Two-step radiosynthesis of [18F]FE-β-CIT and [ 18F]PR04.MZ

Article abstract of DOI:10.1002/jlcr.3032

The cocaine-derived dopamine reuptake inhibitors FE-β-CIT (8-(2-fluoroethyl)-3-(4-iodophenyl)-8-azabicyclo[3.2.1]octane-2-carboxylic acid methyl ester) (1) and PR04.MZ(8-(4-fluorobut-2-ynyl)-3-p-tolyl-8-azabicyclo[3.2. 1]octane-2-carboxylic acid methyl es

Full text of DOI:10.1002/jlcr.3032

Note
Received 29 September 2012,
Revised 29 November 2012,
Accepted 14 January 2013
Published online 1 April 2013 in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/jlcr.3032
Two-step radiosynthesis of [18F]FE-b-CIT and
[18F]PR04.MZ
a
a
a
Patrick J. Riss,a,b Sabine Hoehnemann, Markus Piel, and Frank Roesch
*
The cocaine-derived dopamine reuptake inhibitors FE-b-CIT (8-(2-fluoroethyl)-3-(4-iodophenyl)-8-azabicyclo[3.2.1]octane-2-
carboxylic acid methyl ester) (1) and PR04.MZ(8-(4-fluorobut-2-ynyl)-3-p-tolyl-8-azabicyclo[3.2.1]octane-2-carboxylic acid
methyl ester) (2) were labelled with 18F-fluorine using a two-step route. 2-[18F]Fluoroethyltosylate and 4-[18F]fluorobut-2-
yne-1-yl tosylate were used as labelling reagents, respectively. Radiochemically pure (>98%) [18F]FE-b-CIT and [18F]
PRD04.MZ (32–86 GBq/mmol) were obtained after a synthesis time of 100 min in about 25% non-decay-corrected
overall yield.
Keywords: dopamine transporter; fluorine-18; automated synthesis; tropanes
processes. Moreover, automated synthesis modules have become
available that allow for automation of complex processes including
Introduction
Dopamine transporter (DAT) imaging agents are highly sought
two subsequent HPLC purifications and current good manufactur-
after for studies of the availability of this presynaptic binding site
ing practise (cGMP) compliant productions involving the synthesis
in preclinical research and clinical application.1–4
of 18F-labelled building blocks as intermediates. To the best of our
knowledge, 18F-labelling of [18F]FE-b-CIT ([18F]1) has not been
described yet. This is unfortunate because despite providing useful
characteristics such as the opportunity of introducing an 123I-label,
and an 11C-label, only 18F will provide a sufficiently long half-life
that correlates with the kinetic profile of the ligand to facilitate
kinetic modelling in quantitative positron emission tomography
studies. Initial direct aliphatic nucleophilic radiofluorination of
[18F]PRD04.MZ ([18F]2) was ]performed under conventional condi-
tions. Unfortunately, only low radiochemical yields of 13 Æ 3% were
obtained after the reaction times of 60–70 min. For these reasons
and given the scope of an 18F-labelled DAT ligand, we have
elaborated a suitable two-step method for the radiosynthesis of
the 18F-labelled 3-phenyltropanes[18F]FE-b-CIT ([18F]1), which has
not been described yet, and [18F]PR04.MZ ([18F]2).20
Two viable strategies for the synthesis of 18F-labelled radioli-
gands for the presynaptic DAT, direct substitution and 18F-fluoro
alkylation, have been reported.5–11
In recent years, a variety of 18F-labelled 3-phenyltropane analo-
gues have been reported, and the available labelling strategies were
thoroughly elaborated. Although direct labelling superficially
appears to be the more elegant route, it suffers from low radioche-
mical yields, and most approaches to overcome this issue do not
translate well into the available automated synthesis modules, for
example, due to the use of microwave reactors.12,13 This is a major
issue for clinical application, because routine production of sufficient
doses requires starting radioactivities well above the amounts
considered to be safe for manual handling.1–8,12,13
Radiolabelling of cocaine-analogue radioligands with the
positron emitter 18F(t1/2 = 109.7 min) via nucleophilic fluorination
was often complicated by (i) low N-alkylation yields, (ii) poor
accessibility of sp2-carbon-bound fluorine, resulting in laborious
reaction pathways and (iii) a/b-epimerisation at the C-2 position
in the tropane skeleton.8,12,14–17
Experimental
Automated synthesis of 18F-fluorinated labelling agents ([18F]5–6)
The use of 18F-fluoroalkylating agents has repeatedly been
described as a reliable route to furnish the desired radioligands for
DAT imaging.5–8,12 Until recently, automated direct nucleophilic
radiofluorination was less popular for this purpose.10,11 Despite
being less straightforward than direct nucleophilic substitution, the
18F-fluoroalkylation route has some exceptional advantages for
radiolabelling of cocaine derivatives. The reaction proceeds under
milder, less basic conditions, labelling precursors are more readily
accessible and the labelled product is often easier to separate from
the labelling precursor by preparative HPLC.5–8,16,18,19
All operations were performed in a lead-shielded cell, containing an in-house
designed homemade synthesis module. He 5.0 gas was used for all opera-
tions at a pressure of 1.4 Â 105 Pa. [18F]fluoride in H128O was extracted from
aInstitut für Kernchemie, Universität Mainz, Fritz Strassmann Weg 2, 55128
Mainz, Germany
bDepartment of Chemistry, UiO Postboks 1033 - Blindern, 0315 Oslo, Norge
In addition, 2-[18F]fluoroethyl tosylate has become commercially
available, and satellite distribution of other 18F-fluoroalkylating
agents can be anticipated, thus enabling one-step labelling
*Correspondence to: Patrick J. Riss, Department of Chemistry, UiO, Postboks
1033 - Blindern, 0315 Oslo, Norge
E-mail: patrick.riss@kjemi.uio.no
J. Label Compd. Radiopharm 2013, 56 356–359
Copyright © 2013 John Wiley & Sons, Ltd.

P. J. Riss et al.
the target water using a Waters (GmbH, Eschborn, Germany) light QMA
strong anion exchange cartridge. Elution from the resin was performed using
Results and discussion
a solution of KryptofixW (Merck KGaA, Darmstadt, Germany) K222 (15 mg, Chemistry
0.4 mmol) and potassium carbonate (2.1 mg, 15 mmol) in dry MeCN (1 ml).
Reference compounds 1–2 and the corresponding nortropane
congeners used as labelling precursors 3–4 were synthesized
The eluate was concentrated to dryness in 15 min, using a five-step drying
procedure. The labelling precursor (15 mg, 35–40 mmol), dissolved in MeCN
(1ml), was added subsequently, and the mixture was stirred at 88ꢀC for from natural cocaine as published elsewhere.20–22,25 1,2-Bis-
3 min. The reaction mixture was diluted with 50% water in MeCN (1ml)
and transferred into a 2-ml inject loop of a Knaur (Wissenschaftliche
Geraetebau Dr. Ing. Herbert Knauer GmbH, Berlin, Germany) semi-
preparative HPLC system, equipped with a Lichrospher RP18 (Merck KGaA,
Darmstadt, Germany) 5 mm column. A mixture of MeCN and water (55/45)
was used as mobile phase at a flow rate of 4.7 ml/min. The product fraction
was collected, diluted with water (30 ml) and passed through a Merck
Lichrospher EN (Merck KGaA, Darmstadt, Germany) cartridge. After drying
of the cartridge in a stream of He, the product was eluted in dimethyl
sulfoxide (DMSO) (1 ml) warmed to 65ꢀC, directly into the preheated
second reaction vessel, containing precursor. 2-[18F]Fluoroethyl tosylate
([18F]5) was obtained in a decay-corrected yield of 60 Æ 8% after 55 min;
4-[18F]Fluorobut-2-yne-1-yl tosylate ([18F]6) was obtained in a radio-
chemical yield of 64 Æ 4%.
(4-toluenesulfonyloxy)ethane (7) as a starting material for the synthe-
sis of [18F]5 is commercially available. The starting material required
for the radiosynthesis of [18F]6, 1,4-bis-(4-toluenesulfonyloxy)but-2-
yne (8), was synthesized in a yield of 22% from commercially available
but-2-yne-1,4-diol according to a phase-transfer catalytic procedure
described for 1,4-bis-(4-toluenesulfonyloxy)but-2-ene.7
Radiochemistry
The synthesis of both radiotracers was performed in a two-step
procedure, by first synthesizing the prosthetic group [18F]5 or
[18F]6 followed by the 18F-fluoroalkylation of the respective
precursor 3 or 4 (Scheme 1).
Preparation of the radiofluorinated alkylating agents was
conducted using a homemade synthesis module (Figure 1). The
process was based on a report of Block et al.21,24 and modified
to our needs. Starting from aqueous [18F]fluoride, potassium car-
bonate and KryptofixW K 222, removal of volatiles by azeotropic
co-evaporation with MeCN afforded the potassium cryptate
Screening of reaction conditions
Screening reactions21 were conducted in conical bottom-brown glass
Wheaton (Wheaton, Millville, NJ, USA) reaction vessels in triplicate using
2mg of nortropanes 3 and 4 in DMSO (1ml) at 80–120ꢀC. Lithium iodide
(5mg, 37 mmol) was added to the vials at 80ꢀC but not at higher tempera-
tures because of the volatility of 2-[18F]fluoroethyl iodide. Samples (100 ml) complex [K+ ⊂ K222]18FÀ in a sequence taking roughly 15 min.
were withdrawn after 3, 7, 10 and 20min, quenched in water26 (400 ml)
A solution of the appropriate ditosylate in acetonitrile was added
to the residue and heated to reflux for 3 min. The reaction mix-
and analyzed by HPLC.
ture was diluted with HPLC eluent, and the product was purified
Exemplified radiosynthesis of [18F]2
using semi-preparative reversed phase HPLC.
The product fraction was collected and diluted with H2O, and
the product was separated by solid-phase extraction on a MerckW
After the addition of 6 to the reaction vessel, the reaction mixture was
stirred at 120ꢀC for 15 min; after which the reaction was quenched by the
Lichrolut EN cartridge. This cartridge was dried for 15 min to
addition of the eluent (4ml) to the reaction vessel. The resultant mixture
remove excess moisture from the trapped reagent. This process
was drawn into a syringe containing water (5ml) and passed through a
afforded the labelling reagents [18F]5 and [18F]6 in radiochemical
Merck strong cation exchange (SCX) cartridge to remove the DMSO
yields of about >60% after a total synthesis time of 55 min
prior to HPLC purification. The trapped radioactivity was eluted with dilute
(Table 1). In our experience, the use of HPLC-purified [18F]5 and
ammonia solution 1M (1ml) and purified on
a semi-preparative
[18F]6 results in higher yields in the subsequent labelling reactions
and reduced formation of non-radioactive by-products that might
complicate the purification. We thus resorted to two subsequent
HPLC purifications rather than only one after the second step.
The trapped radioactive reagent was eluted into a Wheaton reac-
tion vessel, and aliquots of about 50 MBq of [18F]5 and [18F]6 were
used for the screening of reaction conditions. [18F]1 was synthesized
by the reaction of 3 and [18F]5 in DMSO, and the time-dependent
labelling yield was investigated as a function of the reaction
temperature in the absence and in the presence of LiI. To avoid
light-induced de-iodination, a brown-glass vessel was used for
Phenomenex Luna (Phenomenex, Aschaffenburg, Germany) RP18(2)
10 mm 250 Â 10 mm HPLC column, using 40% 0.1-M ammonium
acetate/60% acetonitrile at pH 4.7 as mobile phase. The product fraction
was collected after a total retention time of 13–15 min, diluted with 10 ml
of water and passed through a Merck Lichrolut SCX cartridge (200 mg),
conditioned with 5 ml of 1-M hydrochloric acid and 20ml of water.
Subsequently, the cartridge was washed with 3ml of water and the
product eluted in a crimp-sealed vial through a sterile filter using
phosphate-buffered saline (PBS) (1 ml). Average non-decay-corrected
yield was 25% for [18F]1 and [18F]2, respectively.
Analytical HPLC of [18F]1 and [18F]2
The radiochemical purity and the specific radioactivity of [18F]1 and [18F]
2 were determined on a Sykam HPLC system (Sykam GmbH, Eresing,
Germany), equipped with a Berthold LB501 (Berthold Technologies
GmbH & Co. KG, Bad Wildbad Germany) radioactivity detector and a
Knauer (Knauer GmbH, Berlin, Germany) ultraviolet (UV) detector. A
Phenomenex Luna RP18 5 mm 250 Â 4.6 mm analytical HPLC column
was used as stationary phase. For [18F]1, the mobile phase consisted of
50% 0.01-M phosphoric acid in MeCN; for [18F]2, 30% 0.05-M ammonium
acetate buffer at pH 4.6 in MeCN was used. An aliquot of 20 ml was
withdrawn from the formulated tracer solution and diluted with eluent.
Approximately 50 kBq of this solution was injected into the HPLC.
Products [18F]1 and [18F]2 were eluted after 8–10 and 10–11 min, respec-
tively. Specific activities ranged from 32 to 86 GBq/mmol.
Scheme 1. Radiosynthesis of [18F]1 and [18F]2. a: [K+ ⊂ K222]18FÀ, MeCN, 88ꢀC,
3 min; b: 3, [18F]5, dimethyl sulfoxide (DMSO), 120ꢀC, 20 min; and c: 4, [18F]6, DMSO,
120ꢀC, 15 min.
J. Label Compd. Radiopharm 2013, 56 356–359
Copyright © 2013 John Wiley & Sons, Ltd.
www.jlcr.org

P. J. Riss et al.
Figure 1. Scheme of the synthesis module. SPE, solid-phase extraction; UV, ultraviolet
Table 1. Labelling conditions, retention times and radiochemical yields
Conditions
tR (min)
tR(QC)/min
RCY (%)
[18F]1
[18F]2
[18F]5
[18F]6
DMSO, 20 min, 120ꢀC
DMSO, 15 min, 120ꢀC
MeCN, 3 min, 88 ꢀC
11–12
13–15
8–10
8
10
6
64 Æ 5
67 Æ 5.5
60 Æ 8
64 Æ 4
10–11
8
DMSO, dimethyl sulfoxide; QC, quality control; RCY, radiochemical yield.
[18F]1. The results are depicted in Figure 2. In an analogue fashion,
[18F]2 was synthesized from 4 and [18F]6. A radiochemical yield of
about 64–67% was achieved after a reaction time of 15–20 min at
120ꢀC using DMSO as solvent and 2 mg of the precursor. Following
the 18F-fluoroalkylation reaction, both products, [18F]1 and [18F]2,
were purified using semi-preparative HPLC.
[18F]1 was eluted after a reasonable retention time (tR: 11–12 min),
which was necessary to separate it from a polar non-significant by-
product, which was hypothesized to correspond to either the
saponified acid or the 2a-epimer. For the concentration and isolation
of the radiotracer, the product fraction was diluted with water and
trapped on a C18 cartridge. The cartridge was washed with water,
and the product was eluted with EtOH, followed by 0.9% NaCl
solution. This procedure gave [18F]1 as a colourless, clear solution
Figure 2. Radiochemical yields (RCYs) of [18F]1 as a function of reaction tempera-
ture and time using 2 mg of the precursor 3. DMSO, dimethyl sulfoxide.
of neutral pH with a radiochemical purity >98%. No significant UV
impurities were detected.
www.jlcr.org
Copyright © 2013 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2013, 56 356–359

P. J. Riss et al.
The HPLC purification of [18F]2 yielded the product after a
retention time of 13–15 min. [18F]2 was isolated and formulated
using a proton-loaded Merck Lichrolut SCX resin. About 95% of
the total radioactivity in the diluted product fraction was trapped
on the SCX resin. Subsequently, the cartridge was washed with
water, and the product was eluted through a sterile filter into a
crimp-sealed vial using isotonic PBS. Less than 5% of the total
trapped radioactivity remained on the SCX cartridge. The for-
mulation was a colourless, clear solution with a pH value of 7.3.
Quality control by analytical HPLC showed a radiochemical purity
>98%. No significant UV impurities were detected. In both cases,
radio-HPLC conditions were validated by radio-TLC to ensure the
absence of co-eluting impurities in the product fraction. The formu-
lations of [18F]1 and [18F]2 prepared in this procedure contained
either one of both tracers in specific activities of 32–86 GBq/mmol
after 100 min starting from [18F]fluoride. In comparison with direct
nucleophilic radiofluorination under conventional conditions (13%),
formulated [18F]2 in sterile PBS was obtained in a two-fold higher
yield (25%), despite using the slightly more elaborate two-step
procedure described herein.20
[2] L. Müller, C. Halldin, L. Farde, P. Karlsson, H. Hall, C. G. Swahn, J.
Neumeyer, Y. Gao, R. Milius, Nucl. Med. Biol. 1993, 20, 249.
[3] P. H. Elsinga, K. Hatano, K. Ishiwata, Curr. Med. Chem. 2006, 13, 2139.
[4] J. S. Stehouwer, M. M. Goodman, PET Clin. 2009, 4, 101.
[5] M. M. Goodman, C. D. Kilts, R. Keil, B. Shi, L. Martarello, D. Xing,
J. Votaw, T. D. Ely, P. Lambert, M. J. Owens, V. M. Camp, E. Malveaux,
J. M. Hoffman, Nucl. Med. Biol. 2000, 27, 1.
[6] C. Lundquist, C. Halldin, N. Ginovart C. G. Swahn, L. Farde, Nucl. Med.
Biol. 1997, 24, 621.
[7] F. Dollé, F. Hinnen, P. Emond, S. Mavel, Z. Mincheva, W. Saba, M.-A.
Schöllhorn-Peyronneau, H. Valette, L. Garreau, S. Chalon, C. Halldin,
J. Helfenbein, J. Legaillard, J.-C. Madelmont, J.-B. Deloye, M. Bottlaender
and D. Guilloteau, J Labelled Compd. Rad. 2006, 49, 687.
[8] R. P. Klok, P. J. Klein, J. D. M. Herscheid, A. D.Windhorst, J. Label.
Compd. Radiopharm. 2006, 49, 77.
[9] F. Dollé, J. Helfenbein, F. Hinnen, S. Mavel, Z. Mincheva, W. Saba,
M.-A. Schöllhorn-Peyronneau, H. Valette, L. Garreau, S. Chalon,
C. Halldin, J.-C. Madelmont, J.-B. Deloye, M. Bottlaender, J. Le Gailliard,
D. Guilloteau, P. Emond, J. Label. Compd. Radiopharm. 2007, 50, 716.
[10] B. Kuhnast, F. Hinnen, P. Emond, J. Le Gailliard, J. B. Deloye, D.
Guilloteau, F. Dollé J Label Compd. Rad. 2009, 52(Suppl. 1), S279.
[11] S. J. Lee, S. J. Oh, W. Y. Moon, M. S. Choi, J. S. Kim, D. Y. Chi, D. H.
Moon, J.S. Ryu, Nucl. Med. Biol. 2011,38, 593.
[12] P. J. Riss, F. Roesch, Bioorg. Med. Chem. 2009, 17, 7630.
[13] S.-Y. Lu, F.T. Chin, J.A. McCarron, V.W. Pike, J. Label. Compd. Radio-
pharm. 2004, 47, 289.
[14] M. Haaparanta, J. Bergman, A. Laakso, J. Hietala, O. Solin Synapse
1996, 23, 321.
[15] J. Bergman, O. Solin, Nucl. Med. Biol. 1997, 24, 677.
[16] T. Chaly, V. Dhawan, K. Kazumata, A. Antonini, C. Margouleff,
J. R. Dahl, A. Belakhlef, D. Margouleff, A. Yee, S. Wang, G. Tamagnan,
J. L.Neumeyer, D. Eidelberg, Nucl. Med. Biol. 1996, 23, 999.
[17] D. Xing, P. Chen, R. Keil, C. D. Kilts, B. Shi, V. M. Camp, G Malveaux,
T. Ely, M. J. Owens, J Votaw, M. Davis, J. M. Hoffman, R. A. E. BaKay,
T. Subramanian, R. L. Watts, M. M. Goodman, J. Med. Chem. 2000,
43, 639.
Although [18F]2 has been produced in even higher yields
(32–36%) in a much shorter time (35 min) using a monomodal
laboratory microwave reactor, we feel that ease of automation
using commercially available synthesis modules combined with
conventional heating is probably a more useful route in a radio-
tracer production context.
Conclusions
[18F]1 and [18F]2 have been prepared using a two-step labelling
strategy for the first time. This method is compatible with cGMP-
compliant synthesis modules and does not require exceptional
conditions. The formulated tracers were obtained after a total
duration of about 100 min in a non-decay-corrected yield of 25%.
[18] N. Harada, H. Ohba, D. Fukumoto, T. Kakiuchi, H. Tsukada, Synapse
2004, 54, 37.
[19] T. Chaly, R. M. Baldwin, J. L. Neumeyer, M. J. Hellman, V. Dhawan,
P. K. Garg, G. Tamagnan, J. K. Staley, M.S. Al-Tikriti, Y. Hou, S. S. Zoghbi,
X. H. Gu, R. Zong, D. Eidelberg, Nucl. Med. Biol. 2004, 31, 125.
[20] P. J. Riss, F. Debus, R. Hummerich, U. Schmidt, P.Schloss, H. Lueddens,
F. Roesch, ChemMedChem 2009, 4, 1480.
[21] D. Block, H. H. Coenen, G. Stöcklin, J Labelled Compd. Rad. 1988, 25,
201.
[22] P. C. Meltzer, A. Y. Liang, A. L. Brownell, D. R. Elmaleh, B. K.
Conflict of Interest
The authors did not report any conflict of interest.
Madras, J. Med. Chem. 1993, 36, 855.
[23] L. Xu, M. L. Trudell, J Heterocylic Chem. 1996, 33, 2037.
[24] D. Block, H. H. Coenen, G. Stöcklin, J Labelled Compd. Rad. 1987, 24,
1029–1042.
[25] P. J. Riss, R. Hummerich, P. Schloss, Org. Biomol. Chem. 2009, 7, 2688.
[26] A. Bauman, M. Piel, R. Schirrmacher, F. Roesch, Tetrahedron Lett.
2003, 44, 9165.
References
[1] P. J. Riss, K. Stockhofe, F. Roesch, J. Label, Compd. Radiopharm. 2013, doi:
10.1002/jlcr.3018
J. Label Compd. Radiopharm 2013, 56 356–359
Copyright © 2013 John Wiley & Sons, Ltd.
www.jlcr.org