Synthesis, radiolabeling and preliminary in vivo evaluation of [18F]FE-PE2I, a new probe for the dopamine transporter

Article abstract of DOI:10.1016/j.bmcl.2009.06.032

A new dopamine transporter (DAT) ligand, (E)-N-(3-iodoprop-2-enyl)-2β-carbofluoroethoxy-3β-(4′-methyl-phenyl) nortropane (FE-PE2I, 6), derived from PE2I (1), was prepared and found to be a potent inhibitor of rodent DAT in vitro. Compound 6 was radiolabel

Full text of DOI:10.1016/j.bmcl.2009.06.032

Bioorganic & Medicinal Chemistry Letters 19 (2009) 4843–4845
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis, radiolabeling and preliminary in vivo evaluation of [18F]FE-PE2I,
a new probe for the dopamine transporter
a
Magnus Schou a, , Carsten Steiger , Andrea Varrone a,b, Denis Guilloteau c, Christer Halldin a,b
*
a Karolinska Institutet, Department of Clinical Neuroscience, Psychiatry Section, Stockholm, Sweden
b Stockholm Brain Institute, Stockholm, Sweden
c Inserm U619, Université François Rabelais de Tours, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 27 November 2008
Revised 8 June 2009
Accepted 10 June 2009
Available online 13 June 2009
A
new dopamine transporter (DAT) ligand, (E)-N-(3-iodoprop-2-enyl)-2b-carbofluoroethoxy-3b-(40-
methyl-phenyl) nortropane (FE-PE2I, 6), derived from PE2I (1), was prepared and found to be a potent
inhibitor of rodent DAT in vitro. Compound 6 was radiolabelled with fluorine-18 (t1/2 = 109.8 min) for
PET studies in monkeys. In vivo PET measurements showed a regional distribution in brain that corre-
sponds to the known distribution of DAT. This binding was specific, reversible and the kinetics of
[
18F]6 binding in brain were faster than for its lead compound, [11C]1. The possible presence of a hydroxy-
Keywords:
PET
Radioligand
Dopamine transporter
PE2I
methyl-radiometabolite formed by oxidation in the 3b-benzylic position of [18F]6 warrants further
detailed evaluation of the metabolism of [18F]6. [18F]6 is a potential radioligand for imaging DATs in
the human brain with PET.
Ó 2009 Elsevier Ltd. All rights reserved.
Tropane
The dopaminergic (DA) reuptake system has since long been of
central interest in relation to the pathophysiology of several neuro-
psychiatric and neurodegenerative disorders as well as being a tar-
get for drugs of abuse. Since the first positron emission
tomography (PET) imaging study of the DA transporter (DAT)
was conducted with [11C]nomifensine in a hemi-parkinsonian pa-
tient in 1987,1 there have been several studies demonstrating the
utility of DAT imaging for diagnostic applications. In addition, by
using PET and [11C]cocaine, knowledge has been gained in the
bio-distribution and pharmacokinetics of this drug inside the living
human brain that would not have been obtained otherwise.2
Given the high concentration of DAT in brain (Bmax ꢀ 200 pmol/
g in human putamen),3 and that the binding potential (BP) of a PET
radioligand is proportional to the concentration of target sites
(BP = Bmax/Kd), the DAT is a relatively straight-forward target for
radioligand development. Consequently, there are several success-
ful examples of radioligands that enable DAT imaging in vivo, most
of which have been derived from cocaine. PE2I (1, Fig. 1), FECNT
(2), FECT (3) and b-CIT (4) are examples of such radioligands.4–7
Although radioligands exist that have been used in clinical settings
already, none considered optimal for the accurate quantification of
DATs is available today. Problems associated with current radioli-
gands are slow kinetics, lack of selectivity and/or problems with
blood brain barrier (BBB) permeable radiometabolites.4,8,9
Figure 1. Structures of cocaine and some DAT radioligands developed for PET.4–7
In view of our aim to develop an improved DAT radioligand that
could overcome some of the abovementioned problems, we
decided to use 1 as a lead compound. Compound 1 has excellent
pharmacological properties, being a potent (Ki = 17 nM) and selec-
tive DAT inhibitor.10 In addition, [11C]1 has been used for PET imag-
ing of DATs in both pre-clinical and clinical settings, thereby
showing the potential for analogs of [11C]1 to be useful as PET
imaging agents. However, two clear disadvantages also exist for
* Corresponding author. Present address: AstraZeneca R&D Södertälje, Medicinal
Chemistry, S-151 85 Södertälje, Sweden.
E-mail address: magnus.schou@astrazeneca.com (M. Schou).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.06.032

4844
M. Schou et al. / Bioorg. Med. Chem. Lett. 19 (2009) 4843–4845
[
11C]1. These originate in its metabolism and its slow binding
kinetics in brain. In a detailed report on the metabolism of
11C]1, Shetty et al. showed the presence of two radiometabolites
[
of [11C]1 inside the rodent brain.9 These radiometabolites are pro-
duced by sequential oxidation of the benzylic position of the 3b-
phenyl ring, first to the alcohol and then to the corresponding acid.
Radiometabolites in brain pose a threat to the accurate quantifica-
tion of a PET measurement, since the PET camera merely measures
radioactivity in brain and cannot distinguish between chemical
entities. It is therefore important to perform radiometabolite mea-
surements and, if possible, introduce the radiolabel in a position of
the molecule that prevents formation of BBB permeable
radiometabolites.
Scheme 2. Reagents and conditions: (a) [18F]Fꢁ, K2.2.2, K2CO3, o-DCB, 135 °C,
10 min; (b) TBAH, DMF, 85 °C, 20 min.
A potential disadvantage of using 1 as a lead compound for radi-
oligand development is thus that analogs of 1 may show similar
benzylic oxidation radiometabolites as those of [11C]1. However,
Shetty et al. also found only a small level of the 2b-carboxylic acid
in the rat (all in urine), suggesting that the ester moiety of 1 is rel-
atively metabolically stable. In the light of these results, we
decided to use the 2b-carboxy position for introducing a fluoroeth-
yl group, since this would serve for several purposes. The first
being to avoid extensive formation of the BBB permeable product,
2-[18F]fluoroethanol, which was shown in rodent brain during PET
studies with [18F]2 (in that case produced by oxidative N-dealkyla-
tion).8 The second purpose for introducing an 18F-label into an ana-
log of 1, is that it facilitates imaging at later time-points (due to the
longer half-life of 18F (t1/2 = 109.8 min, compared to 11C (t1/
2 = 20.4)), which is of special interest since [11C]1 displays slow
kinetics in the human brain in vivo.11 In addition, it is also well-
known that substitution in the 2b ester position on this scaffold
is well tolerated at the DAT (i.e., this moiety can be modified with
retention of affinity). Another general advantage with fluorine-18
is its lower positron energy as compared with carbon-11
(650 keV vs 960 keV, respectively), which results in improved spa-
tial resolution of the obtained PET images. This is important when
imaging the substantia nigra (SN), which is a small region of the
brain responsible for dopamine production in brain and therefore
of central interest in research on the DA system.
nished the Michael addition product, which was demethylated
and alkylated with the E-isomer of tributyltin propenyl chloride.
Iodination and hydrolysis of the methyl ester yielded 4 in 18%
overall yield from cocaine.12
Compound 6 was assayed for affinity at the DAT using compe-
tition assays with [125I]PE2I on rodent tissue. As expected, 6 had re-
tained its affinity at rodent DAT (Ki = 12 1.7 nM (n = 3) vs
Ki = 16 nM for 1). To further evaluate the potential 6 as a candidate
PET radioligand for imaging DAT in the human brain, we radiola-
belled 6 with fluorine-18. The precursor for radiolabelling, [18F]flu-
oroethyl bromide ([18F]8) was prepared from [18F]fluoride ion by
adapting a previously described method ( Scheme 2).13 Briefly,
bromoethyl tosylate (7) in o-dichlorobenzene was added to a dried
[K2.2.2][18F]fluoride–carbonate complex. After a 10 min reaction,
[
18F]8 was distilled off into an ice-cold reaction vessel containing
DMF. [18F]8 was obtained in 38% radiochemical yield (uncorrected
for decay) from [18F]fluoride within 45 min after end of bombard-
ment (EOB).
[
18F]6 was obtained by heating an alkaline mixture of 4 and
[
18F]8 for 20 min. After completed heating, [18F]6 was purified by
semi-preparative HPLC. After isolation of the product fraction
and evaporation of solvents, [18F]6 was formulated in a mixture
of physiologically buffered saline, ethanol and propylene glycol be-
fore final sterilization by filtration (0.22
lipore, Sweden). The overall radiochemical yield of [18F]6 from
18F]fluoride was 7% (uncorrected for decay) and the total synthesis
time was 90 min. The radiochemical purity of [18F]6 was >95% and
the specific radioactivity was 113–385 GBq/ mol. The structure of
l
m MillexÒ GV filter; Mil-
The synthesis of the target compound 6 is depicted in Scheme 1
and started from naturally occurring cocaine. The first seven steps
from cocaine to 4 are well described in the literature.10,12 Shortly,
hydrolysis of the benzoate moiety in cocaine and subsequent elim-
ination of the resultant alcohol in a mixture of methanol and
hydrochloric acid produced anhydroecgonine methyl ester. React-
ing this unsaturated ester with p-tolylmagnesium bromide fur-
[
l
[
18F]6 was also confirmed by comparison of the fragmentation
spectra of reference 6 and the carrier associated with[18F]6 (major
fragments, m/z (relative abundance): 394.15 (12%), 366.15 (18%),
248.05 (15%), 166.97 (20%)).
Two PET measurements were conducted in cynomolgus mon-
keys with [18F]6 for 180 min each. Monkey 1 underwent one base-
line study (59 MBq). Monkey 2 underwent a displacement study
with intravenous administration of GBR 12909 (2.5 mg/kg) at
30 min after injection of [18F]6 (67 MBq). The results are shown
in Figure 2 (below).
After injection of [18F]6, radioactivity readily entered brain in a
high amount, with approximately 540% SUV present in the caudate
at 13 min after injection. The rank order for regional radioactivity
uptake in brain was caudate > putamen > thalamus ꢀ mid-
brain ꢀ frontal cortex ꢀ cerebellum. The ratios of radioactivity to
the cerebellum peaked at 105 min and were 7.5 and 7.9 for the
caudate and putamen, respectively. In the displacement experi-
ment, the binding ratio to cerebellum decreased with 81% (to
1.4) in the caudate and 83% (to 1.3) in the putamen at the end of
the PET measurement. Thus, the distribution of radioactivity in
brain was reversible, specific and in accordance with the known
distribution of DATs in primate brain. Rather surprisingly, [18F]6
also showed much faster kinetics in brain than [11C]1,7,11 which
Scheme 1. Reagents and conditions: (a) POCl3, DCM, rt, 4 h; (b) 2-fluoroethanol,
Et3N, DCM, 0 °C?rt.

M. Schou et al. / Bioorg. Med. Chem. Lett. 19 (2009) 4843–4845
4845
Baseline
Displacement
600
500
400
300
200
100
0
750
500
250
0
Caudate
Caudate
Putamen
Thalamus
Midbrain
Putamen
Thalamus
Midbrain
Cerebellum
Cerebellum
0
25 50 75 100 125 150 175 200
Time (min)
0
25 50 75 100 125 150 175 200
Time (min)
Figure 2. Time–activity curves following injection with [18F]6. SUV represents standard uptake values (%injected dose per gram tissue multiplied with body weight). The
arrow denotes the administration of the displacement compound, GBR12909.
is of benefit to the patient by reducing the time spent in the PET
camera.
with the experiments. DIMI (Projects EC-FP6, LSVB-CT-2005-
512146) is also gratefully acknowledged for financial support.
The time-course for the metabolism of [18F]6 was studied by
radio-HPLC on acetonitrile extracts from venous monkey plasma.14
The recovery from the analytical procedure was greater than 91%.
Two major radiometabolite fractions were observed, probably cor-
responding to oxidation products of the 3b-benzylic position (i.e.,
to the hydroxyl and carboxylic acid derivatives of [18F]6) as previ-
ously shown for the closely structurally related [11C]1.9 Although it
appears that the 3b-benzylic oxidation product is formed to a low-
er extent as compared after injection of [18F]6 to after injection of
References and notes
1. Aquilonius, S. M.; Bergstrom, K.; Eckernas, S. A.; Hartvig, P.; Leenders, K. L.;
Lundquist, H.; Antoni, G.; Gee, A.; Rimland, A.; Uhlin, J., et al Acta. Neurol. Scand.
1987, 76, 283.
2. Fowler, J. S.; Volkow, N. D.; Wang, G. J.; Gatley, S. J.; Logan, J. Nucl. Med. Biol.
2001, 28, 561.
3. Madras, B. K.; Gracz, L. M.; Fahey, M. A.; Elmaleh, D.; Meltzer, P. C.; Liang, A. Y.;
Stopa, E. G.; Babich, J.; Fischman, A. J. Synapse 1998, 29, 116.
4. Muller, L.; Halldin, C.; Farde, L.; Karlsson, P.; Hall, H.; Swahn, C. G.; Neumeyer,
J.; Gao, Y.; Milius, R. Nucl. Med. Biol. 1993, 20, 249.
[
11C]1, the presence of M1 in plasma warrants further detailed
evaluation of the metabolites of [18F]6 by radio-LC and LC–MS/
MS. This, as well as a more detailed description of the time-course
for metabolite formation in monkeys, will be reported elsewhere.
In summary, a new fluoroethyl analogue of PE2I, compound 6,
was prepared and found to be a potent inhibitor of rodent DAT in
vitro. As a candidate radioligand, 6 was radiolabelled with 18F at
high specific radioactivity and in good radiochemical yield. The
binding of [18F]6 to DAT was specific, reversible and showed faster
kinetics in brain than [11C]1. The possible presence of a hydroxy-
methyl-radiometabolite formed by oxidation in the 3b-benzylic
position of [18F]6 warrants a further detailed evaluation by LC–
MS/MS. This will be reported elsewhere.
5. Goodman, M. M.; Kilts, C. D.; Keil, R.; Shi, B.; Martarello, L.; Xing, D.; Votaw, J.;
Ely, T. D.; Lambert, P.; Owens, M. J.; Camp, V. M.; Malveaux, E.; Hoffman, J. M.
Nucl. Med. Biol. 2000, 27, 1.
6. Halldin, C.; Farde, L.; Lundkvist, C.; Ginovart, N.; Nakashima, Y.; Karlsson, P.;
Swahn, C. G. Synapse 1996, 22, 386.
7. Halldin, C.; Erixon-Lindroth, N.; Pauli, S.; Chou, Y. H.; Okubo, Y.; Karlsson, P.;
Lundkvist, C.; Olsson, H.; Guilloteau, D.; Emond, P.; Farde, L. Eur. J. Nucl. Med.
Mol. Imaging 2003, 30, 1220.
8. Zoghbi, S. S.; Shetty, H. U.; Ichise, M.; Fujita, M.; Imaizumi, M.; Liow, J. S.; Shah,
J.; Musachio, J. L.; Pike, V. W.; Innis, R. B. J. Nucl. Med. 2006, 47, 520.
9. Shetty, H. U.; Zoghbi, S. S.; Liow, J. S.; Ichise, M.; Hong, J.; Musachio, J. L.;
Halldin, C.; Seidel, J.; Innis, R. B.; Pike, V. W. Eur. J. Nucl. Med. Mol. Imaging 2007,
34, 667.
10. Emond, P.; Garreau, L.; Chalon, S.; Boazi, M.; Caillet, M.; Bricard, J.; Frangin, Y.;
Mauclaire, L.; Besnard, J. C.; Guilloteau, D. J. Med. Chem. 1997, 40, 1366.
11. Jucaite, A.; Odano, I.; Olsson, H.; Pauli, S.; Halldin, C.; Farde, L. Eur. J. Nucl. Med.
Mol. Imaging 2006, 33, 657.
12. Stepanov, V.; Schou, M.; Jarv, J.; Halldin, C. Appl. Radiat. Isot. 2007, 65, 293.
13. Zhang, M. R.; Tsuchiyama, A.; Haradahira, T.; Yoshida, Y.; Furutsuka, K.; Suzuki,
K. Appl. Radiat. Isot. 2002, 57, 335.
14. Halldin, C.; Swahn, C.-G.; Farde, L.; Sedvall, G. Radioligand Disposition and
Metabolism. In PET for Drug Development and Evaluation; Comar, D., Ed.; Kluwer
Academic, 1995; pp 55–65.
Acknowledgements
The authors thank Arsalan Amir, Kenneth Dahl, Siv Eriksson, Ju-
lio Gabriel, Guennadi Jogolev, Gudrun Nylén, Phong Truong and the
Karolinska Institutet PET group for excellent technical assistance