before radiolabeling had been carried
1
0
50
1
8
out. Accordingly, only 20–50% of azeotropically dried [ F]KF/
8-crown-6 could be resolubilized in DMF. In an attempt to cir-
1
1
8
À
[a]
F
(100–750 MBq) was eluted from an anion exchange resin with
(0.5 mg, 4 mmol; in 1 mL H O) in MeOH (500 mL) and methanol was
cumvent this problem, a solution of the precursor and
K
2
CO
3
2
evaporated under He flow at 70–808C. The residue was taken up in DMF
200 mL) containing [Ar I]X (12 mmol), Cu(OTf) (4 mg, 12 mmol) and 18-
crown-6 (8.8 mg, 30 mmol). The resulting solution was heated at 858C for
0 min under Ar. Afterwards, the reaction mixture was cooled to room
temperature and diluted with water (2 mL). Reaction time: 30 min. All
syntheses were carried out manually. RCCs were determined by radio-
HPLC. Each experiment was carried out at least in triplicate. The standard
deviation of RCC did not exceed 10% of its mean value.
(MeCN) CuOTf in DMF was added directly to the vial contain-
4
18
(
2
2
ing dried [ F]KF/18-crown-6 (conventional protocol using
[6]
3
mgK CO and 15 mgK in MeCN). After addition, the mix-
2 3 222
2
b
ture was heated at 858C for 20 min (one-pot radiosynthesis
18
under “high base” conditions). Unexpectedly, F-labeled model
arenes were obtained in only 0.5–5% RCCs. The concurrent for-
18
mation of [ F]fluoromesitylene (up to 5%) was observed, indi-
cating a loss of selectivity. We assumed that (MeCN) CuOTf is
4
18
À
unstable under basic conditions. Therefore,
F
was eluted
In accordance with the Sanford/Scott protocol, the corre-
with essentially the same amount of K CO as that used in the
2
3
+
À
18
18
sponding [ArÀIÀMes] BF was treated with [ F]KF/18-crown-
experiments with small aliquots of [ F]KF/18-crown-6 (i.e., 0.1
instead of 3.5 mgK CO ; “low base” conditions). Radioactivity
4
[
12]
6
and (MeCN) CuOTf in DMF at 858C for 20 min under air.
4
2
3
18
Using small aliquots of [ F]KF/18-crown-6 (10–30 MBq) in DMF
recovery from the anion exchange resin was about 88% and
radioactivity loss during azeotropic drying did not exceed 3%
(decay-corrected; 12% not corrected for decay). Under “low
base” conditions, acceptable RCCs of 28, 56, and 26% were
18
obtained via redissolution of dried [ F]KF/18-crown-6, high
18
and selective F-incorporation was observed (Table 2) afford-
18
18
18
ing [ F]fluorobenzene, 4-[ F]fluoroanisole and 3-[ F]FBA in
Chem. Eur. J. 2015, 21, 5972 – 5979
www.chemeurj.org
5974
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
1
8
18
achieved for [ F]fluorobenzene, 4-[ F]fluoroanisole, and 3-
Once the counterproductive role of the base had been as-
certained, we studied whether the “minimalist” approach
18
[
F]FBA, respectively.
In the next set of experiments, the respective arylBPin radio-
could be applied in copper-mediated aromatic nucleophilic ra-
18
[16]
fluorination precursor, Cu(OTf) (py)4 and [ F]KF/K
in DMF
222
diofluorination.
According to this radiolabeling protocol,
2
18
were heated at 110 8C for 20 min under air, as described by
[ F]fluoride is directly eluted from the anion-exchange resin
with alcoholic solutions of the labeling precursors bearing
a quaternary ammonium, diaryliodonium, or triarylsulfonium
[
8]
Gouverneur and co-workers (Table 3). As with the Sanford/
18
Scott approach, this procedure afforded [ F]fluorobenzene, 4-
1
8
18
18
[
F]fluoroanisole, and 3-[ F]FBA in good RCCs of 67, 66 and
functionality. After alcohol removal, the resulting [ F]fluoride
18
7
6% if small aliquots of previously dried [ F]KF/K (ca. 30
onium salt is then simply heated in a suitable solvent. This
method eliminates not only the need for time-consuming
azeotropic drying steps but also, most importantly with re-
222
MBq) were used. In this case, about 20–25% of the radioactivi-
ty could not be resolubilized and remained on the vessel walls
after azeotropic drying. To avoid such high activity losses
a one-pot radiofluorination procedure was tested. Under “high
base” conditions, one-pot radiosyntheses provided radiofluori-
nated arenes in only 5–7% RCCs.
18
spect to copper-mediated aromatic F-labeling, the necessity
for a base or any other additives. Under “minimalist” condi-
tions, the copper mediated radiofluorination was carried out as
18
À
follows. F was eluted from the solid support with a solution
of a (mesityl)(aryl)iodonium salt precursor in MeOH almost
quantitatively (>97% recovery). After methanol removal, the
respective copper compound in DMF was added to the residue
and the resulting solution was heated at 858C for 20 min
before the reaction was quenched with an excess of water.
Gratifyingly, with (MeCN) CuOTf, (tBuCN) CuOTf or Cu(OTf) , ex-
In contrast, “low base” (0.06 mg K CO instead of 2.8 mg)
2
3
18
conditions afforded F-labeled model compounds in 41–64%
RCCs.
Table 3. Cu(OTf)
nates: “High base” vs. “low base” protocols.
2
(py)
4
-mediated radiofluorination of pinacolyl arylboro-
4
2
2
[
a]
18
cellent RCCs of all three F-labeled model compounds were
achieved from the corresponding tetrafluoroborates (Table 4).
Whereas
radiofluorinations
with
(MeCN) CuOTf
and
4
(
tBuCN) CuOTf could be carried out under air, radiolabeling
2
with Cu(OTf)2 required inert conditions. In the case of
[
b]
Entry
Product
Aliquot
One-Pot
RCC [%]
18
18
[
F]fluorobenzene and 4-[ F]fluoroanisole, an excellent radio-
RCC [%]
18
fluorination selectivity was observed. For 3-[ F]FBA some ero-
“
High base”
“Low base”
[d]
18
[
c]
sion of selectivity (RCC of [ F]fluoromesitylene=5%), as well
as the formation of several additional unidentified side-prod-
ucts (overall <7%) took place. The three copper salts afforded
comparable RCCs of radiofluorinated arenes. Advantageously
1
2
3
67
66
76
5
7
7
64
DMF solutions of (MeCN) CuOTf remained stable under ambi-
4
42
41
ent conditions for a prolonged period of time (at least 6 h). (4-
+
À
À
MeOPhÀIÀMes) X salts with different counterions X =OTs,
ClO , Br and OTf were tested as radiolabeling precursors. (4-
4
+
À
18
MeOPhÀIÀMes) OTs afforded 4-[ F]fluoroanisole in slightly
higher RCC than that of the corresponding tetrafluoroborate
precursor (Table 4, entry 4). The remaining iodonium salts were
not suitable for this labeling procedure.
[
a] All syntheses were carried out manually. RCCs were determined by
radio-HPLC. Each experiment was carried out at least in triplicate. The
standard deviation of RCC did not exceed 10% of its mean value. [b] Syn-
theses of the radiolabeled compounds were carried out according to the
literature. Thereafter, the reaction mixture was cooled down to ambient
temperature and diluted with H O (2 mL). [c] Syntheses of the radiola-
beled compounds were carried out according to the literature. Briefly,
[
1c]
In the next step, the feasibility of the novel radiofluorination
18
procedure for the preparative-scale production of F-labeled
2
[8]
18
arenes was evaluated. Accordingly, 4-[ F]fluoroanisole was pre-
1
8
[
F]fluoride (95–170 MBq) was eluted from an anion-exchange resin with
pared from the corresponding iodonium tetrafluoroborate pre-
a solution of K222 (13 mg) and K CO (2.8 mg) in 80% MeCN (1 mL). The
2
3
18
À
cursor starting from 12 GBq of F . Radiochemical yield (RCY)
of the isolated labeled product amounted to 59% within
solvent was evaporated under reduced pressure at 958C. The residue was
azeotropically dried with MeCN (2”1 mL). Thereafter, solutions of
2 4
Cu(OTf) (py) (15 mg, 22 mmol) and the corresponding arylboronic acid pi-
4
5 min.
nacol ester (60 mmol) in DMF (each in 150 mL) were added and the reac-
tion vial was purged with air. The reaction mixture was heated at 110 8C
for 20 min, cooled to room temperature and diluted with water (2 mL).
With an efficient radiofluorination procedure in hand, we
turned to the preparation of clinically relevant PET tracers.
1
8
À
18
18
18
[
d]
F
(100–750 MBq) was eluted from an anion exchange resin with
CO (0.06 mg, 0.43 mmol) and K2.2.2 (0.27 mg, 0.72 mmol)
6-[ F]Fluorodopamine (6-[ F]FDA, [ F]-4) targets specifically
[18]
a solution of K
2
3
catecholamine synthesis, storage, and secretion pathways.
Although originally developed as a sympathoneural imaging
in 80% MeCN (1 mL). The solvent was evaporated under reduced pres-
sure at 958C and the residue was azeotropically dried with MeCN (1 mL;
18
agent, 6-[ F]FDA has been confirmed as an excellent tracer for
the diagnosis and localization of chromaffin tumors, neuroblas-
tomas, ganglioneuromas, and metastatic pheochromocyto-
two times) under air. A solution of Cu(OTf)
DMF (150 mL) followed by a solution of arylboronic acid pinacol ester
60 mmol) in DMF/MeCN (5:1; 180 mL) were added to the residue and the
2 4
(py) (3.6 mg, 5.3 mmol) in
(
reaction mixture was heated at 110 8C for 20 min under air. The reaction
mixture was cooled to room temperature and diluted with water (2 mL).
[13,19]
mas.
Chem. Eur. J. 2015, 21, 5972 – 5979
www.chemeurj.org
5975
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Using our novel procedure, the iodonium salt precursor 6
afforded 71–94% RCC of protected 6-[ F]fluorodopamine
Table 4. Copper-mediated aromatic nucleophilic radiofluorination under “mini-
malist” conditions.
18
[a]
18
[
F]-1 (Scheme 3C; Table 4, entry 5). Since deprotection
with HI at 1308C was incomplete, even after 20 min (55%
conversion), the radiolabeled intermediate was first treated
with hydrazine hydrate in EtOH to quantitatively remove
the N-phthalimido group. Thereafter, the O-4-methylbenzyl
[
17]
18
groups were cleaved with HI to give 6-[ F]FDA. Starting
18
from 2.4 GBq of [ F]fluoride, a dose of 550 MBq of 6-
Entry Product
[(MeCN)
RCC [%]
4
CuOTf] Cu(OTf)
2
2
[(tBuCN) CuOTf]
18
[
F]FDA (46% RCY) was obtained as a ready-to-use solution
RCC [%] RCC [%]
in 4% ethanolic phosphate buffer within a total synthesis
time of 130 min. Optimization of the deprotection chemis-
try is under way and will be reported in due course.
1
2
90
86
90
78
68
72
–
90
77
–
18
18
18
4-[ F]Fluoro-l-phenylalanine (4-[ F]FPhe, [ F]-7) is well-
suited for the visualization of increased protein synthesis
rate, especially in peripheral and cerebral tumors. However,
18
[
b]
examples of clinical applications of 4-[ F]FPhe are rather
rare owing to its poor accessibility via nucleophilic radio-
3
[
14,24]
fluorination.
Last year a novel preparation method for
4
69
85
18
[24,25]
[
F]-7 was presented by Neumann et al.
this protocol, 4-[ F]FPhe was prepared in 20% RCY with
According to
18
À1
>
37 GBqmmol specific activity, from the respective (4-ani-
syl)(aryl)iodonium salt precursor 8 (Scheme 4A).
MeCN) CuOTf-mediated radiofluorination of iodonium
5
6
71–94
81–92
83
–
–
(
4
salt 9 under “minimalist” conditions afforded the radiofluori-
18
nated intermediate [ F]-2 in 81–92% RCC (Scheme 4B,
Table 4, entry 6). Deprotection of the labeled intermediate
was accomplished with 12 n HCl at 1408C for 10 min. Final-
–
18
ly, HPLC purification afforded [ F]-7 in a 4% ethanolic phos-
18
phate buffer solution. Enantiomerically pure 4-[ F]FPhe (>
9
6
1
4
8% ee; (R)-isomer was not observed) was isolated in 53–
6% RCY in two steps with specific activity of
09 GBqmmol (14 GBq 4-[ F]FPhe were obtained from
a
À1
18
7
4
93
92
–
1
8
À
9.5 GBq of F within 117 min). Thus, RCY and specific ac-
18
tivity of [ F]-7 were within the same range as those ach-
ieved for 4-(2-[ F]fluoroethyl)-l-tyrosine ([ F]FET). [ F]FET
group. represents a very popular surrogate tracer for phenylalanine
18
18
[26] 18
[
17]
-MeBzl=4-methylbenzyl
group;
NPhth=N-phthalimido
18
[a] [ F]Fluoride (0.1–10 GBq) was eluted from an anion exchange resin with
and is widely used, especially for diagnosis of brain
precursor (12 mmol) in MeOH (500 mL). Methanol was evaporated under He
flow at 70–808C, The residue was dissolved in a solution of a copper salt
(
[27]
18
tumors. However, in contrast to 4-[ F]FPhe, protein incor-
18
[25b]
12 mmol) in DMF (300 mL) and the resulting solution was heated at 858C for poration of [ F]FET is very low.
Tumor retention of 4-
1
8
18
2
0 min. The reaction mixture was diluted with water (2 mL). RCCs were deter- [ F]FPhe should be higher than that of [ F]FET, which
mined by radio-HPLC. Radiofluorinations with (MeCN)
tBuCN) CuOTf were carried out under air. Radiofluorinations with Cu(OTf)
were carried out under inert conditions. All syntheses were carried out manual-
4
CuOTf and
18
could make 4-[ F]FPhe well-suited for imaging of brain
(
2
2
tumors. In our opinion, this justifies its future preclinical
ly. Each experiment was carried out at least in triplicate. The standard deviation and, probably, clinical evaluation.
of RCC usually did not exceed 20% of its mean value. Where the standard devi-
ation of RCCs exceeded 20% of its mean value, the range of RCCs is given.
DAA1106 (3; Scheme 5) is a very potent and selective ag-
onist at the translocator protein 18 kDa (TSPO). TSPO is
a biomarker of brain inflammation and reactive gliosis,
which are characteristic for a number of neurodegenerative
[b] Iodonium tosylate precursor was used.
18
[28]
Several procedures for the preparation of 6-[ F]FDA have
and psychiatric diseases, stroke and brain tumors.
[
20]
18
11
been reported. The first radiosynthesis of [ F]-4 via aromatic
[ C]DAA1106 is intensely used for monitoring of these pathol-
[21]
[29]
11
nucleophilic radiofluorination, published by Ding et al., con-
sisted of four cumbersome reaction steps (Scheme 3A). Quite
ogies. However, the short half-life of C (t =20 min) pre-
1
=
2
cludes commercialization and hampers its broad application.
18
[30]
[8]
recently, the synthesis of 6-[ F]FDA using the respective (4-ani-
Suzuki et al.,
as well as Gouverneur and co-workers,
18
[22]
18
syl)(aryl)iodonium precursor 5 and [ F]KF/K in toluene was
reported (Scheme 3B).
showed that F-labeling of the corresponding (4-anisyl)(aryl)io-
2
22
[
23]
18
After deprotection of the radiola-
donium salt 10 or arylBPin 11 precursor, afforded [ F]DAA1106
in 46 and 59% RCC, respectively. However, [ F]DAA1106 has
18
18
beled intermediate with HI, 6-[ F]FDA was isolated in 23%
RCY over two steps.
not yet been isolated or biologically tested.
Chem. Eur. J. 2015, 21, 5972 – 5979
www.chemeurj.org
5976
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
18
18
18
Scheme 3. Preparation of 6-[ F]fluorodopamine (6-[ F]FDA, [ F]-4).
18
18
Scheme 5. Preparation of [ F]DAA1106 ([ F]-3).
18
18
18
Scheme 4. Radiosynthesis of 4-[ F]fluoro-l-phenylalanine (4-[ F]FPhe, [ F]-
).
7
1
8
Figure 1. [ F]DAA1106 MR-PET six days after anterior cerebral artery occlu-
sion, showing transverse (1) and horizontal images (2), with the ischemic
lesion in the anterior cingulate cortex visible as hyperintensity (white arrow-
heads in MR- and PET-MR-images). The peri-infarct zone is highlighted by
red circles. Section levels are indicated by white dashed lines in the insert
Under our standard “minimalist” conditions, precursor 12
was efficiently radiofluorinated to give [ F]DAA1106 in 89–
18
9
6% RCC (Scheme 5C). The RCY of the tracer with a specific ac-
À1
tivity of 66 GBqmmol after HPLC isolation and formulation as
(
bottom left).
a solution in EtOH amounted to 60% (1.2 GBq were obtained
1
8
À
[31]
from 2.4 GBq of F in a total synthesis time of 45 min).
Additionally, the amount of Cu in the final product solution
cingulate cortex. At day 6 after induction of stroke, the inflam-
matory processes in the ischemic core and the surrounding
18
18
18
of 4-[ F]fluoroanisole, 4-[ F]FPhe, and [ F]DAA1106 was deter-
mined by using inductively coupled plasma mass spectrometry
18
tissue were studied using [ F]DAA1106-PET. PET images were
projected on a structural MR image from the same animal, ac-
(
ICP-MS). The Cu content varied from 0.6–1.4 ppm. This is sig-
18
nificantly below any level of concern according to the ICH
quired on the following day. The [ F]DAA1106 PET signal over-
lapped with the ischemic lesion, visible as a hyperintensity in
[
32]
Guideline of Elemental Impurities (Q3D).
18
18
Visualization of neuroinflammation by [ F]DAA1106-PET was
studied in a rat stroke model (Figure 1). Ischemic stroke was in-
duced by occlusion of the anterior cerebral artery (ACA) using
the MR image (Figure 1, white arrowheads). The [ F]DAA1106
PET image displayed the ischemic lesion at most section levels
with excellent signal-to-noise ratio. A mismatch with the MR
image was observed in the most rostral part (Figure 1, red cir-
cles), where the radioactive signal extended beyond the in-
[33]
stereotaxic injection of the vasoconstrictor endothelin-1.
ACA occlusion resulted in an ischemic lesion in the anterior
Chem. Eur. J. 2015, 21, 5972 – 5979
www.chemeurj.org
5977
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
farcted core. This was due to inflammatory processes in the
[5] a) E. Lee, J. M. Hooker, T. Ritter, J. Am. Chem. Soc. 2012, 134, 17456–
17458; b) H. Ren, H. Y. Wey, M. Strebl, R. Neelamegam, T. Ritter, J. M.
Hooker, ACS Chem. Neurosci. 2014, 5, 611 –615.
[
34]
peri-infarct zone after stroke.
[
6] N. Ichiishi, A. F. Brooks, J. J. Topczewski, M. E. Rodnick, M. S. Sanford, P. J.
Scott, Org. Lett. 2014, 16, 3224–3227.
[
7] J. H. Chun, S. Lu, Y. S. Lee, V. W. Pike, J. Org. Chem. 2010, 75, 3332–
Conclusion
3338.
We have established a robust “low base” protocol for the effi-
[8] M. Tredwell, S. M. Preshlock, N. J. Taylor, S. Gruber, M. Huiban, J. Passchi-
18
er, J. Mercier, C. Genicot, V. Gouverneur, Angew. Chem. Int. Ed. 2014, 53,
cient copper-mediated preparation of [ F]arenes from diarylio-
donium salts and aryl pinacol boronates. This method over-
comes the problem of insufficient stability of copper catalysts
like (MeCN) CuOTf and Cu(OTf) (py) under the basic condi-
7
751–7755; Angew. Chem. 2014, 126, 7885–7889.
9] Y. D. Ye, S. D. Schimler, P. S. Hanley, M. S. Sanford, J. Am. Chem. Soc.
013, 135, 16292–16295.
[10] N. Ichiishi, A. J. Canty, B. F. Yates, M. S. Sanford, Org. Lett. 2013, 15,
134–5137.
[
2
4
2
4
5
tions used in the previously reported procedures. Whereas
these “high” base radiofluorination procedures afforded very
low RCCs on a preparative scale, our approach enables the
preparation of radiolabeled arenes in reasonable yields. Fur-
[
[
11] See Supporting Information for complete experimental details.
12] We found that the application of synthetic instead of ambient air signif-
icantly decreased deviations of RCCs. Consequently, in all experiments
described here synthetic air was applied.
[
13] D. S. Goldstein, G. Eisenhofer, J. A. Flynn, G. Wand, K. Pacak, Hyperten-
sion 2004, 43, 907–910.
thermore, we have developed
a novel radiofluorination
method that combines the advantages of our previously re-
ported “minimalist” approach with the exceptional capabilities
of copper-mediated aromatic nucleophilic radiofluorination.
The versatility and scope of the novel radiofluorination proce-
[
[
14] K. Mineura, M. Kowada, F. Shishido, Surg. Neurol. 1989, 31, 468–469.
15] B. D. Zlatopolskiy, R. Kandler, D. Kobus, F. M. Mottaghy, B. Neumaier,
Chem. Commun. 2012, 48, 7134–7136.
[16] R. Richarz, P. Krapf, F. Zarrad, E. A. Urusova, B. Neumaier, B. D. Zlatopol-
18
skiy, Org. Biomol. Chem. 2014, 12, 8094–8099.
dure was demonstrated by the preparation of three F-labeled
model arenes and by the production of clinical doses of three
[
[
[
17] N. Malik, B. Zlatopolskiy, C. Solbach, W. Voelter, S. Reske, H.-J. Machulla,
J. Radioanal. Nucl. Chem. 2011, 288, 563–569.
18] G. Eisenhofer, D. Hovevey-Sion, I. J. Kopin, R. Miletich, K. L. Kirk, R. Finn,
D. S. Goldstein, J. Pharm. Exp. Ther. 1989, 248, 419–427.
19] a) I. Ilias, B. Shulkin, K. Pacak, Trends Endocrinol. Metab. 2005, 16, 66–72;
b) D. S. Goldstein, G. Eisenhofer, B. B. Dunn, I. Armando, J. Lenders, E.
Grossman, C. Holmes, K. L. Kirk, S. Bacharach, R. Adams, P. Herscovitch,
I. J. Kopin, J. Am. Coll. Cardiol. 1993, 22, 1961–1971.
18
18
18
PET tracers, [ F]FDA, 4-[ F]FPhe, and [ F]DAA1106, in good to
excellent radiochemical yields. Furthermore, this novel method
circumvents time-consuming azeotropic drying and avoids the
application of base and other additives, such as cryptands. Ad-
18
ditionally, [ F]DAA1106 was evaluated in a rat stroke model
and demonstrated excellent potential for visualization of TSPO
overexpression associated with neuroinflammation after ische-
mic stroke.
[20] a) O. Eskola, T. J. Gronroos, A. Naum, P. Marjamaki, S. Forsback, J. Berg-
man, S. Lankimaki, J. Kiss, T. Savunen, J. Knuuti, M. Haaparanta, O. Solin,
Eur. J. Nucl. Med. Mol. Imaging 2012, 39, 800–810; b) M. A. Channing,
J. L. Musachio, J. J. Kusmierz in Radiochemical Syntheses, Wiley, 2012,
pp. 125–138.
[
[
[
21] Y. S. Ding, J. S. Fowler, S. J. Gatley, S. L. Dewey, A. P. Wolf, D. J. Schlyer, J.
Med. Chem. 1991, 34, 861–863.
22] S. DiMagno. Fluorination of aromatic ring systems; US Patent 13/
Acknowledgements
125,209 (, 2012) (supplement 1).
We thank F. Wharton for her careful proofreading of the manu-
script, and Dr. H. Frauendorf and G. Udvarnoki for the measure-
ment of mass spectra.
23] A. Vavere, E. Butch, B. Shulkin, S. Snyder, J. Nucl. Med. 2014, 55, 1410.
[24] C. Lemaire, M. Guillaume, L. Christiaens, A. J. Palmer, R. Cantineau, Int. J.
Rad. Appl. Instrum. A. 1987, 38, 1033–1038.
[
25] a) K. Neumann, K. Glaspy, A. Vavere, S. Snyder, S. DiMagno, J. Nucl. Med.
013, 54, 1069 (supplement 2); b) W. Bodsch, H. H. Coenen, G. Stocklin,
2
Keywords: copper · fluorination · imaging agents · positron
emission tomography · radiopharmaceuticals
K. Takahashi, K. A. Hossmann, J. Neurochem. 1988, 50, 979–983.
26] K. Hamacher, H. H. Coenen, Appl. Radiat. Isot. 2002, 57, 853–856.
27] a) C. Juhasz, S. Dwivedi, D. O. Kamson, S. K. Michelhaugh, S. Mittal, Mol.
Imaging 2014, 13, 1–16; b) M. Hutterer, M. Nowosielski, D. Putzer, N. L.
Jansen, M. Seiz, M. Schocke, M. McCoy, G. Gobel, C. la Fougere, I. J. Vir-
golini, E. Trinka, A. H. Jacobs, G. Stockhammer, Neuro Oncol. 2013, 15,
[
[
[
1] a) J. Ermert, BioMed Res. Int. 2014, 15; b) K. Kettenbach, H. Schieferstein,
T. L. Ross, Biomed Res. Int. 2014, 361329; c) T. Rühl, W. Rafique, V. T. Lien,
P. J. Riss, Chem. Commun. 2014, 50, 6056–6059; d) B. D. Zlatopolskiy, P.
Krapf, R. Richarz, H. Frauendorf, F. M. Mottaghy, B. Neumaier, Chem. Eur.
J. 2014, 20, 4697–4703; e) M. Tredwell, V. Gouverneur, Angew. Chem.
Int. Ed. 2012, 51, 11426–11437; Angew. Chem. 2012, 124, 11590–11602;
f) B. D. Zlatopolskiy, R. Kandler, F. M. Mottaghy, B. Neumaier, Appl.
Radiat. Isot. 2012, 70, 184–192.
341–351; c) V. Dunet, C. Rossier, A. Buck, R. Stupp, J. O. Prior, J. Nucl.
Med. 2012, 53, 207–214; d) H. J. Wester, M. Herz, W. Weber, P. Heiss, R.
Senekowitsch-Schmidtke, M. Schwaiger, G. Stçcklin, J. Nucl. Med. 1999,
4
0, 205–212.
[
28] a) E. R. O’Brien, V. Kersemans, M. Tredwell, B. Checa, S. Serres, M. S. Soto,
V. Gouverneur, D. Leppert, D. C. Anthony, N. R. Sibson, J. Nucl. Med.
2
014, 55, 275–280; b) R. Rupprecht, V. Papadopoulos, G. Rammes, T. C.
[
2] a) A. F. Brooks, J. J. Topczewski, N. Ichiishi, M. S. Sanford, P. J. Scott,
Chem. Sci. 2014, 5, 4545–4553; b) S. H. Liang, N. Vasdev, Angew. Chem.
Int. Ed. 2014, 53, 11416–11418; c) S. H. Liang, N. Vasdev, Angew. Chem.
Baghai, J. Fan, N. Akula, G. Groyer, D. Adams, M. Schumacher, Nat. Rev.
Drug Discovery 2010, 9, 971–988.
2
014, 126, 11600–11602; d) T. Liang, C. N. Neumann, T. Ritter, Angew.
Chem. Int. Ed. 2013, 52, 8214–8264; Angew. Chem. 2013, 125, 8372–
423; e) E. L. Cole, M. N. Stewart, R. Littich, R. Hoareau, P. J. H. Scott,
[29] a) A. Takano, R. Arakawa, H. Ito, A. Tateno, H. Takahashi, R. Matsumoto,
Y. Okubo, T. Suhara, Int. J. Neuropsychopharmacol. 2010, 13, 943–950;
b) J. Maeda, T. Suhara, M. R. Zhang, T. Okauchi, F. Yasuno, Y. Ikoma, M.
Inaji, Y. Nagai, A. Takano, S. Obayashi, K. Suzuki, Synapse 2004, 52, 283–
291; c) M.-R. Zhang, T. Kida, J. Noguchi, K. Furutsuka, J. Maeda, T.
Suhara, K. Suzuki, Nucl Med Biol 2003, 30, 513–519.
8
Curr. Top. Med. Chem 2014, 14, 875–900.
[
3] E. Lee, A. S. Kamlet, D. C. Powers, C. N. Neumann, G. B. Boursalian, T.
Furuya, D. C. Choi, J. M. Hooker, T. Ritter, Science 2011, 334, 639–642.
4] A. S. Kamlet, C. N. Neumann, E. Lee, S. M. Carlin, C. K. Moseley, N. Ste-
phenson, J. M. Hooker, T. Ritter, PloS one 2013, 8, e59187.
[
[30] M.-R. Zhang, K. Kumata, K. Suzuki, Tetrahedron Lett. 2007, 48, 8632–
8635.
Chem. Eur. J. 2015, 21, 5972 – 5979
www.chemeurj.org
5978
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
1
8
[
31] Owing to the low solubility of [ F]DAA1106 in water an ethanolic solu-
tion of the tracer was diluted with aqueous PEG-400 immediately
before application. See the Supporting Information for additional ex-
perimental details.
[34] S. Rojas, A. Martin, M. J. Arranz, D. Pareto, J. Purroy, E. Verdaguer, J.
Llop, V. Gomez, J. D. Gispert, O. Millan, A. Chamorro, A. M. Planas, J.
Cereb. Blood Flow Metab. 2007, 27, 1975–1986.
[
32] Guidline for Elemental Impurities, Q3D, ICH, 2013.
33] H. Endepols, H. Mertgens, H. Backes, U. Himmelreich, B. Neumaier, R.
Graf, G. Mies, J. Neurosci. Meth. 2014, in press, DOI: 101016/j.jneu-
meth.2014.11003.
Received: October 9, 2014
Published online on February 24, 2015
[
Please note: Minor changes have been made to this manuscript and Sup-
porting Information since its publication in Chemistry—A European Journal
Early View. The Editor.
Chem. Eur. J. 2015, 21, 5972 – 5979
www.chemeurj.org
5979
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim