Optimization and synthesis of an 18F-labeled dopamine D3 receptor ligand using [18F]fluorophenylazocarboxylic tert-butylester

Article abstract of DOI:10.1002/jlcr.3361

There is still no efficient fluorine-18-labeled dopamine D₃ subtype selective receptor ligand for studies with positron emission tomography. We aim at improving the D₃ selectivity and hydrophilicity of a candidate ligand by changing the substitution pattern to a 2,3-dichlorophenylpiperazine and hydroxylation of the butyl chain. The compound [¹⁸F]3 exhibited D₃ affinity of K_i = 3.6 nM, increased subtype selectivity (K_i(D₂/D₃) = 60), and low affinity to 5-HT_1A and α₁ receptors (K_i (5-HT_1A/D₃) = 34; K_i (α₁/D₃) = 100). The two-step radiosynthesis was optimized for analog [¹⁸F]4 by reducing the necessary concentration of the precursor amine (57 mM), which reacted with [¹⁸F]fluorophenylazocarboxylic tert-butylester under basic conditions. The optimization of the base (Cs ₂CO₃, 23 mM) and the adjustment of reaction temperature led to the radiochemical yield of 63% after 5 min at 35C. The optimized reaction conditions were transferred on to the synthesis of [¹⁸F]3 with an overall non-decay corrected yield of 8-12% in a specific activity of 32-102 GBq/μmol after a total synthesis time of 30-35 min. This provides a D ₃ radioligand candidate with improved attributes concerning selectivity and radiosynthesis for further preclinical studies.

Full text of DOI:10.1002/jlcr.3361

Research article
Received 15 October 2015,
Accepted 17 November 2015
Published online 28 December 2015 in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/jlcr.3361
18
Optimization and synthesis of an F-labeled
dopamine D3 receptor ligand using
18
[ F]fluorophenylazocarboxylic tert-butylester
Natascha Nebel,a Simone Maschauer,a Carsten Hocke,a Harald Hübner,b
b
a
*
Peter Gmeiner, and Olaf Prante
There is still no efficient fluorine-18-labeled dopamine D3 subtype selective receptor ligand for studies with positron
emission tomography. We aim at improving the D3 selectivity and hydrophilicity of a candidate ligand by changing the
substitution pattern to a 2,3-dichlorophenylpiperazine and hydroxylation of the butyl chain. The compound [18F]3 exhibited
D3 affinity of Ki = 3.6 nM, increased subtype selectivity (Ki(D2/D3) = 60), and low affinity to 5-HT1A and α1 receptors (Ki (5-HT1A
/
D3) = 34; Ki (α1/D3) = 100). The two-step radiosynthesis was optimized for analog [18F]4 by reducing the necessary
concentration of the precursor amine (57 mM), which reacted with [18F]fluorophenylazocarboxylic tert-butylester under basic
conditions. The optimization of the base (Cs2CO3, 23 mM) and the adjustment of reaction temperature led to the
radiochemical yield of 63% after 5 min at 35°C. The optimized reaction conditions were transferred on to the synthesis of
[
18F]3 with an overall non-decay corrected yield of 8-12% in a specific activity of 32-102 GBq/μmol after a total synthesis time
of 30-35 min. This provides a D3 radioligand candidate with improved attributes concerning selectivity and radiosynthesis
for further preclinical studies.
Keywords: dopamine D3 receptor; positron emission tomography; phenylazocarboxylic esters; nucleophilic aromatic 18F-fluorination
For the subtype selective visualization and a more detailed
investigation of the D3 receptor, positron emission tomography
Introduction
In the central nervous system (CNS), dopamine receptors are
generally involved in the control of locomotion, cognition, affect,
and neuroendocrine secretion.1 The various dopamine receptor
subtypes (D1-like receptors D1 and D5; D2-like receptors D2, D3
and D4) are differently expressed in the CNS. Whereas D1 and
D2 dopamine receptors are the most abundant subtypes in the
CNS, the D3 receptor is strongly expressed in the nucleus
accumbens and the islands of Calleja of the limbic system.2,3
Being preferentially located in brain regions, which have an
impact on emotional and cognitive functions, the D3 receptor
is especially useful in addressing behavioral properties, such as
reinforcement and reward.1,4 Surveying the D3 receptor as a
target for neuroleptics showed that inhibition of the locomotor
system, a side effect of D2 receptor-blocking, does not occur with
D3 receptor-blocking during treatment of schizophrenia.3,5,6
Furthermore, various pharmacological studies have investigated
the activation and inhibition of the D3 system as an interesting
therapeutic target for the treatment of cocaine and alcohol
abuse as well as for motoric dysfunctions, such as Parkinson’s
disease, dyskinesia, and restless leg syndrome.2,7–9
(PET) is in principle well suited to detect, locate, and quantify
receptors taking advantage of its combined molecular
specificity, sensitivity, and spatial resolution that is unique
among imaging modalities.11,12 However, the commonly used PET
radioligands, such as [11C]raclopride and [18F]fallypride (Figure 1),
are nonselective D2/D3 radioligands. Recently, the D3 subtype-
preferring radioligand [11C]PHNO has been characterized in more
detail (Figure 1); however, its use for the in vivo imaging of the
D3 receptor is strongly limited.8,13 Thus, developing an efficient
dopamine D3 selective 18F-labeled tracer, with its favorable
physical properties, remains a challenging task to the present.12,14
aDepartment of Nuclear Medicine, Molecular Imaging and Radiochemistry,
Friedrich Alexander University Erlangen-Nürnberg (FAU), Schwabachanlage 6,
D-91054, Erlangen, Germany
bDepartment of Chemistry and Pharmacy, Medicinal Chemistry, Emil Fischer
Center, Friedrich Alexander University Erlangen-Nürnberg (FAU), Schuhstrasse
19, D-91052, Erlangen, Germany
However, the development of D3 subtype selective ligands is
still challenging because of highly conserved structures between
the D2-like subtypes D2 and D3. In addition, dopamine receptor
ligands also frequently exhibit good binding affinities to other
monoamine receptors such as the serotonin receptor (5-HT)
and the adrenergic α1 receptor.4,10
*Correspondence to: Olaf Prante, Molecular Imaging and Radiochemistry,
Friedrich-Alexander University.
E-mail: olaf.prante@uk-erlangen.de
Contract/grant sponsor: Deutsche Forschungsgemeinschaft; contract/grant
number: DFG; PR 677/6-3
J. Label Compd. Radiopharm 2016, 59 48–53
Copyright © 2015 John Wiley & Sons, Ltd.

N. Nebel et al.
Figure 1. Commonly used D2/D3 radioligands for PET.
A new method for the synthesis of radiolabeled aromatic
bearing compounds using [18F]fluorophenylazocarboxylic esters,
in this case tert-butyl-2-(4-[18F]fluorophenyl) azocarboxylate
([18F]1) led to the 18F-labeled ligand [18F]4 with good affinity
(Ki (D3) = 3.5 nM) toward the D3 receptor (Scheme 1, Table 1).15
However, the binding of [18F]4 is not selective for the dopamine
Experimental
General
All solvents were of p.a. grade and purchased from Carl Roth GmbH
(Karlsruhe, Germany), all chemicals besides the ones mentioned were
purchased from Alfa Aesar (Karlsruhe, Germany), Carl Roth GmbH
D3 receptor, exhibiting nanomolar affinity toward the interfering (Karlsruhe, Germany), Fluka (Steinheim, Germany), Merck (Darmstadt,
Germany) or Sigma Aldrich (Steinheim, Germany), in the highest
available quality and used without further purification. Special emphasis
was on anhydrous solvents for 18F-radiosyntheses.
TLC plates (silica gel 60, F254, Merck) were bought ready for use, and
for column chromatography silica gel 60 with a particle size of 40-60 μm
(Merck) was used. NMR-Spectra were recorded with an AVANCE 360/600
(Bruker) and TMS as standard. High-resolution mass spectra (HRMS) were
processed with a GCmate (JEOL) with EI+ as ionization mode and a
resolution of M/ΔM = 3000.
5-HT1A receptor and the adrenergic α1 receptor.
The applied 18F-labeling method starting from azocarboxylic
ester [18F]1 turned out to be very attractive being fast and
efficient. The two-step radiosynthesis involved only one final
HPLC purification and proceeds within a total synthesis time
of only 35–40 min with good overall yields of 30–35% (non-
decay corrected).15 The reaction occurs under mild conditions,
the aromatic core of phenylazocarboxylic esters being highly
activated toward nucleophilic aromatic 18F-substitution, and
The HPLC-System (Series 1100, Agilent) was equipped with a VWD UV-
subsequent reactions at the carbonyl group being easily Lamp (detection at 254 nm, if not otherwise stated) and for radio-HPLC was
additionally connected to a radio-detector (500 TR Series, Packard). The
conditions used for preparative and analytical HPLC are described in methods
1-4. Method 1: analytical column: Chromolith RP-18e, 100 × 4.6 mm, flow rate:
4 mL/min, solvent A: water (0.1% TFA), solvent B: acetonitrile (0.1% TFA),
gradient A/B: 90:10 to 50:50 in 5 min, λ = 214 nm. Method 2:analytical
column: Luna C18(2) 5 μm, 150 × 4.6 mm, flow rate: 1.5 mL/min, solvent:
A: water (0.1% TFA), solvent B: acetonitrile (0.1% TFA), gradient A/B:
85:15 to 10:90 in 30 min. Method 3: semipreparative column: Kromasil
C8, 125 × 8 mm, flow rate: 4 mL/min, solvent: A: water (0.1% TFA), solvent
B: acetonitrile (0.1% TFA), gradient A/B: 75:25 to 40:60 in 30 min. Method 4:
induced and directed through choice of leaving group (ester)
and conditions (acidic/basic).15,16
Aiming at the improvement of the selectivity of [18F]4, a 2,3-
dichloro-substitution pattern, opposed to the 2-methoxy-
substitution at the phenylpiperazine (Scheme 2), could be
beneficial because it has in the past offered both good affinity
and selectivity toward the D3 receptor.17–20 However, a
dichloro-substitution increases lipophilicity, which could lead
to unsatisfactory bioavailability and biodistribution, such as
high non-specific binding and non-passage of the blood brain preparative column: Luna C18(2) 5 μm, 250 × 10 mm, flow rate: 4 mL/min,
barrier.13,18,21,22 Consequently, a hydroxyl function on the
solvent A: water 0.1% TFA, solvent B: acetonitrile 0.1% TFA, gradient A/B:
70:30 to 40:60 in 25 min.
propyl-linking-chain was introduced at the same time,
following the method of Grundt et al.,19 to lower the lipophilic
character of the ligand, leading to the 2,3-dichloro substituted
azaphenylcarboxamide
compound of this study.
Chemistry
[
18F]3 (Scheme 2) as the target
Compounds 115 and 219 were synthesized as described previously.
Analytical data of the compounds were in accordance with the
respective literature data.
Moreover, another aim of this research was to evaluate the
second step of the radiosynthesis leading to [18F]4 (Scheme 2)
regarding possible optimization. The reaction forming the
carboxamide occurs only under very basic conditions and the
threshold of the amount of base (K2CO3) was to be determined
or even its replacement due to its poor solubility in the
already-evaluated solvent ethanol.15 Finally, the optimized
radiosynthesis was transferred on to the synthesis of [18F]3,
providing a D3 radioligand candidate with improved attributes
toward dopamine D3 receptor imaging concerning in vitro
selectivity and an optimized radiosynthesis.
(E)-2-(4-Fluorophenyl)-N-(-3-hydroxy-4-(4-(2,3-dichlorophenyl)
piperazine-1-ylbutyl)diazene-1-carboxamide (3)
The synthesis of 3 followed a procedure described in the literature
(Scheme 1).17 Briefly,
a solution of tert-butyl 2-(4-fluorophenyl)
azocarboxylate 1 (230 mg, 1.00 mmol) in ethyl acetate (EtOAc) (3 mL)
was stirred together with K2CO3 (450 mg, 3.26 mmol) and amine 2
(160 mg, 0.50 mmol) was added. The mixture was stirred at room
temperature for approximately 72 h. Afterwards, the mixture was diluted
Scheme 1. Synthesis of reference compound 3. Reagents and conditions: (a) K2CO3, EtOAc, room temperature, 72 h.
J. Label Compd. Radiopharm 2016, 59 48–53
Copyright © 2015 John Wiley & Sons, Ltd.
www.jlcr.org

N. Nebel et al.
Table 1. Binding affinities of 3 and 4 to the human dopamine receptor subtypes D1, D2long, D2short, D3, D4.4, and D5, the
serotonin receptors 5-HT1A and 5-HT2 and the adrenergic α1 receptor (mean values of n = 2). a values taken from Ref. 15.
Ki values (nM)
[3H]spiperone
[3H]WAY 600135
[3H]ketanserin
[3H]prazosin
D1
10000
n.d.
D5
>70000
n.d.
D2long
490
38
D2short
190
36
D3
3.6
3.5
D4.4
1800
94
5-HT1A
120
6.4
5-HT2
56
43
α1
360
5.2
3
4a
with H2O and the product extracted with EtOAc (3 × 30 mL). The
combined organic phases were washed with a saturated NaCl solution,
dried over Na2SO4 and the solvent removed under reduced pressure.
The crude product was first purified by column chromatography (Rf:
MeOH/CH2Cl2 1/9 = 0.4) and then by preparative HPLC (method 4 tR:
14.64 min) to give 3 (32 mg, 0.06 mmol, 14%) as orange solid. 1H NMR
concentration per each assay tube of 80 and 30 μg/mL for p5-HT1A and
pα1, respectively, in the presence of the radioligands [3H]WAY100635
(specific activity of 80 Ci/mmol, Biotrend, Cologne, Germany) (0.22 nM final
concentration, KD = 0.11 nM, Bmax = 70 fmol/mg), and [3H]prazosin
(specific activity of 83 Ci/mmol, PerkinElmer, Rodgau, Germany) (0.21 nM
final concentration, KD = 0.17 nM, Bmax = 160 fmol/mg), respectively. Protein
(360 MHz, CDCl3) δ (ppm) = 1.70–1.82 (m, 1H), 1.86–2.06 (m, 1H), 3.09– concentration was determined by the method of Lowry using bovine serum
3.47 (m, 7H), 3.53–3.68 (m, 1H), 3.69–3.87 (m, 2H), 3.93 (d, J = 11.5 Hz,
1H), 4.32 (bs, 1H), 6.98 (dd, J = 7.9, 1.6 Hz, 1H), 7.18 (d, J = 3.2 Hz, 1H),
7.19–7.24 (m, 2H), 7.22–7.31 (m, 1H), 7.95–8.01 (m, 2H). 13C NMR
(91 MHz, CDCl3) δ (ppm) = 29.71, 34.31, 36.96, 48.02, 55.61, 62.95,
63.58, 112.00 116.59 (d, JCF = 23.2 Hz), 119.84, 121.52, 126.63 (d,
albumin as standard.24 The resulting competition curves of the receptor
binding experiments were analyzed by nonlinear regression using the
algorithms in PRISM 6.0 (GraphPad Software, San Diego, CA). The data were
initially fit using a sigmoid model to provide an IC50 value, representing the
concentration corresponding to 50% of maximal inhibition. IC50 values
were transformed to Ki values according to the equation of Cheng and
Prusoff.25
J
CF = 9.7 Hz), 136.09, 147.60 (d, JCF = 3.0 Hz), 152.23, 161.57 (d,
JCF = 53.9 Hz), 166.32 (d, JCF = 257.0 Hz). HRMS (467.1291) found:
467.1289. Purity: HPLC method 1: 214 nm, tR = 4.04 min, 97%, method
2: 254 nm, tR = 11.16 min, 96%.
Radiochemistry
Fluorine-18 was provided by PET Net GmbH (Erlangen, Germany) and
delivered as n.c.a. [18F]fluoride, fixed on an anion exchange cartridge
(QMA, Waters) with an average radioactivity range of 0.6-1.0 GBq. The
identity of radioligands was determined using reference compounds
bearing the natural fluorine isotope. The reference compound was
coinjected with the radioligand for HPLC analysis and the identity
confirmed by identical retention times using HPLC methods 1 and 2.
When using radio-TLC (silica gel 60, F254, Merck) to determine the
radiochemical yield (RCY), the Rf-value of the reference compounds was
used as reference (Rf (3) = Rf (4) = 0.4 (MeOH/CH2Cl2 1/9)).
Receptor binding studies
Radioligand binding assays were carried out following a procedure
described previously.23 Briefly, for competition binding experiments
with the human D2L, D2S, D3 and D4.4 receptors, preparations of
membranes from CHO cells stably expressing the corresponding
receptor were used together with [3H]spiperone (specific activity of
81 Ci/mmol, PerkinElmer, Rodgau, Germany) at a final concentration
of 0.20-0.30 nM. The assays were carried out at a protein concentration
of 1.5-6 μg/assay tube, with observed KD values of 0.085, 0.095, 0.095
and 0.28 nM and corresponding Bmax values of 610, 1500, 4000, and
Optimization experiments for the radiosynthesis of [18F]4
2100 fmol/mg protein for the D2L, D2S, D3 and D4.4 receptors,
respectively. Competition binding experiments with the human D1, D5
and 5-HT2A receptors were performed in an analogous manner.
tert-Butyl2-(4-[18F]fluorophenyl)azocarboxylate([18F]1)wassynthesized,
Membranes from HEK 293 cells transiently transfected with these target and purified using a C18-cartridge (SepPak tC18, Waters) as described
receptors were used at a protein concentration of 3-6 μg/mL, together
with [3H]SCH23990 (specific activity 60 Ci/mmol, Biotrend, Cologne,
Germany) (final concentration 0.47 nM, KD = 0.49 nM or 0.70 nM,
Bmax = 3000 fmol/mg or 1300 fmol/μg for D1 or D5, respectively) or
[3H]ketanserin (specific activity of 53 Ci/mmol, PerkinElmer, Rodgau,
Germany) (final concentration 0.24 nM, KD = 0.19 nM, Bmax = 700 fmol/mg),
before.15 The product of reaction step one, was eluted from the cartridge
with2 mL of ethanol. To start the second reaction step aliquots (500 μL) of
thiseluatewereaddedtoreactionvialscontainingcommerciallyavailable
4-amino-1-(4-(2-methoxyphenyl)piperazine-1-yl)butane and a base (K2CO3,
Li2CO3 or Cs2CO3) in ethanol (500 μL). The exact reaction conditions are listed
in Table 2. Aliquots (20 μL) from the reaction mixture were drawn after 5, 10,
respectively. Binding studies with porcine 5-HT1A and α1 receptors were and 20 min, diluted in acetonitrile 1:500, and the RCY was determined by
carried out as described.23 Homogenates from porcine cerebral cortex
radio-TLC (Instant Imager, Perkin-Elmer) (Rf: MeOH/CH2Cl2 1/9 = 0.4). All
were prepared, and assays were run with membranes at a protein experiments were conducted as duplicates (n = 2).
Scheme 2. Synthesis of compounds [18F]4 and [18F]3. Reagents and conditions: (a) [18F]fluoride, K222, K2CO3, acetonitrile, 85°C, 30 s15; (b) 4-amino-1-(4-(2-
methoxyphenyl)piperazine-1-yl)butane, K2CO3, ethanol, 35°C, 10 min; (c) 2, Cs2CO3, ethanol, room temperature, 5 min.
www.jlcr.org
Copyright © 2015 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2016, 59 48–53

N. Nebel et al.
Table 2. Results and conditions of the optimization experiments for the radiosynthesis of [18F]4 and [18F]3.1 As a side product
[18F]fluorobenzene occurs. Results are given as mean values with standard deviation (n = 2).
Entry
1
c (Amine2)
c (Base)
Temp.
35°C
Time (min)
RCY [18F]4 (%)
RCY [18F]fluorobenzene (%)
76 mM
217 mM Li2CO3
5
0
0
10
20
5
10
20
5
10
20
5
10
20
5
10
20
5
10
20
5
10
20
5
10
20
2
3
4
5
6
7
8
76 mM
57 mM
38 mM
0 mM
108 mM Li2CO3
217 mM K2CO3
108 mM K2CO3
54 mM K2CO3
23 mM Cs2CO3
23 mM Cs2CO3
23 mM Cs2CO3
35°C
35°C
35°C
35 °C
35°C
35°C
r.t.3
0
0
62 11
5
5
2
6
4
4
10
12
22
13
17
19
4
7
6
1
2
4
3
2
2
1
2
2
1
70
74
8
24 13
42
0
2
6
6
1
0 mM
0
57 mM
57 mM
63
72
72
21 18
58 25
9
4
3
3
6
3
3
7
12
11
6
2
76
8
13
1
Entry
9
c (2)
c (Base)
Temp.
r.t.
Time (min)
RCY [18F]3 (%)
RCY [18F]fluorobenzene (%)
57 mM
23 mM Cs2CO3
5
10
20
60 11
58 17
18
3
3
5
16
13
65
9
1solvent in all cases was EtOH (500 μL) and the reaction was started after addition of an aliquot (500 μL) of [18F]1 in EtOH.
24-amino-1-(4-(2-methoxyphenyl)piperazine-1-yl)butane
3r.t., room temperature
Radiosynthesis of (E)-2-(4-[18F]fluorophenyl)-N-(3-hydroxy-4- Determination of experimental LogD (pH 7.4)
(4-(2,3-dichlorophenyl) piperazine-1-yl)butyl)diazene-1-
An aliquot of 10 μL of [18F]3 (ca. 25 kBq) in solution (PBS) was added to
carboxamide ([18F]3) using optimized conditions
PBS (500 μL) and n-octanol (500 μL). After vortexing for 1 min, the
mixture was centrifuged. Three aliquots of 100 μL were taken from each
phase and analyzed in a γ-counter. The experiment was conducted in
triplicates (n = 3).
Using the optimized conditions, [18F]fluoride was eluted from the QMA-
cartrigde with a solution of Kryptofix 2.2.2 (15 mg, 39.8 μmol) and K2CO3
(1 M, 15 μL) in acetonitrile (900 μL). Water was removed by evaporation
of acetonitrile (3 × 500 μL) under a stream of nitrogen at 85°C. The
precursor 2-(4-(N,N,N-trimethylammonium)phenyl)azocarboxylic acid
tert-butylester triflate (2.5 mg, 6.0 μmol) in acetonitrile (400 μL) was added
to the dried K+/Kryptofix 2.2.2/18F complex, and the solution was stirred
for 30 s at 85°C. Afterwards, the yellow solution was diluted with HCl
(0.2 M, 15 mL) and passed through a cartridge (SepPak tC18, Waters).
The cartridge was washed with acetonitrile/0.2 M HCl (20:80, 5 mL) and
water (2 mL). Tert-butyl 2-(4-[18F]fluorophenyl)azocarboxylate ([18F]1)
was then eluted with ethanol (1 mL) from the cartridge into a second
reactor containing 2 (15 mg, 54 μmol) and Cs2CO3 (7.5 mg, 23 μmol). After
the solution was stirred for 5 min at room temperature, the solvent was
removed and acetonitrile/water (0.1% TFA) (1:1, 500 μL) added. The
product was purified by semipreparative HPLC (method 3: tR = 8.5 min)
to afford [18F]3 in an overall yield of 8-12% (non-decay corrected, n = 4)
and specific activities of 32-102 GBq/μmol. The identity of the radioligand
was determined by coinjection with its reference compound using HPLC
methods 1 (tR = 3.6 min) and 2 (tR = 9.12).
Stability in human serum
An aliquot of [18F]3 (5-10 MBq) was added to human serum (400 μL) and
incubated at 37°C. Aliquots (25 μL) were taken after 5, 10, 15, 30, 45, 60,
and 120 min and quenched in acetonitrile/H2O (10% TFA) (1:1, 100 μL) or
in MeOH/CH2Cl2 (1:1, 100 μL). The samples were centrifuged and the
supernatants analyzed by radio HPLC (method 1) or TLC (CH2Cl2 /EtOH
1:9, Rf: 0.4).
Results and discussion
Chemistry
As depicted in Scheme 1, amine 2 was reacted with tert-butyl
2-(4-fluorophenyl)azocarboxylate 1 in the presence of K2CO3 in
a nucleophilic substitution reaction to afford target compound
J. Label Compd. Radiopharm 2016, 59 48–53
Copyright © 2015 John Wiley & Sons, Ltd.
www.jlcr.org

N. Nebel et al.
3. The yield of this reaction was relatively low (14%) in 23 mM (83% RCY, Figure 2) after a reaction time of 5 min.
comparison with the synthesis of a series of structural analogs, Contrary to the increase in RCY while reducing the concentration
including 4 (24%) as reported recently.15,17 However, the of K2CO3, the RCY of [18F]4 using Cs2CO3 was unchanged in the
phenylazo carboxamide 3 was obtained in sufficient purity concentration range of 23–217 mM (Figure 2). Using Cs2CO3 as
(96%) to proceed with in vitro binding experiments as described base, the reaction proceeded very fast, reaching its maximum
in the following.
RCY already after 5 min, whereas with K2CO3 as base, the
maximum RCY was achieved after 20 min (Figure 2).
The byproduct, which forms during the reaction, was identified as
[18F]fluorobenzene through coinjection of commercially available
fluorobenzene using HPLC method 1 (tR (18F]4)=2.7 min, tR ([18F]
fluorobenzene = 3.2 min). The formation of [18F]fluorobenzene
occured with both alkali-bases and was higher using Cs2CO3 than
using K2CO3 (Figure 2, dashed lines). Figure 2 shows that the
concentration of alkali-base does not influence the formation of
[18F]fluorobenzene when using Cs2CO3, and it is also formed
without the presence of amine as reactant (Table 2, entry 5 and
6). The concentration of amine could be reduced from 76 to
57 mM with consistently high RCY of about 70% for both K2CO3
and Cs2CO3 (Table 2, entry 3 and 7).
A further reduction of amine to 38 mM lead to a significant loss
of RCY to 42% after 20 min (Table 2, entry 4). Finally, the influence
of the reaction temperature was studied. The RCY was high
(72-76%) after 20 min at room temperature and at 35°C when
using Cs2CO3 at a concentration of 23 mM, proceeding faster at
35°C (Table 2, entry 7 and 8).
Biological evaluation
Receptor binding studies investigated the affinity of the
compound 3 to subtypes of the dopamine receptors and to
the related biogenic amine receptors 5-HT1A, 5-HT2, and α1
(Table 1). The binding affinity of 3 toward the D3 receptor (Ki
(D3) = 3.6 nM) was in the same nanomolar range as for the
reference compound 4 (Ki (D3) = 3.5 nM). In comparison with 4,
compound 3 showed an increased D3 selectivity over the other
dopamine receptors subtypes by a factor of 5 with Ki (D2/D3)
= 53. The least selectivity was observed between the 5-HT2 and
D3 receptor (Ki (5-HT2/D3) = 16; Table 1). The selectivity for the
D3 receptor over the serotonergic receptor 5-HT1A, and α1 was
improved compared with reference compound 4 (Ki (5-HT1A
D3) = 34 and Ki (α1/D3) = 100).
/
Radiochemistry
Optimization experiments for the radiosynthesis of [18F]4
To sum up, the optimized reaction conditions were
The first step of the radiosynthesis of [18F]4 depicted in Scheme 2, determined to be Cs2CO3 as base in a concentration of 23 mM
dealing with the 18F-fluorination of the trimethylammonium and the amine in a concentration of 57 mM at room temperature
triflate precursor, exactly followed the procedure described or 35°C. These reaction conditions were applied in the
previously.15 The second step of the radiosynthesis leading to radiosynthesis of compound [18F]3.
the phenylazacarboxamide [18F]4 was optimized and tested
primarily regarding threshold and choice of the alkali-base. The
Radiosynthesis of (E)-2-(4-[18F]fluorophenyl)-N-(3-hydroxy-4-
concentration of the amine (4-amino-1-(4-(2-methoxyphenyl)
piperazine-1-yl)butane), which can attack the carboxyl of the
(4-(2,3-dichlorophenyl) piperazine-1-yl)butyl)diazene-1-
[18F]fluorophenylazocarboxylic ester [18F]1 under basic conditions
carboxamide ([18F]3) using optimized conditions
in a nucleophilic substitution reaction, was lowered along with
readjustment of the reaction temperature (Figure 2 and Table 2). The radiosynthesis of [18F]3 (Scheme 2) was realized adopting
The base K2CO3 is of low solubility in ethanol and a high the optimized reaction conditions, that is, 23 mM Cs2CO3 and
concentration of 217 mM thereof was chosen for the initial 57 mM amine 2 at room temperature (Table 2, entry 9).
reaction (Figure 2). Reduction of the concentration down to Compared with the radiosynthesis of [18F]4, the formation of
54 mM was possible and even offered a 15% higher RCY of
[
18F]3 was faster, leading to a RCY of 60 11% after 5 min. After
[
18F]4. Further reduction to a concentration of 25 mM resulted in a total synthesis time of 30–35 min starting from [18F]fluoride,
significant loss of RCY (Figure 2).
[
18F]3 was obtained after HPLC separation in an overall yield of
Trying the two alternative alkali bases Li2CO3 and Cs2CO3 8-12% (non-decay corrected, n = 4) and a specific activity of
underlined that the concentration of base can be based on its 32–102 GBq/μmol (n = 4).
solubility alongside other factors as ion-size and ion-pairing.26
In first in vitro experiments, [18F]3 was stable in human serum
The in-ethanol and water-insoluble Li2CO3 led to negligible RCY in vitro at 37°C. Monitoring with radio-HPLC and radio-TLC
(Table 2, entry 1 and 2), whereas the in-ethanol-soluble Cs2CO3 showed no decomposition over 2 h. The logD7.4 was determined
led to similar RCY as with K2CO3 even at a low concentration of to be 2.56. These results are promising for high stability in vivo
Figure 2. Dependence of the RCY on choice and concentration of the alkali base. The solid lines represent the product [18F]4 and the dashed lines represent the side
product [18F]fluorobenzene (n = 2). Conditions: 4-Amino-1-(4-(2-methoxyphenyl)piperazine-1-yl)butane (76 mM), EtOH (1 mL), 35°C.
www.jlcr.org
Copyright © 2015 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2016, 59 48–53

N. Nebel et al.
[4] F. Böckler, P. Gmeiner, Biochim. Biophys. Acta 2007, 1768, 871–887.
[5] J. N. Joyce, M. J. Millan, Drug Discov. Today 2005, 10, 917–925.
[6] K. A. Neve, J. K. Seamans, H. Trantham-Davidson, J. Recept. Signal
Transduct. Res. 2004, 24, 165–205.
[7] L. R. Riddle, R. Kumar, S. A. Griffin, P. Grundt, A. H. Newman, R. R.
Luedtke, Neuropharmacology 2011, 60, 284–294.
[8] D. Erritzoe, A. Tziortzi, D. Bargiela, A. Colasanti, G. E. Searle, R. N.
Gunn, J. D. Beaver, A. Waldman, D. J. Nutt, M. Bani, E. Merlo-Pich, E.
A. Rabiner, A. Lingford-Hughes, Neuropsychopharmacology 2014,
39, 1703–1712.
[9] S. R. Cote, V. C. Chitravanshi, C. Bleickardt, H. N. Sapru, E. V.
Kuzhikandathil, Behav. Brain Res. 2014, 263C, 46–50.
and for penetrating the BBB, for which logP values between 2
and 3.5 are considered optimal.21
Conclusion
The radiosynthesis of compound [18F]4 was optimized resulting in
a strategy using the alkali-base Cs2CO3, which is soluble in ethanol
and effective in a low concentration of 23 mM, in contrast to
217 mM using K2CO3, and accelerates the reaction time to 5 min.
The concentration of the reacting amine was also reduced by a
fourth and temperature simply set to room temperature. This [10] J. L. Neisewander, T. H. Cheung, N. S. Pentkowski,
Neuropharmacology 2014, 76 B, 301–319.
optimized reaction was applied in the radiosynthesis of the
interesting new 2,3-dichlorophenylpiperazine [18F]3, achieving an
[11] E. D. Morris, M. V. Lucas, J. R. Petrulli, K. P. Cosgrove, Yale J. Biol. Med.
2014, 87, 33–54.
overall non-decay corrected yield of 8-12% and a specific activity
[12] G. Muehllehner, J. S. Karp, Phys. Med. Biol. 2006, 51, R117–137.
of 32-102 GBq/μmol after a total synthesis time of 30-35 min. The
hydroxylated phenylazacarboxamide [18F]3 showed nanomolar
affinity (Ki (D3) = 3.6 nM) and a selectivity of >16-fold for the
dopamine D3 receptor, providing a D3 radioligand candidate with
improved attributes concerning selectivity and a reliable and easy
radiosynthesis to facilitate further preclinical studies.
[13] O. Prante, M. Dörfler, P. Gmeiner, in The Dopamine Receptors (Ed.: K.
A. Neve), Humana Press, 2009, 101–135.
[14] M. Slifstein, E. A. Rabiner, R. N. Gunn, in Imaging of the Human Brain
in Health and Disease (Eds: P. Seeman, B. Madras), Academic Press,
2014, 265–281.
[15] S. K. Fehler, S. Maschauer, S. B. Höfling, A. L. Bartuschat, N.
Tschammer, H. Hübner, P. Gmeiner, O. Prante, M. R. Heinrich, Chem.
Eur. J. 2014, 20, 370–375.
[16] H. Jasch, S. B. Hoefling, M. R. Heinrich, J. Org. Chem. 2012, 77, 1520–1532.
[17] N. Nebel, S. Maschauer, A. L. Bartuschat, S. K. Fehler, H. Hübner, P.
Gmeiner, T. Kuwert, M. R. Heinrich, O. Prante, C. Hocke, Bioorg.
Med. Chem. Lett. 2014, 24, 5399–5403.
[18] C. Hocke, O. Prante, S. Löber, H. Hübner, P. Gmeiner, T. Kuwert,
Bioorg. Med. Chem. Lett. 2005, 15, 4819–4823.
[19] P. Grundt, K. M. Prevatt, J. Cao, M. Taylor, C. Z. Floresca, J. K. Choi, B.
G. Jenkins, R. R. Luedtke, A. H. Newman, J. Med. Chem. 2007, 50,
4135–4146.
[20] A. H. Newman, P. Grundt, G. Cyriac, J. R. Deschamps, M. Taylor, R.
Kumar, D. Ho, R. R. Luedtke, J. Med. Chem. 2009, 52, 2559–2570.
[21] S. M. Ametamey, M. Honer, P. A. Schubiger, Chem. Rev. 2008, 108,
1501–1516.
Acknowledgements
This work was supported by the Deutsche Forschungsgemeinschaft
(DFG; PR 677/6-3). This study is part of the Neurotrition Project,
which is supported by the Emerging Fields Initiative of the
Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU).
Conflict of interests
The authors report no conflict of interest.
[22] C. Hocke, P. Cumming, S. Maschauer, T. Kuwert, P. Gmeiner, O.
Prante, Nucl. Med. Biol. 2014, 41, 223–228.
References
[23] H. Hübner, C. Haubmann, W. Utz, P. Gmeiner, J. Med. Chem. 2000, 43,
756–762.
[24] O. H. Lowry, N. J. Rosebrough, A. L. Farr, R. J. Randall, J. Biol. Chem.
1951, 193, 265–275.
[25] Y. Cheng, W. H. Prusoff, Biochem. Pharmacol. 1973, 22, 3099–3108.
[26] C. Galli, Org. Prep. Proced. Int. 1992, 24, 285–307.
[1] M. Jaber, S. W. Robinson, C. Missale, M. G. Caron, Neuropharmacology
1996, 35, 1503–1519.
[2] D. Payer, G. Balasubramaniam, I. Boileau, Prog. Neuropsychopharmacol.
Biol. Psychiatry 2014, 52, 4–8.
[3] P. Sokoloff, B. Giros, M. P. Martres, M. L. Bouthenet, J. C. Schwartz,
Nature 1990, 347, 146–151.
J. Label Compd. Radiopharm 2016, 59 48–53
Copyright © 2015 John Wiley & Sons, Ltd.
www.jlcr.org