Synthesis and in vivo Evaluation of Fluorine-18 and Iodine-123 Pyrazolo[4,3-e]-1,2,4-triazolo[1,5-c]pyrimidine Derivatives as PET and SPECT Radiotracers for Mapping A2AReceptors

Article abstract of DOI:10.1002/cmdc.201600219

Imaging agents that target adenosine type 2A (A_2A) receptors play an important role in evaluating new pharmaceuticals targeting these receptors, such as those currently being developed for the treatment of movement disorders like Parkinson′s di

Full text of DOI:10.1002/cmdc.201600219

DOI: 10.1002/cmdc.201600219
Full Papers
Synthesis and in vivo Evaluation of Fluorine-18 and
Iodine-123 Pyrazolo[4,3-e]-1,2,4-triazolo[1,5-c]pyrimidine
Derivatives as PET and SPECT Radiotracers for Mapping
A2A Receptors
Christine Vala, Thomas J. Morley,* Xuechun Zhang, Caroline Papin, Adriana
Alexandre S. Tavares, H. Sharon Lee, Cristian Constantinescu, Olivier Barret,
Vincent M. Carroll, Ronald M. Baldwin, Gilles D. Tamagnan, and David Alagille[a]
Imaging agents that target adenosine type 2A (A2A) receptors
play an important role in evaluating new pharmaceuticals tar-
geting these receptors, such as those currently being devel-
oped for the treatment of movement disorders like Parkinson’s
disease. They are also useful for monitoring progression and
treatment efficacy by providing a noninvasive tool to map
changes in A2A receptor density and function in neurodegener-
ative diseases. We previously described the successful evalua-
tion of two A2A-specific radiotracers in both nonhuman pri-
mates and in subsequent human clinical trials: [123I]MNI-420
and [18F]MNI-444. Herein we describe the development of both
of these radiotracers by selection from a series of A2A ligands,
based on the pyrazolo[4,3-e]-1,2,4-triazolo[1,5-c]pyrimidine
core of preladenant. Each of this series of 16 ligands was
found to bind to recombinant human A2A receptor in the low
nanomolar range, and of these 16, six were radiolabeled with
either fluorine-18 or iodine-123 and evaluated in nonhuman
primates. These initial in vivo results resulted in the identifica-
tion of 7-(2-(4-(4-(2-[18F]fluoroethoxy)phenyl)piperazin-1-yl)eth-
yl)-2-(furan-2-yl)-7H-pyrazolo[4,3-e][1,2,4]triazolo[1,5-c]pyrimi-
din-5-amine
([18F]MNI-444)
and
7-(2-(4-(2-fluoro-4-
[
123I]iodophenyl)piperazin-1-yl)ethyl)-2-(furan-2-yl)-7H-imida-
zo[1,2-c]pyrazolo[4,3-e]pyrimidin-5-amine ([123I]MNI-420) as PET
and SPECT radiopharmaceuticals for mapping A2A receptors in
brain.
Introduction
The A2A receptor is one of four known adenosine receptors[1,2]
and is found in the central nervous system, localized in the
striatum, nucleus accumbens, and olfactory tubercle.[3,4] These
receptors co-localize with striatal dopamine type 2 (D2) recep-
tors, in the form of heterodimers, and the two receptors are
functionally linked.[5,6] This functional link makes the A2A recep-
tor an alternative therapeutic target, to the D2 receptor, for the
treatment of movement disorders, such as Parkinson’s and
Huntingdon’s disease (PD and HD),[7–10] as well as ischemic
brain damage[11,12] and epilepsy.[13] Furthermore, dysregulation
of A2A receptors has been linked to psychiatric disorders such
as schizophrenia,[14,15] attention deficit hyperactivity disorder,[16]
mood disorder,[17] depression,[18] and addiction.[19,20]
whether drug candidates cross the blood–brain barrier and
engage with their target in a dose-dependent manner, thereby
helping guide the dose selection in early clinical trials.
[11C]TMSX was the first example of an A2A radiotracer, but suf-
fered from poor brain uptake as well as chemical instabili-
ty.[26,27] More recent examples of A2A radiotracers are based
around the pyrazolotriazolopyrimidine core of the A2A agonist
preladenant,[28] such as [11C]preladenant,[29] [11C]SCH442416,[30]
and [18F]MRS5425.[31,32] Recently we reported [123I]MNI-420,[33]
a SPECT radiotracer, and [18F]MNI-444, a PET radiotracer for
imaging A2A receptors in brain.[34] Evaluation of both of these
radiotracers in nonhuman primates followed by human sub-
jects demonstrated that both were suitable radioligands for
imaging A2A receptors, presenting the highest binding in brain
regions with an expected high density of A2A and a tracer
uptake that was blocked by the preladenant in a dose-depen-
dent manner.
An A2A-specific radiotracer could be used to evaluate drugs
that target the A2A receptor[21–25] by providing information on
[a] Dr. C. Vala, Dr. T. J. Morley, Dr. X. Zhang, Dr. C. Papin, Dr. A. A. S. Tavares,
Dr. H. S. Lee, Dr. C. Constantinescu, Dr. O. Barret, Dr. V. M. Carroll,
Dr. R. M. Baldwin, Dr. G. D. Tamagnan, Dr. D. Alagille
Molecular NeuroImaging, a division of inviCRO,
Herein we describe the medicinal chemistry program that
led to the identification of both of these radioligands from
a series of 16 compounds designed to allow radiolabeling with
fluorine-18 or iodine-123. The compounds were evaluated in
vitro for their binding affinity to recombinant human A2A re-
ceptor, and the most promising compounds were radiolabeled
with either [18F]fluoride or [123I]iodide. The results of these ini-
60 Temple Street, Suite 8B, New Haven, CT 06510 (USA)
E-mail: tmorley@mnimaging.com
Supporting information (detailed protocols for all chemical and radio-
chemical reactions, and characterization data for all synthesized com-
pounds) and the ORCID identification number(s) for the author(s) of this
article can be found under http://dx.doi.org/10.1002/cmdc.201600219.
ChemMedChem 2016, 11, 1 – 9
1
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
tial in vivo results in nonhuman primates are presented as jus-
tification for selection of the two imaging agents.
group. Vinyl iodides 1 f and 2 f necessitated a two-step synthe-
sis through a stannane intermediate (11 and 12), followed by
Boc deprotection for their generation.
N-Alkylation of each of the phenylpiperazines 3a–f and 4a–f
with 5 to give series 1 or 2 was achieved by heating a twofold
excess of the piperazine in DMF (Scheme 1). Compounds 1h,
2g, and 1h were also synthesized from their corresponding
para-iodo- and para-fluorophenylpiperazines using identical
chemistry.[39,40] In all cases the reactions went to completion, as
monitored by TLC, but difficulties in isolating the highly insolu-
ble products led to variable yields (14–69%) (Table 1).
Compounds 1b–h and 2a–h were screened for human A2A
receptor binding affinity using an in vitro competition assay
against [3H]CGS21680 in transfected HEK-293 cells.[41] Ki values
were determined from an eight-point dose–response assay of
the ligand under test. An A2A receptor antagonist,
ZM241385,[42] was used as a positive control (Kd =
0.14 nm in rat striatal membranes), and compound
15 was included in order to compare the results of
this assay with those of previous reports. All of the
screened compounds had nanomolar affinity (1.4–
2.8 nm) for the human A2A receptor, in the same
range as that of preladenant (Ki =1.1 nm).[43] Further-
more, the compounds bearing an extra fluorine on
the phenylpiperazine ring (R1 =F) showed slightly in-
creased binding relative to the unsubstituted series
(R1 =H). It is worth noting that compound 15 had
a Ki value of 1.3 nm, which is consistent with previ-
ous values reported for this compound (1.6 nm),[44]
indicating that the results of this assay are comparable with re-
ported data. This series of compounds was not screened
against off-target binding to other CNS or PDE targets, which
could affect their suitability as PDE10-specific ligands.
The clogP values covered a narrow range (2.6–4.1), with the
unsaturated side chains showing a higher value than the satu-
rated analogues. The values are all in the range of 2–4, which
are recommended for good brain uptake.[45] For comparison,
with previous radiotracers the clogP values of [11C]TMSX,
Results and Discussion
Chemistry
Preparation of the set of 16 ligands 1 and 2 was accomplished
using a divergent synthesis strategy, broadly following the
published route to preladenant,[35] and according the general
synthetic route shown in Scheme 1. We synthesized a series of
phenylpiperazines 3 and 4, whereby the para substituents (R2)
contained either a fluorine or iodine, which were then alkylat-
ed by a common N7-ethyl tosylate pyrazolotriazolopyrimidine
derivative 5.[35]
Scheme 1. Preparation of compound series 1 and 2: a) NEt3, DMF, 808C, 16 h.
Our initial SAR efforts focused on derivatizing the phenol pi-
perazine core (6 and 7, Scheme 2) with small fluoro- and iodo-
containing aliphatic side chains in order to provide a handle
from which the radioisotope could be incorporated. The side
chains were chosen for their well-documented ability to readily
incorporate the radiohalogen in the final ligand. To this end
phenols 6 and 7 were used to synthesize the O-alkylated phe-
nylpiperazine products 1a–e and 2a–e from their correspond-
ing tosylates 8a–e,[36–38] after removal of the Boc protecting
Scheme 2.
Synthesis of fluoro- and iodophenylpiperazines: a) NaH, DMF, RT, 34–95%; b) TFA, CH2Cl2, RT, or HCl(aq), RT, 39–99%; c) trans-3-(tributylstannyl)allyl tosylate,
NaH, DMF, RT, 83% for 11 and 79% for 12; d) I2, CH2Cl2, RT, 84% for 13 and 78% for 14.
&
ChemMedChem 2016, 11, 1 – 9
www.chemmedchem.org
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
perature stability of formulated [18F]1b was monitored by
radio HPLC for up to 8 h, and the RCP remained >95% over
this period.
Table 1. Binding affinities of 1b–h and 2a–h for recombinant human A2A
receptor.[a]
Compd
R1
R2
Yield [%][b]
Ki [nm][c]
clogP
Our initial efforts in synthesizing a trimethylstannyl deriva-
tive as precursor for [123I]1h proved unsuccessful due to the
highly unstable nature of the stannyl product. Instead, a direct
radioiodination method using phenylpiperazine 15 was used,
which yielded the desired para-[123I]iodo product [123I]1h
(Scheme 4). The related compound [123I]2h, bearing an extra
fluorine atom on the phenyl ring, was found to be unreactive
to the same direct radiolabeling; instead, the trimethylstannyl
derivative 21 was used. This precursor was synthesized by N-al-
kylation of phenylpiperazine 22 with tosylate 5. The
1a
1b
1c
1d
1e
1 f
1g
1h
2a
2b
2c
2d
2e
2 f
H
H
H
H
H
H
H
H
F
F
F
F
F
FCH2O
FCH2CH2O
(E)-FCH2CH=CHCH2O
14
28
58
18
20
54
63
69
57
15
37
15
20
26
30
41
64
NT
2.8
2.4
2.4
1.8
1.5
2.6
1.5
2.1
1.8
1.4
1.6
2.1
2.0
1.7
2.0
1.3
2.78
2.86
3.19
3.61
2.61
4.16
2.90
3.88
3.07
3.15
3.47
3.90
2.90
4.45
3.10
4.08
2.59
FCH2CꢀCCH2O
FCH2CH2OCH2CH2O
(E)-ICH=CHCH2O
F
I
FCH2O
FCH2CH2O
(E)-FCH2CH=CHCH2O
[
[
123I]iododestannylation of 21 gave para-[123I]iodo product
123I]1h in 80% radiochemical yield. The related [123I]1 f was
FCH2CꢀCCH2O
FCH2CH2OCH2CH2O
F
F
F
H
(E)-ICH=CHCH2O
synthesized via the same [123I]iododestannylation chemistry,
but by using vinyl tributylstannane 23 (synthesized from 5 and
1) as the precursor. In all three cases, the radiolabeling reac-
tions were quenched by the addition of a basic sodium meta-
bisulfite solution before purification of the crude product by
2g
2h
15
F
I
H
[a] See Scheme 1 for scaffold structure. [b] Isolated yield. [c] Binding affin-
ity for recombinant human A2A receptor; NT: not tested.
t
reversed-phased HPLC. The products were isolated on a C18
light Sep-Pak, eluted with ethanol and formulated into physio-
logical saline. In all cases the RCP of the formulated products
was >95%.
[11C]SCH442416, and [18F]MRS5425 are 2.93, 3.18, and 1.95, re-
spectively. For the potential fluorine-18 radiotracers, those
compounds bearing the glycol side chain (1e and 2e) showed
the lowest clogP values; we chose these two compounds and
fluoroethyl compound 1b for fluorine-18 labeling and imaging
in nonhuman primates. For the four potential iodine-123 radio-
tracers, compounds 1 f, 1h, and 2h were chosen for radiolab-
eling and in vivo evaluation.
Imaging
Compounds [18F]1b, [18F]1e, and [18F]2e were evaluated in
adult female baboons (Papio anubis). Brain PET imaging was
used to quantify radiotracer brain uptake, kinetics, and binding
properties. The radiotracers were injected intravenously as
a bolus, and images were acquired over 3 h. All three tracers
Radiochemistry
Synthesis of the tosylate precursors for radiofluorina-
tion began by coupling of 6 or 7 with tosylate 5 to
give 16 or 17 (Scheme 3).[44] O-Alkylation of 16 (R1 =
H) with 1,2-ethyl-p-ditosylate gave the radiolabeling
precursor 18. O-Alkylation of 16 or 17 with diethy-
lene glycol di(p-ditosylate) gave precursors 19 and
20. Radiolabeling of these precursors with fluorine-
18 was achieved using a commercial GE TRACERlabꢁ
FXF-N module by displacement of the tosylate with
[18F]fluoride in DMSO in the presence of potassium
carbonate and Kryptofix-222. The crude product was
purified by reversed-phased HPLC, the radiolabeled
t
product was isolated on a C18 light Sep-Pak, eluted
with ethanol and formulated into a physiological
saline solution containing ascorbic acid (1.9 mm) to
decrease radiolysis of the radiotracer, and polysor-
bate-80 (3.4 mm; 3.3% final ethanol concentration).
Compounds [18F]1b, [18F]1e, and [18F]2e were ob-
tained in an overall decay-corrected radiochemical
Scheme 3.
yield of 10–20% in a synthesis time of 60 min. All
Radiosynthesis of PET tracers: a) NEt3, DMF, 908C, 54% for 16 and 49% for 17;
b) (TsOCH2)2, NaH, DMF, RT, 21%; c) (TsOCH2CH2)2O, NaH, DMF, RT, 14% for 19 and 8% for
20; d) [18F]fluoride, K2CO3/K222, DMSO, 1008C, 10 min, 23% for [18F]1b and 8% for [18F]1e
and [18F]2e.
three 18F-labeled radiotracers were obtained with
>99% radiochemical purity (RCP) and with specific
activity between 2 and 6 CimmolÀ1. The room tem-
ChemMedChem 2016, 11, 1 – 9
www.chemmedchem.org
3
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
~30 min for [18F]1e and [18F]2e, followed by a signifi-
cant washout within 90 min of imaging. For compar-
ison, the PET A2A radiotracers [11C]TMSX and
[11C]SCH442416 showed maximum uptake ratio to
cerebellum in striatum of 1.6 in rhesus macaque
(Macaca mulatta) and 2.2 in pig-tailed macaque
(Macaca nemestrina), respectively.[30,46] In terms of ki-
netics, [18F]1b is the preferred radiotracer, presenting
higher peak uptake in whole brain and faster wash-
out kinetics than [18F]1e and [18F]2e. Further studies
have been undertaken to demonstrate the specificity
of [18F]1b for the A2A receptor, by performing suc-
cessful blocking studies in nonhuman primate using
both preladenant and tozadenant, which are de-
tailed elsewhere.[34] No significant bone uptake, an
indication of in vivo radio-defluorination, was ob-
served in any of the three tested radiotracers. How-
ever, the slow washout and, particularly for [18F]1e
and [18F]2e, an increase in signal over time observed
in cerebellum could indicate the presence of small
amounts of radioactive metabolites. Limited metab-
olite analysis was evaluated by HPLC after plasma
denaturation at 5, 10, 30, 60, and 120 min post-injec-
tion. These results showed a very similar pattern be-
tween all tracers tested with a plasma parent frac-
tion of 80, 45, 31, 16, and 10% for [18F]1b; 81, 54,
Scheme 4. Radiosynthesis of SPECT radiotracers: a) [123I]iodide, H3PO4, peracetic acid,
DMF, RT, 60 min, 13–49% (n=5); b) tosylate 5, NEt3, DMF, 808C, 42%; c) [123I]iodide,
H3PO4, peracetic acid in dioxane, RT, 20 min, 79% (n=2); d) [123I]iodide, H3PO4, peracetic
acid, DMF, RT, 20 min, 8–17% (n=4).
entered the baboon brain rapidly, presenting peak uptake in
whole brain of 1.1–2.4% of injected dose (%ID; Table 2). The
distribution in brain regions was consistent with known A2A
39, 23, and 8% for [18F]1e; and a slightly faster metabolism of
68, 32, 18, 10, and 5% for [18F]2e.
The in vivo properties of [123I]1h, [123I]2h, and [123I]1 f were
also evaluated in three adult female baboons. In these cases
brain SPECT imaging was used to assess radiotracer brain pen-
etration, kinetics and binding properties in vivo. The radiotrac-
ers were injected intravenously as a bolus, and images were
acquired over 2 h post-injection. Both [123I]1h and [123I]2h rap-
idly entered the baboon brain; however, [123I]1 f exhibited neg-
ligible brain penetration (Figure 2). The distribution of both
Table 2. Imaging results in baboon brain.
Radiotracer
Peak brain
Peak
Max. SUVr
BPND
uptake [%ID]
brain SUV
(in striatum)
[18F]1b
[18F]1e
[18F]2e
2.4
1.1
1.8
0.5
0.9
1.4
5.5
2.5
3.5
0.9
0.8
1.4
5.6
4.1
6.0
1.1
1.2
4.1
3.16
1.73
4.24
–
–
–
[
123I]1h and [123I]2h is consistent with known A2A densities,
being highest in the striatum and lowest in the cortical areas
and cerebellum. Conversely, [123I]1 f showed no apparent bind-
ing to any of the brain regions investigated. The peak standar-
dized uptake values (SUV) and standardized uptake values
ratios (SUVr), calculated using the cerebellum as reference
region, are shown in Table 2, with [123I]2h showing by far the
greatest uptake in A2A-rich areas (4.1 versus 1.5)
These results indicate that [123I]2h showed the greatest
promise of the three potential SPECT radiotracers tested, as
123I]1 f showed very low brain penetration, and [123I]1h has
[
123I]1 f
[
[
123I]1h
123I]2h
distribution patterns (Figure 1), where the highest uptake was
found in the striatum (caudate and putamen), and the lowest
uptake was found in cortical areas and cerebellum. Peak stand-
ardized uptake values (SUV) in the striatum were found to be
3.5–5.5. Standardized uptake value ratios (SUVr) were calculat-
ed using the cerebellum as reference region, which was as-
sumed to have negligible A2A expression. The maximum SUVr
(occurring between 30–60 min post-injection), surrogate esti-
mate of the specific binding to A2A for the three tracers, is re-
ported in Table 2 and ranked from 4.1 to 6.0. Binding poten-
tials, BPND, estimated using noninvasive Logan graphical analy-
sis with cerebellum as a reference region ranked from 1.73 to
4.24. The brain kinetics were more favorable for [18F]1b with
a fast brain uptake within 10 min of injection, relative to
[
a lower accumulation in desired A2A-rich areas of the brain
(SUV=1.45 versus 4.08 for [123I]2h). Based on these data
[
123I]2h was advanced for further in vivo evaluation in nonhu-
man primate and was shown to be blocked, in a dose-depen-
dent manner, by the A2A-specific ligand preladenant.[33] This
radioligand has since also been successfully evaluated in
human clinical trials.[47]
&
ChemMedChem 2016, 11, 1 – 9
www.chemmedchem.org
4
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Figure 1. Representative PET results following a single intravenous bolus injection of A) [18F]1b, B) [18F]1e, or C) [18F]2e in baboons. Left: SUV images (sum 0–
180 min post-injection). Right: SUV time–activity curves. Note the high uptake in A2A-rich regions, such as caudate and putamen, relative to lower uptake in
regions with decreased levels of A2A, such as the cortex and cerebellum.
Figure 2. Brain SPECT imaging results of A) [123I]1h (left), B) [123I]2h (center), and C) [123I]1 f (right).
Conclusions
An additional three of this series were radiolabeled with
A small series of potential A2A imaging agents was synthesized,
the structures of which are based on the pyrazolotriazolopyri-
midine core of preladenant. In vitro evaluation of the series
showed that all bind to recombinant human A2A receptors
with nanomolar affinity. Three of this series were radiolabeled
with fluorine-18 and evaluated in baboon. All three radioli-
gands rapidly entered the brain, with a peak whole-brain %ID
of 1.1–2.4%, and showed an uptake pattern consistent with
A2A receptor distribution. The SUV values for each of these
tracers was good, between 5.5 and 2.5, with [18F]1b showing
the highest SUV (5.5). Furthermore, [18F]1b showed high (5.6)
peak SUVr in the striatum and a rapid washout from the brain.
[
123I]iodine and also evaluated in baboon. Two of these three
failed to either enter the brain, or accumulate significantly in
A2A-rich regions. However, [123I]2h showed high brain uptake
(SUV=4.1) in the desired brain regions.
The high brain uptake, rapid kinetics, and high target-to-
nontarget ratios compared with other A2A imaging agents indi-
cated that [18F]1b ([18F]MNI-444) and [123I]2h ([123I]MNI-420)
warranted further evaluation in nonhuman primates.[33,34]
These in vivo evaluations have already been performed and
are detailed elsewhere, and showed that both are specific for
the A2A receptor. Furthermore, human evaluation of both
[
123I]2h[33] and [18F]1b have been successfully performed.[48]
ChemMedChem 2016, 11, 1 – 9
www.chemmedchem.org
5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
5) The reactor vial was cooled to 408C, its contents were diluted
with preparative HPLC mobile phase (1.5 mL) and the mixture was
transferred to the loop-loading vial.
Experimental Section
Chemistry: Commercially available reagents were purchased from
Sigma–Aldrich (St. Louis, MO, USA), and used as supplied without
further purification. Reactions were carried out under an atmos-
phere of nitrogen. Analytical TLC was carried out with aluminum-
backed plates coated with Merck Kieselgel 60 GF254. Visualization
was realized by UV light (254 nm) or with a ninhydrin stain. Flash
chromatography was performed with silica gel (60 ꢂ, 200–
6) The reactor vial was rinsed with water and the rinse was also
transferred to the loop-loading vial.
7) The contents of the loop-loading vial was transferred to the
HPLC loop, via a fluid detector, injected onto the prep-HPLC
column (Agilent Eclipse XDB C18, 9.4ꢃ250 mm) and eluted with
a mixture of acetonitrile/water/triethylamine (40:60:0.2 v/v/v) at
a flow rate of 4 mLminÀ1. The eluent was monitored by UV (l=
254 nm) and radioactive detectors connected in series.
1
425 mesh). H and 13C NMR spectra were recorded at 238C on an
AC Bruker 500 MHz instrument. Chemical shifts (d) are expressed in
parts per million (ppm) downfield from tetramethylsilane and are
referenced to the atom resonance of the solvent (CHCl3: d=
7.26 ppm for 1H and d=77.2 for 13C). For 1H NMR, data are present-
ed as follows: chemical shift, multiplicity (s, singlet; d, doublet; dt,
doublet of triplet; m, multiplet; bs, broad signal), coupling con-
stants (J in Hz), and integration). For 13C NMR data, signals corre-
sponding to CH, CH2, or CH3 groups are assigned from DEPT. High-
resolution mass spectrometry (HRMS) data were recorded with an
electron ionization spray (ESI) technique at the School of Chemical
Science, University of Illinois. All compounds tested were >95%
pure, as determined by elemental and/or HPLC analysis. Elemental
analyses were performed by Atlantic Microlab (Norcross, GA, USA).
In vitro binding assays were performed by CEREP (Redmond, WA,
USA) using an in vitro competition assay against [3H]CGS21680 in
transfected HEK-293 cells.[41]
8) The product-containing fraction was collected in a flask contain-
ing water (15 mL) and the diluted product fraction was transferred
onto a pre-conditioned solid-phase extraction cartridge (Waters
t
Sep-Pak C18 light, pre-rinsed with 6 mL ethanol and 6 mL water).
9) The cartridge was washed with water (10 mL) and the product
was eluted from the cartridge with ethanol (1 mL) into a second
flask containing the drug product matrix (17 mL of normal saline).
The cartridge was rinsed with a further portion of the formulation
matrix (2 mL) and the rinse was added to the second flask.
10) The contents of the second flask were pushed through a steri-
lizing filter (Millex LG 0.22 mm, Millipore) into a vented final prod-
uct vial, which had been assembled under aseptic conditions.
General procedure for 123I radioiodination: The shipping vial con-
taining [123I]iodide (20–120 mCi) and NaOH (0.1n, 1.0 equiv) was re-
constituted with dilute phosphoric acid solution (0.8m, 1.05 equiv),
followed by dilute peracetic acid solution (19% solution in water,
50 mL), followed by a solution of the precursor in THF (50 mg in
50 mL). The mixture was thoroughly mixed and allowed to stand at
ambient temperature for the desired reaction time, after which
was added sodium metabisulfite solution (10 mg) in saturated
sodium bicarbonate solution (100 mL), followed by triethylamine
(50 mL). The mixture was further diluted with HPLC mobile phase
(0.4 mL), withdrawn from the reaction vial, injected onto a Waters
NovaPak C18 (4 mm, 4.6ꢃ250 mm) HPLC column and eluted at
1 mLminÀ1. The collected product fraction was diluted to 10 mL
with USP sterile water for injection and loaded onto a Waters
Radiochemistry: Radiolabeling was performed on a GE TRACER-
labꢁ FXF-N module using helium pressure or vacuum for fluid trans-
fers. [18F]Fluoride was purchased from Cardinal Health (Hartford,
CT, USA). QC HPLC testing was performed with a Waters HPLC
system (1525 or e2695), monitoring the column effluent with UV
(Waters 2489, l=280 nm) and radioactivity detectors (Bioscan FC-
200) with the data analyzed by Waters Empower software (Version
Build 3.3471). Radioactivity was measured with a Capintec CRC-25
dose calibrator. Radionuclidic identity was determined by g spec-
troscopy using a Canberra 802 multichannel analyzer. pH measure-
ments were performed with pH indicator strips (Range 0–14,
Sigma–Aldrich). Cryptand-222 (Kryptofix-222) content was mea-
sured using a TLC test as published.[39] Residual solvent analysis
was performed using a PerkinElmer Clarus 500 gas chromatograph
fitted with a Restek Stabilwax (30 mꢀ0.53 mmꢃ1 mm) column. Py-
rogen content was measured by the LAL test with a Charles River
Laboratories Endosafe PTS-100. Sterility tests were performed by
inoculating a sample of the drug product in fluid thioglycollate
medium and tripticase soy broth (Northeast Laboratories) and
monitoring for two weeks, per USP guidelines.
t
SepPak C18 light. The SepPak was rinsed with a 25% ethanol in
water solution before the product was eluted from the SepPak
using USP-grade ethanol (600 mL) through
a sterilizing filter
(0.22 mm) into a sterile final product vial. USP sterile saline (9 mL)
was then used to rinse the SepPak and filter into the final product
vial. The product identity was confirmed with analytical radio-HPLC
by co-injection with the cold standard using the same HPLC
column and mobile phase.
General procedure for 18F radiofluorination: A GE TRACERlabꢁ
FXF-N synthesis module was operated in the following sequence:
1) The [18F]fluoride in target water (0.5 to 3 mL) was transferred
from the shipping vial and trapped on a short ion-exchange solid-
phase extraction (SPE) cartridge (ORTG, TN).
2) The cartridge was eluted into the pre-heated (608C) reactor vial
with a solution of cryptand-222 (7.1 mg, 19 mmol) and potassium
carbonate (0.75 mg, 5.4 mmol) dissolved in a mixture of acetonitrile
(0.7 mL) and water (0.3 mL).
3) The reactor was heated at 958C under vacuum with a stream of
helium for 4 min before addition of acetonitrile (1 mL) and evapo-
ration continued for 2 min. A second portion of acetonitrile (1 mL)
was added and drying was continued for a further 2 min before
the reactor was cooled to 708C.
4) A solution of tosylate precursor (1 mg) in anhydrous DMSO
(1.0 mL) was added to the reactor vial, which was then sealed and
heated at 1008C for 10 min.
In vivo PET imaging: Three adult female baboons (Papio anubis)
weighing on average 17.6 kg (15.4–19.3 kg) were used for PET
scanning following intravenous bolus injection of [18F]1b, [18F]1e,
or [18F]2e (170.6Æ14.8 MBq, mean ÆSD, n=3). The animals were
handled in accordance with institutional animal care protocols in
compliance with Federal regulations. All PET data were acquired
using a Biograph mCT PET/CT camera (Siemens Healthcare Molecu-
lar Imaging, Knoxville, TN, USA). Animals were fasted for up to 24 h
prior to imaging. On the imaging day, the animal was first anesthe-
tized with intramuscular ketamine (10 mgkgÀ1) and glycopyrrolate
(0.1–0.2 mgkgÀ1) or propofol 1% emulsion (2.0–4.0 mgkgÀ1), trans-
ferred to the PET camera, and immediately intubated with an en-
dotracheal tube for continued anesthesia with 2.5% isoflurane ad-
ministered through a re-breathing circuit. A transmission scan,
which was later used for attenuation and scatter correction, was
acquired using the camera CT. Subsequently and immediately fol-
lowing radiotracer administration, a list mode emission scan was
&
ChemMedChem 2016, 11, 1 – 9
www.chemmedchem.org
6
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
[1] B. B. Fredholm, J. F. Chen, R. A. Cunha, P. Svenningsson, J. M. Vaugeois,
Int. Rev. Neurobiol. 2005, 63, 191–270.
[2] B. B. Fredholm, A. P. IJzerman, K. A. Jacobson, K. N. Klotz, J. Linden, Phar-
macol. Rev. 2001, 53, 527–552.
acquired for a period of 180 min. The dynamic PET sinograms were
corrected for photon attenuation and scatter and reconstructed
using interactive reconstruction methods.
Reconstructed scans were imported into PMOD 3.308 software
(PMOD Technologies, Zurich, Switzerland) and motion correction
was performed by creating an average image of consecutive scans
with absence of motion, which was then used as a reference for
rigid matching of all scans in the study. An in-house-developed
baboon brain magnetic resonance imaging (MRI) template was
used for image co-registration and placement of volumes of inter-
est (VOIs) for quantification of radiotracer uptake. The T1-weighted
MR images were acquired with a GE Signa unit (General Electric,
Milwaukee, WI, USA) at 1.5 T with a spoiled gradient recall se-
quence with the following settings: TR=25 ms, TE=5 ms, NEX=2,
matrix=256ꢃ256, field of view=16 cm. The MR images were re-
duced from an initial matrix of 256ꢃ256 with 240 axial slices (axial
slice thickness=0.7 mm) to a 128ꢃ128 matrix with 120 axial slices
(axial slice thickness=1.4 mm). The following brain regions were
manually drawn on the MRI: striatum, caudate, putamen, frontal
cortex, temporal cortex, parietal cortex, occipital cortex, cerebellum
and whole brain. Average images of [18F]1b, [18F]1e or [18F]2e PET
data were generated by averaging the scans presenting the high-
est uptake. The generated average image was then co-registered
to the MRI template using an automatic rigid matching tool. The
transformation matrix was saved and subsequently applied to the
dynamic PET series. Finally, MRI derived VOIs were applied to the
co-registered PET images for image quantification.
[3] S. N. Schiffmann, F. Libert, G. Vassart, J. E. Dumont, J. J. Vanderhaeghen,
Brain Res. 1990, 519, 333–337.
[4] P. Svenningsson, C. Le Moine, G. Fisone, B. B. Fredholm, Prog. Neurobiol.
1999, 59, 355–396.
[5] K. Fuxe, S. Ferre, M. Canals, M. Torvinen, A. Terasmaa, D. Marcellino, S. R.
Goldberg, W. Staines, K. X. Jacobsen, C. Lluis, A. S. Woods, L. F. Agnati, R.
Franco, J. Mol. Neurosci. 2005, 26, 209–219.
[6] J. F. Chen, R. Moratalla, F. Impagnatiello, D. K. Grandy, B. Cuellar, M. Ru-
binstein, M. A. Beilstein, E. Hackett, J. S. Fink, M. J. Low, E. Ongini, M. A.
Schwarzschild, Proc. Natl. Acad. Sci. USA 2001, 98, 1970–1975.
[7] A. R. Carta, A. Kachroo, N. Schintu, K. Xu, M. A. Schwarzschild, J. Wardas,
M. Morelli, J. Neurochem. 2009, 111, 1478–1489.
[8] J. D. Salamone, A. J. Betz, K. Ishiwari, J. Felsted, L. Madson, B. Mirante, K.
Clark, L. Font, S. Korbey, T. N. Sager, J. Hockemeyer, C. E. Muller, Front.
Biosci. 2008, 13, 3594–3605.
[9] P. Popoli, D. Blum, A. Martire, C. Ledent, S. Ceruti, M. P. Abbracchio,
Prog. Neurobiol. 2007, 81, 331–348.
[10] J. Hierholzer, M. Cordes, S. Venz, L. Schelosky, C. Harisch, W. Richter, U.
Keske, N. Hosten, J. Maurer, W. Poewe, R. Felix, J. Nucl. Med. 1998, 39,
954–960.
[11] J.-F. Chen, P. K. Sonsalla, F. Pedata, A. Melani, M. R. Domenici, P. Popoli, J.
Geiger, L. V. Lopes, A. De Mendonca, Prog. Neurobiol. 2007, 83, 310–
331.
[12] A. Monopoli, G. Lozza, A. Forlani, A. Mattavelli, E. Ongini, Neuroreport
1998, 9, 3955–3959.
[13] A. Huber, M. Guttinger, H. Mohler, D. Boison, Neurosci. Lett. 2002, 329,
289–292.
[14] A. Kurumaji, M. Toru, Brain Res. 1998, 808, 320–323.
[15] K. Alsene, J. Deckert, P. Sand, H. de Wit, Neuropsychopharmacology
2003, 28, 1694–1702.
[16] Y. Molero, C. Gumpert, E. Serlachius, P. Lichtenstein, H. Walum, D. Jo-
hansson, H. Anckarsꢄter, L. Westberg, E. Eriksson, L. Halldner, Genes
Brain Behav. 2013, 12, 305–310.
[17] D. van Calker, K. Biber, Neurochem. Res. 2005, 30, 1205–1217.
[18] M. El Yacoubi, J. Costentin, J.-M. Vaugeois, Neurology 2003, 61, S82–
S87.
[19] R. M. Brown, J. L. Short, J. Pharm. Pharmacol. 2008, 60, 1409–1430.
[20] C. V. Gomes, M. P. Kaster, A. R. Tome, P. M. Agostinho, R. A. Cunha, Bio-
chim. Biophys. Acta Biomembr. 2011, 1808, 1380–1399.
[21] B. B. Fredholm, R. A. Cunha, P. Svenningsson, Curr. Top. Med. Chem.
2003, 3, 413–426.
[22] M. A. Schwarzschild, Prog. Neurobiol. 2007, 83, 261–262.
[23] R. J. Gillespie, S. J. Bamford, R. Botting, M. Comer, S. Denny, S. Gaur, M.
Griffin, A. M. Jordan, A. R. Knight, J. Lerpiniere, S. Leonardi, S. Lightowler,
S. McAteer, A. Merrett, A. Misra, A. Padfield, M. Reece, M. Saadi, D. L. Sel-
wood, G. C. Stratton, D. Surry, R. Todd, X. Tong, V. Ruston, R. Upton,
S. M. Weiss, J. Med. Chem. 2009, 52, 33–47.
[24] V. Katritch, V.-P. Jaakola, J. R. Lane, J. Lin, A. P. IJzerman, M. Yeager, I. Ku-
fareva, R. C. Stevens, R. Abagyan, J. Med. Chem. 2010, 53, 1799–1809.
[25] R. A. Cunha, S. Ferre, J.-M. Vaugeois, J.-F. Chen, Curr. Pharm. Des. 2008,
14, 1512–1524.
Standardized uptake values (SUV) were calculated as concentration
in the VOI divided by the injected dose divided by the animal
weight. In addition, the SUV values obtained for each brain region
were divided by the SUV values determined in the cerebellum, the
selected reference region, to generate SUV ratios (SUVr). Both SUV
and SUVr values were then plotted against time.
General synthetic method: pyrazolotriazolopyrimidine tosylate
and arylpiperazine coupling: Tosylate 5 (1.0 equiv) and arylpipera-
zine (1.9 equiv) were dissolved in DMF before the addition of trie-
thylamine (7 equiv). This mixture was heated at 808C for 16 h after
which it was allowed to cool to room temperature, and the excess
solvent was evaporated under reduced pressure. Mixtures of ethyl
acetate, hexane or methanol, as described in individual sections,
were added to the residue and the resulting solid product was ob-
tained by filtration.
Abbreviations
A2A: adenosine type 2A, D2: dopamine type 2, Boc: tert-butoxycar-
bonyl, DCY: decay-corrected yield, DMAP: 4-dimethylaminopyri-
dine, DME: dimethoxyethanol, DMF: N,N-dimethylformamide,
DMSO: dimethyl sulfoxide, HD: Huntington’s disease, HPLC: high-
performance liquid chromatography, ID: injected dose, mCPBA:
meta-chloroperbenzoic acid, NMP: N-methylpyrrolidine, NT: not
tested, PD: Parkinson’s disease, PET: positron emission tomography,
RCP: radiochemical purity, RT: room temperature, SPECT: single
photon emission computed tomography, SUV: standardized uptake
values, SUVr: standardized uptake value ratios, TFA: trifluoroacetic
acid, TLC: thin-layer chromatography, Ts: para-toluenesulfonyl.
[26] K. Ishiwata, W. F. Wang, Y. Kimura, K. Kawamura, K. Ishii, Ann. Nucl. Med.
2003, 17, 205–211.
[27] K. Ishiwata, J. Noguchi, S. Wakabayashi, J. Shimada, N. Ogi, T. Nariai, A.
Tanaka, K. Endo, F. Suzuki, M. Senda, J. Nucl. Med. 2000, 41, 345–354.
[28] R. A. Hodgson, P. J. Bedard, G. B. Varty, T. M. Kazdoba, T. Di Paolo, M. E.
Grzelak, A. J. Pond, A. HadjTahar, N. Belanger, L. Gregoire, A. Dare, B. R.
Neustadt, A. W. Stamford, J. C. Hunter, Exp. Neurol. 2010, 225, 384–390.
[29] X. Zhou, S. Khanapur, A. P. Huizing, R. Zijlma, M. Schepers, R. A. J. O. Di-
erckx, A. van Waarde, E. F. J. de Vries, P. H. Elsinga, J. Med. Chem. 2014,
57, 9204–9210.
[30] R. M. Moresco, S. Todde, S. Belloli, P. Simonelli, A. Panzacchi, M. Riga-
monti, M. Galli-Kienle, F. Fazio, Eur. J. Nucl. Med. Mol. Imaging 2005, 32,
405–413.
[31] A. K. Bhattacharjee, L. Lang, O. Jacobson, B. Shinkre, Y. Ma, G. Niu, W. C.
Trenkle, K. A. Jacobson, X. Chen, D. O. Kiesewetter, Nucl. Med. Biol. 2011,
38, 897–906.
Keywords: adenosine 2A receptor · movement disorders ·
Parkinson’s disease · PET · SPECT
ChemMedChem 2016, 11, 1 – 9
www.chemmedchem.org
7
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
[32] S. Khanapur, S. Paul, A. Shah, S. Vatakuti, M. J. B. Koole, R. Zijlma,
R. A. J. O. Dierckx, G. Luurtsema, P. Garg, A. van Waarde, P. H. Elsinga, J.
Med. Chem. 2014, 57, 6765–6780.
[33] A. A. d. S. Tavares, J. Batis, O. Barret, D. Alagille, C. Vala, G. Kudej, A.
Koren, K. P. Cosgrove, K. Nice, J. H. Kordower, J. Seibyl, G. D. Tamagnan,
Nucl. Med. Biol. 2013, 40, 403–409.
[34] O. Barret, J. Hannestad, D. Alagille, C. Vala, A. Tavares, C. Papin, T. J.
Morley, K. Fowles, H. Lee, J. Seibyl, D. Tytgat, M. Laruelle, G. Tamagnan,
J. Nucl. Med. 2014, 55, 1712–1718.
[43] R. A. Hodgson, R. Bertorelli, G. B. Varty, J. E. Lachowicz, A. Forlani, S.
Fredduzzi, M. E. Cohen-Williams, G. A. Higgins, F. Impagnatiello, E. Nico-
lussi, L. E. Parra, C. Foster, Y. Zhai, B. R. Neustadt, A. W. Stamford, E. M.
Parker, A. Reggiani, J. C. Hunter, J. Pharmacol. Exp. Ther. 2009, 330, 294–
303.
[44] B. R. Neustadt, J. Hao, N. Lindo, W. J. Greenlee, A. W. Stamford, D. Tul-
shian, E. Ongini, J. Hunter, A. Monopoli, R. Bertorelli, C. Foster, L. Arik, J.
Lachowicz, K. Ng, K.-I. Feng, Bioorg. Med. Chem. Lett. 2007, 17, 1376–
1380.
[35] P. G. Baraldi, F. Fruttarolo, M. A. Tabrizi, D. Preti, R. Romagnoli, H. El-
Kashef, A. Moorman, K. Varani, S. Gessi, S. Merighi, P. A. Borea, J. Med.
Chem. 2003, 46, 1229–1241.
[45] H. Pajouhesh, G. R. Lenz, NeuroRX 2005, 2, 541–553.
[46] M. Mishina, K. Ishiwata, Y. Kimura, M. Naganawa, K. Oda, S. Kobayashi, Y.
Katayama, K. Ishii, Synapse 2007, 61, 778–784.
[36] M. D. Bartholomꢄ, V. Gottumukkala, S. Zhang, A. Baker, P. Dunning, F. H.
Fahey, S. T. Treves, A. B. Packard, J. Med. Chem. 2012, 55, 11004–11012.
[37] R. Iwata, S. Furumoto, C. Pascali, A. Bogni, K. Ishiwata, J. Labelled
Compd. Radiopharm. 2003, 46, 555–566.
[38] J. L. Musachio, J. R. Lever, Tetrahedron Lett. 1989, 30, 3613–3616.
[39] K. G. Liu, A. J. Robichaud, Tetrahedron Lett. 2005, 46, 7921–7922.
[40] S. N. VanderWel, P. J. Harvey, D. J. McNamara, J. T. Repine, P. R. Keller, J.
Quin, R. J. Booth, W. L. Elliott, E. M. Dobrusin, D. W. Fry, P. L. Toogood, J.
Med. Chem. 2005, 48, 2371–2387.
[47] A. A. S. Tavares, J. C. Batis, C. Papin, D. Jennings, D. Alagille, D. S. Russell,
C. Vala, H. Lee, R. M. Baldwin, I. G. Zubal, K. L. Marek, J. P. Seibyl, O.
Barret, G. D. Tamagnan, J. Nucl. Med. 2013, 54, 1760–1767.
[48] O. Barret, J. Hannestad, C. Vala, D. Alagille, A. Tavares, M. Laruelle, D.
Jennings, K. Marek, D. Russell, J. Seibyl, G. Tamagnan, J. Nucl. Med.
2015, 56, 586–591.
[41] P. Puntambekar, J. Van Buren, M. Raisinghani, L. S. Premkumar, V. Ram-
kumar, J. Neurosci. 2004, 24, 3663–3671.
Received: April 22, 2016
Revised: June 15, 2016
[42] A. Uustare, A. Vonk, A. Terasmaa, K. Fuxe, A. Rinken, Life Sci. 2005, 76,
1513–1526.
Published online on && &&, 0000
&
ChemMedChem 2016, 11, 1 – 9
www.chemmedchem.org
8
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

FULL PAPERS
C. Vala, T. J. Morley,* X. Zhang, C. Papin,
A. A. S. Tavares, H. S. Lee,
C. Constantinescu, O. Barret, V. M. Carroll,
R. M. Baldwin, G. D. Tamagnan,
D. Alagille
&& – &&
Brain mapping: We synthesized 16 po-
tential ligands based on the core struc-
ture of the adenosine 2A (A2A) receptor
antagonist preladenant, and screened
them against the A2A receptor for bind-
ing. Of these 16, we identified six for ra-
diolabeling with fluorine-18 or iodine-
123 as potential radiotracers and carried
out an in vivo evaluation in nonhuman
primates, resulting in the selection of
[18F]MNI-444 and [123I]MNI-420 as candi-
dates for measuring A2A receptor densi-
ty in brain in clinical trials.
Synthesis and in vivo Evaluation of
Fluorine-18 and Iodine-123
Pyrazolo[4,3-e]-1,2,4-triazolo[1,5-
c]pyrimidine Derivatives as PET and
SPECT Radiotracers for Mapping A2A
Receptors
ChemMedChem 2016, 11, 1 – 9
www.chemmedchem.org
9
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ