Dopamine D3 receptor antagonists: The quest for a potentially selective PET ligand. Part 3: Radiosynthesis and in vivo studies

Article abstract of DOI:10.1016/j.bmcl.2009.07.055

Compound 1 is a potent and selective antagonist of the dopamine D₃ receptor. With the aim of developing a carbon-11 labeled ligand for the dopamine D₃ receptor, 1 was selected as a potential PET probe. [¹¹C]1 was obtained

Full text of DOI:10.1016/j.bmcl.2009.07.055

Bioorganic & Medicinal Chemistry Letters 19 (2009) 5056–5059
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Dopamine D3 receptor antagonists: The quest for a potentially selective PET
ligand. Part 3: Radiosynthesis and in vivo studies
Idriss Bennacef a, Cristian A. Salinas a, Thomas A. Bonasera a, Roger N. Gunn a, Hélène Audrain b,
Steen Jakobsen b, Nabeel Nabulsi c, David Weinzimmer c, Richard E. Carson c, Yiyun Huang c, Ian Holmes d,
g
e
e
a,f,
*
Fabrizio Micheli e, , Christian Heidbreder , Gabriella Gentile , Tino Rossi , Marc Laruelle
*
a GlaxoSmithKline Clinical Imaging Centre, Imperial College London, Hammersmith Hospital, Du Cane Road, London, UK
b Aarhus PET Centre, Aarhus University Hospital, Aarhus, Denmark
c PET Centre, Yale University School of Medicine, New Haven CT, USA
d Cancer Therapeutics, Melbourne, Australia
e GlaxoSmithKline, Neurosciences Centre of Excellence for Drug Discovery, Verona, Italy
f Division of Neurosciences and Mental Health, Imperial College London, Neuroscience Laboratories, Burlington Danes Building, Hammersmith Hospital, UK
g Altria Client Service, Richmond VA, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Compound 1 is a potent and selective antagonist of the dopamine D3 receptor. With the aim of developing
a carbon-11 labeled ligand for the dopamine D3 receptor, 1 was selected as a potential PET probe. [11C]1
was obtained by palladium catalyzed cross coupling using [11C]cyanide and 4 with a specific activity of
Received 19 May 2009
Revised 7 July 2009
Accepted 7 July 2009
Available online 25 July 2009
55.5 25.9 GBq/lmol (1.5 0.7 Ci/l
mol). [11C]1 was tested in porcine and non-human primate models to
assess its potential as a radioligand for PET imaging of the dopamine D3 receptor. We conclude that in
both species and despite appropriate in vitro properties, [11C]1 does not show any specific signal for
the dopamine D3 receptor.
Keywords:
Dopamine D3
Antagonist
PET
Ó 2009 Elsevier Ltd. All rights reserved.
Ligand
In 1990, Sokoloff et al. first cloned, characterized and reported the
dopamine (DA) D3 receptor subtype.1 The expression pattern of D3
mRNA in rat and human brain are comparable, with high expression
primarily in limbic system areas (globus pallidus, caudate, putamen,
andmedialthalamus). Various anatomical, pharmacological, and ge-
neticobservationssuggestthattheDA D3 receptormight be involved
in Parkinson’s disease, schizophrenia, and drug addiction.2
Although potent and selective DA D3 receptor antagonists can be
found in the literature3 and despite tremendous efforts from a num-
berof groups, relativelylittle progresshas beenmadein identifying a
selective DA D3 radioligand.4–6 To the best of our knowledge, only
DA receptor its routine use to examine all D3 DA rich brain regions
is complicated by its adjunct affinity at the D2 receptor. To use
[
11C]PHNO as a tool for imaging D3 receptors throughout the brain
would require the use of a selective D3 blocker, which is difficult.
Thus, the identification of a selective DA D3 PET radioligand would
be an important advance in this research field.
A large number of molecules have been identified as part of our
research towards a DA D3 antagonist, some with nanomolar affin-
ity and a good selectivity profile over other dopaminergic recep-
tors. In this Letter, we describe the synthesis and in vitro
properties of 1 (Fig. 1), its radiosynthesis and subsequent in vivo
evaluation as a PET probe in pig and monkey.12
[
11C]PHNO, a probe firstly described as a DA D2 radioligand can be
used with PET to map the DA D3 receptor.7,8 PHNO is an agonist that
bindstoboththeD2 (pKi = 8.2)andtheD3 (pKi = 9.7)receptors.9 Non-
human primate studies have recently demonstrated that it is possi-
ble to extract the DA D3 signal from the analysis of regional
Synthesis ofcompound 1is depicted inScheme 1. 1-(2-Chloroethyl)-
2-imidazolidinone
2
was reacted with 4-(trifluoromethyl)
11C]PHNO data which indicated that the ubtantia-(>90%) and glo-
NC
CF3
[
O
bus pallidum (60%) binding potentials were mostly D3.10 Although
N
11C]PHNO is currently the best available tool for imaging the D3
N
N
[
* Corresponding authors. Tel.: +39 045 8218515; fax: +39 045 8218196 (F.M.).
E-mail addresses: fm20244@gsk.com (F. Micheli), Marc.A.Laruelle@gsk.com (M.
1
Laruelle).
Figure 1.
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.07.055

I. Bennacef et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5056–5059
5057
N
CF3
O
N
N
N
N
I
1
CF3
CF3
b
O
O
HCl.
HN
a
Cl
N
N
N
N
N
b
CF3
I
O
3
I
2
N
N
N
I
4
Scheme 1. Reagents and conditions: (a) K2CO3, NMP, 100 °C, 2 Â 10 min, cooling, 90%; (b) [trans-2-aminocyclohexyl]amine, CuI, KOH, toluene, MW, overnight, 25%.
piperidine hydrochloride and potassium carbonate in NMP at 100 °C,
affording 1-{2-[4-(trifluoromethyl)-1-piperidinyl]ethyl}-2-imidazolid-
inone3ingood yield (90%). Compound 3was treatedwith 3-iodobenzo-
nitrile or 1,3-diiodobenzene, [trans-2-aminocyclohexyl]amine, copper
iodide, and potassium hydroxide in toluene, under microwave irradia-
tion, affording 1or 1-(3-iodophenyl)-3-{2-[4-(trifluoromethyl)-1-pipe-
ridinyl]ethyl}-2-imidazolidinone 4, which was purified by preparative
HPLC.13
As reported previously, compound 1 is a potent (pKi = 9.1)
antagonist for the DA D3 receptor.11 The selectivity profile of 1
was assessed by filtration binding assays on human DA D3 and
D2 receptors and demonstrated a 100-fold selectivity for the D3
DA over D2 DA receptor. Further assays indicated that it was selec-
cyanide and urea functional groups are suitable sites for labeling
with carbon-11. Therefore, radiosynthesis of [11C]1 was carried
out and PET imaging experiments in pigs and monkeys conducted
to investigate its potential as a PET ligand to image the D3 receptor
in vivo.
Because of its robustness, radiolabeling with [11C]cyanide label-
ing route was chosen.17 Radiosynthesis of [11C]1 was accomplished
via a palladium(0)-catalyzed cross coupling between the iodide 4
and [11C]cyanide using Pd2dba3ÁCHCl3/dppf and KHCO3 as base in
either NMP (80 °C for 3 min) or DMF (100 °C for 5 min). The crude
product was purified by reverse-phase HPLC and the fraction con-
taining the product was diluted with water and concentrated
through a C18 SepPak cartridge. Elution of the retained product
from the cartridge with ethanol and formulation with saline
yielded 899 326 MBq (24.3 8.8 mCi) of [11C]1 at the end of syn-
thesis (EOS) (n = 7). Specific activity at EOS was 55.5 25.9 GBq/
tive (1000-fold) over serotoninergic receptors (5HT1A, 5HT1B,
5HT1D, 5HT2A, 5HT2C, 5HT6, and 5HT7) and adrenergic receptors (al-
pha1A and alpha1B).14 The high affinity for the DA D3 receptor was
retained in rat and pig native tissues (Table 1).11
Compound 1 displayed a suitable pharmacokinetic profile for iv
administration in rat, with good brain penetration (brain to blood
concentration ratio = 6.4) and high free (protein-unbound) fraction
(15.1%) in the brain. Binding to plasma protein was low in both rat
(72%) and pig (61%).11
Lipophilicity measurements at pH 7.4 of 1 were similar as deter-
mined by HPLC (CHI-log D = 2.9)15 and by a phosphate buffer (pH
7.4): octanol shake flask method (log D = 2.7, n = 3; maximum
range, +5%). These lipophilicity measures are in the accepted range
for CNS PET tracers (log D between 1 and 3).16
l
mol (1.5 0.7 Ci/lmol) and both chemical and radiochemical
purities were greater than 99%. (Scheme 2).18
To assess the binding profile of [11C]1, firstly, a series of block-
ing studies were performed in the Landrace pig (n = 2) with
SB277011, a high affinity (pKi = 8.0 for D3) and selective (Ki D3/
D2 < 100) D3 antagonist as the blocking agent.19 After bolus injec-
tion of 320 122 MBq (8.7 3.3 mCi), compound [11C]1 readily
entered the brain and a homogeneous distribution was observed
at both baseline and following iv administration of SB277011
(Fig. 2). No lipophilic metabolite was observed and the radioli-
gand displayed a moderate rate of metabolism with ꢀ50% of par-
ent remaining in the plasma at 30 min after radioligand
injection.20
Overall, compound 1 displays acceptable physico-chemical and
in vitro properties necessary for a PET radioligand. Moreover, the
Table 1
DA (human D2, D3, rat and pig D3) receptor affinities, brain penetration, and lipophilicity of 1
hpKi DA D2
6.8
hpKi DA D3
9.1
pKi rat NT
pKi pig NT
Rat brain to blood level ratio
6.4
CHI-log D
Log D
1
9.1
9.1
2.9
2.7
Affinity results. SEM for D3 and D2 filtration binding data is 0.2. NT: native tissue.
N
11C
I
O
O
CF3
CF3
a or b, c
N
N
N
N
N
N
[11C]1
4
Scheme 2. Reagents and conditions: (a) [11C]CNÀ, Pd2dba3, dppf, KHCO3, NMP, 80 °C, 3 min; or (b) [11C]CNÀ, Pd2dba3, dppf, KHCO3, DMF, 100 °C, 5 min; (c) semi-preparative
HPLC.

5058
I. Bennacef et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5056–5059
Figure 4. [11C]1 uptake images summed from 60 to 90 min post-injection at
baseline (B) and after SB277011 (C). Corresponding MRI images (A).
Figure 2. [11C]1 uptake images in the pig brain, summed from 0 to 90 min post-
injection at baseline. Axial (A), coronal (B), and sagital (C) sections (striatum
region).
0.5
Binding potential (BPND) values were estimated indirectly from
total distribution volumes derived from a one tissue compartment
model in the D3 DA target regions of interest (globus pallidus, cau-
date, putamen, and medial thalamus). Only small values for BPND
were observed at baseline for all target regions (ꢀ0.2) and these
values were unaffected following administration of SB277011
(Fig. 3). These data did not support the existence of a specific D3
DA signal with [11C]1 in the pig.
Baseline
SB-277011 2.0 mg/kg
0.4
0.3
0.2
0.1
0
The candidate radioligand was then evaluated in Rhesus mon-
keys using the same SB277011 blocking paradigm.21 In the monkey
[
11C]1 also displayed a moderate rate of metabolism with ꢀ50% of
parent remaining in plasma after 30 min with no lipophilic metabo-
lites. The plasma free fraction of [11C]1 was 32 5% (n = 6). Figure 4
shows a set of brain PET images summed from 60 to 90 min at the
level of the striatum and globus pallidus and the corresponding ana-
tomical MRI image displaying the same slices. In the brain, equilib-
rium in regional uptake was reached by about 60 min after
injection of [11C]1.
Globus pallidus
Caudate
Putamen
Thalamus
Figure 5. Regional binding potentials from baseline-blocking experiments in
rhesus monkey.
To conclude, the synthesis and radiosynthesis of [11C]1 were ro-
bust and straightforward. Physico-chemical and in vitro biological
profiles of 1 were assessed and brain PET imaging in pig and non-
human primate were performed. Despite displaying appropriate
physico-chemical properties, high in vitro binding affinity and
selectivity for the DA D3 receptor, no specific PET signal could be
detected in the pig or monkey in vivo. Hence [11C]1 is not suitable
for imaging the DA D3 receptors. Although the failure of [11C]1 as a
PET radioligand is not completely understood, it indicates that an
in vitro pKi higher than 9.1 for D3 receptors may be required of fu-
ture generation of candidate PET tracers, because of the relatively
low density of this receptor subtype in the brain.
Binding potential was calculated by the equilibrium ratio of re-
gions to cerebellum averaged from 60 to 90 min (Fig. 5). Regional
BPND values were small and unaffected by the SB277011
challenge.22
As observed for the pig, [11C]1 does not display any D3-specific
signal in monkey brains.
0.5
Baseline (n=3)
SB-277011 0.5 mg/kg (n=2)
0.4
Acknowledgments
0.3
0.2
0.1
0.0
The authors would like to thank Dr. Giorgio Bonanomi for the
preparation of derivative 4 and Carla Marchioro’s group for struc-
tural and analytical elucidation.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2009.07.055.
References and notes
globus pallidus
caudate
putamen
medial thalamus
Figure 3. Regional binding potential estimates for [11C]1 at baseline and following
administration of a selective D3 DA antagonist (SB277011) in pig.
1. Sokoloff, P.; Giros, B.; Martres, M. P.; Marie, L.; Schwartz, J. C. Nature 1990, 347,
146.

I. Bennacef et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5056–5059
2. Sokoloff, P.; Dias, J.; Le Foll, B.; Guillin, O.; Lerich, L.; Bezard, E.; Gueremy, C. CNS 14. Data not shown.
5059
Neurol. Disord.: Drug Targets 2006, 5, 25.
15. Valko, K.; Du, C. M.; Bevan, C.; Reynolds, D. P.; Abraham, M. H. Curr. Med. Chem.
2008, 8, 1137.
3. Micheli, F.; Bonanomi, G.; Braggio, S.; Capelli, A. M.; Celestini, P.; Damiani, F.;
Fabio, R. D.; Donati, D.; Gagliardi, S.; Gentile, G.; Hamprecht, D.; Petrone, M.;
Radaelli, S.; Tedesco, G.; Terreni, S.; Worby, A.; Heidbreder, C. Bioorg. Med.
Chem. Lett. 2008, 18, 901.
4. Kuhnast, B.; Valette, H.; Besret, L.; Demphel, S.; Coulon, C.; Ottaviani, M.;
Guillermier, M.; Bottlaender, M.; Dollé, F. Nucl. Med. Biol. 2006, 33, 785.
5. de Vries, E. F. J.; Kortekaas, R.; van Waarde, A.; Dijkstra, D.; Elsinga, P. H.;
Vaalburg, W. J. Nucl. Med. 2005, 46, 1384.
16. Waterhouse, R. N. Mol. Imaging Biol. 2003, 5, 376.
17. Sandell, J.; Halldin, C.; Hall, H.; Thorberg, S. O.; Werner, T.; Sohn, D.; Sedvall, G.;
Farde, L. Nucl. Med. Biol. 1999, 26, 159.
18. The amount of [11C]CO2 at EOB was about 50 GBq (1.35 Ci). In the synthesis of
[
11C]1 for the pig studies, the yield of formulated product and specific activity
were 2.09 1.71 GBq (56.6 46.2 mCi) and 40.7 11.1 GBq (1.1 0.3 Ci/umol)
(n = 5), respectively.
6. Gao, M.; Mock, B. H.; Hutchins, G. D.; Zheng, Q. H. Bioorg. Med. Chem. 2005, 13,
6233.
19. Stemp, G.; Ashmeade, T.; Branch, C. L.; Hadley, M. S.; Hunter, A. J.; Johnson, C.
N.; Nash, D. J.; Thewlis, K. M.; Vong, A. K. K.; Austin, N. E.; Jeffrey, P.; Avenell, K.
Y.; Boyfield, I.; Hagan, J. J.; Middlemiss, D. N.; Reavill, C.; Riley, G. J.; Routledge,
C.; Wood, M. J. Med. Chem. 2000, 43, 1878.
7. McCormick, P. N.; Kapur, S.; Seeman, P.; Wilson, A. A. Nucl. Med. Biol. 2008, 35, 11.
8. Ginovart, N.; Willeit, M.; Rusjan, P.; Graff, A.; Bloomfield, P. M.; Houle, S.;
Kapur, S.; Wilson, A. A. J. Cereb. Blood Flow Metab. 2006, 27, 857.
9. van Vliet, L. A.; Rodenhuis, N.; Dijkstra, D.; Wikstrom, H.; Pugsley, T. A.; Serpa,
K. A.; Meltzer, L. T.; Heffner, T. G.; Wise, L. D.; Lajiness, M. E.; Huff, R. M.;
Svensson, K.; Sundell, S.; Lundmark, M. J. Med. Chem. 2000, 43, 2871.
10. Rabiner, E. A.; Slifstein, M.; Nobrega, J.; Plisson, C.; Huiban, M.; Raymond, R.;
Diwan, M.; Wilson, A. A.; McCormick, P.; Gentile, G.; Gunn, R. N.; Laruelle, M.
Synapse 2009, 63, 782.
[11]. Micheli, F.; Holmes, I.; Arista, L.; Bonanomi, G.; Braggio, S.; Cardullo, F.; Di
Fabio, R.; Donati, D.; Gentile, G.; Hamprecht, D.; Terreni, S.; Heidbreder, C.;
Savoia, C.; Griffante, C.; Worby, A. Bioorg. Med. Chem. Lett. 2009, 19, 4011.
12. The research complied with national legislation and with company policy on
the Care and Use of Animals and with related codes of practice.
13. 1H NMR spectra were obtained on Varian INOVA spectrometer (500 MHz).
Compound 1 (DMSO-d6): d 8.02–7.96 (t, 1H), 7.91–7.83 (dd, 1H), 7.56–7.47 (t,
1H); 7.46–7.38 (d, 1H); 3.86–3.75 (t, 2H); 3.58–3.46 (t, 2H); 3.34–3.26 (t, 2H);
3.04–2.92 (d, 2H); 2.49–2.41 (t, 2H); 2.30–2.16 (m, 1H); 2.01–1.89 (t, 2H);
1.82–1.71 (d, 2H); 1.46–1.31 (dq, 2H). Compound 4 (DMSO-d6): d 8.09–7.98 (m,
1H), 7.50–7.41 (d, 1H), 7.34–7.29 (d, 1H); 7.12–7.04 (t, 1H); 3.80–3.70 (dd, 2H);
3.53–3.45 (dd, 2H); 3.35–3.23 (m, 2H); 3.04–2.89 (d, 2H); 2.51–2.39 (t, 2H);
2.30–2.15 (m, 1H); 2.00–1.86 (m, 2H); 1.80–1.71 (m, 2H); 1.48–1.32 (m, 2H).
LC/MS analysis proved purity >98% for both compounds.
20. The fraction of unchanged [11C]1 was measured by radio-HPLC in extracts of
plasma from samples taken throughout the scanning session. Free fraction of
[
11C]1 in the plasma was determined by ultrafiltration. Brain regions with the
highest concentrations of
[
11C] GSK981352 are shown in yellow, with
progressively lower concentrations displayed in orange and red, respectively.
PET images are displayed on a common scale.
21. One test–retest study and two control-blocking studies were performed in
three different rhesus monkeys using [11C]1. Each monkey received two iv
injections of [11C]1, separated by 3–4 h. A 2-h emission scan was acquired on
the High Resolution Research Tomograph (HRRT) scanner for each injection.
Brain regions with the highest concentrations of [11C]GSK981352 are shown in
red, with progressively lower concentrations displayed in yellow, green, and
blue, respectively. PET images are displayed on a common scale. For the test/
retest study, the radioligand [11C]1 was administered to the animal as a bolus.
In the blocking studies, solution of SB-277011 (2 mg/kg) was infused for a total
of 3 h, starting 1 h before the second scan. Activity in the blood was measured
during the scan. Additional blood samples were taken for analysis of [11C]1 and
its metabolites by HPLC. Free fraction of [11C]1 in the plasma was determined
by ultrafiltration.
22. The apparent increase in BPND is attributed to normal variation in a dataset
with little specific binding signal.