Synthesis and Pharmacological Evaluation of [11C]Granisetron and [18F]Fluoropalonosetron as PET Probes for 5-HT3 Receptor Imaging

Article abstract of DOI:10.1021/acschemneuro.6b00192

Serotonin-gated ionotropic 5-HT₃ receptors are the major pharmacological targets for antiemetic compounds. Furthermore, they have become a focus for the treatment of irritable bowel syndrome (IBS) and there is some evidence that pharmacological modulation of 5-HT₃ receptors might alleviate symptoms of other neurological disorders. Highly selective, high-affinity antagonists, such as granisetron (Kytril) and palonosetron (Aloxi), belong to a family of drugs (the "setrons") that are well established for clinical use. To enable us to better understand the actions of these drugs in vivo, we report the synthesis of 8-fluoropalonosetron (15) that has a binding affinity (K_i = 0.26 ± 0.05 nM) similar to the parent drug (K_i = 0.21 ± 0.03 nM). We radiolabeled 15 by nucleophilic ¹⁸F-fluorination of an unsymmetrical diaryliodonium palonosetron precursor and achieved the radiosynthesis of 1-(methyl-¹¹C)-N-granisetron ([¹¹C]2) through N-alkylation with [¹¹C]CH₃I, respectively. Both compounds [¹⁸F]15 (chemical and radiochemical purity >95%, specific activity 41 GBq/μmol) and [¹¹C]2 (chemical and radiochemical purity ≥99%, specific activity 170 GBq/μmol) were evaluated for their utility as positron emission tomography (PET) probes. Using mouse and rat brain slices, in vitro autoradiography with both [¹⁸F]15 and [¹¹C]2 revealed a heterogeneous and displaceable binding in cortical and hippocampal regions that are known to express 5-HT₃ receptors at significant levels. Subsequent PET experiments suggested that [¹⁸F]15 and [¹¹C]2 are of limited utility for the PET imaging of brain 5-HT₃ receptors in vivo.

Full text of DOI:10.1021/acschemneuro.6b00192

Subscriber access provided by Northern Illinois University
Article
11
Synthesis and Pharmacological Evaluation of [C]granisetron and
18
3
[F]fluoropalonosetron as PET Probes for 5-HT Receptor Imaging
Linjing Mu, Adrienne Müller Herde, Pascal M. Rüefli, Filippo Sladojevich, Selena Milicevic Sephton,
Stefanie D. Krämer, Andrew J Thompson, Roger Schibli, Simon M. Ametamey, and Martin Lochner
ACS Chem. Neurosci., Just Accepted Manuscript • DOI: 10.1021/acschemneuro.6b00192 • Publication Date (Web): 29 Aug 2016
Downloaded from http://pubs.acs.org on September 1, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Chemical Neuroscience is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 43
ACS Chemical Neuroscience
1
2
Synthesis and Pharmacological Evaluation of [11C]granisetron and [18F]fluoropalonosetron as PET
Probes for 5ꢀHT3 Receptor Imaging
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Linjing Mu,† Adrienne Müller Herde,‡ Pascal M. Rüefli,§ Filippo Sladojevich,‡,# Selena Milicevic
Sephton,‡,& Stefanie D. Krämer,‡ Andrew J. Thompson,¶ Roger Schibli,‡ Simon M. Ametamey,‡ and
Martin Lochner*,§,┴
†Department of Nuclear Medicine, University Hospital Zürich, Rämistrasse 100, 8091 Zürich,
Switzerland.
‡Center for Radiopharmaceutical Sciences of ETH, PSI and USZ, VladimirꢀPrelogꢀWeg 4, 8093 Zürich,
Switzerland.
§Department of Chemistry and Biochemistry, University of Bern, Freiestrasse 3, 3012 Bern, Switzerland.
¶Department of Pharmacology, University of Cambridge, Tennis Court Road, Cambridge CB2 1PD, UK.
Abstract
Serotoninꢀgated ionotropic 5ꢀHT3 receptors are the major pharmacological targets for antiemetic
compounds. Furthermore, they have become a focus for the treatment of irritable bowel syndrome (IBS)
and there is some evidence that pharmacological modulation of 5ꢀHT3 receptors might alleviate
symptoms of other neurological disorders. Highly selective, highꢀaffinity antagonists, such as granisetron
(Kytril®) and palonosetron (Aloxi®), belong to a family of drugs (the ‘setrons’) that are well established
for clinical use. To enable us to better understand the actions of these drugs in vivo, we report the
synthesis of 8ꢀfluoropalonosetron (15) that has a binding affinity (Ki = 0.26 ± 0.05 nM) similar to the
-1ꢀ
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 2 of 43
1
2
3
4
5
6
7
8
9
parent drug (Ki = 0.21 ± 0.03 nM). We radiolabeled 15 by nucleophilic 18Fꢀfluorination of an
unsymmetrical diaryliodonium palonosetron precursor and achieved the radiosynthesis of 1ꢀ(methylꢀ11C)ꢀ
Nꢀgranisetron ([11C]2) through Nꢀalkylation with [11C]CH3I, respectively. Both compounds [18F]15
(chemical and radiochemical purity >95%, specific activity 41 GBq/ꢁmol) and [11C]2 (chemical and
radiochemical purity ≥99%, specific activity 170 GBq/ꢁmol) were evaluated for their utility as PET
probes. Using mouse and rat brain slices, in vitro autoradiography with both [18F]15 and [11C]2 revealed a
heterogeneous and displaceable binding in cortical and hippocampal regions that are known to express
5ꢀHT3 receptors at significant levels. Subsequent PET experiments suggested that [18F]15 and [11C]2 are
of limited utility for the PET imaging of brain 5ꢀHT3 receptors in vivo.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Key Words: 5ꢀHT3 receptor; positron emission tomography; granisetron; palonosetron; antagonist;
imaging.
Introduction
Palonosetron (1, Aloxi®, Scheme 2) is a potent second generation 5ꢀHT3 receptor antagonist, and is
the most recently approved drug for preventing acute (0 ꢀ 24 h) and delayed (24 ꢀ 120 h) emesis following
moderate chemotherapy, with improved clinical efficacy when compared to early generation antagonists
such as granisetron (2, Kytril®, Scheme 1).1 In vitro, palonosetron also has a higher binding affinity for 5ꢀ
HT3 receptors (Ki = 40 pM) when compared to other antagonists (e.g. granisetron, Ki = 1.45 nM), and an
extended in vivo plasma elimination halfꢀlife.2ꢀ4 The reasons for these properties are not fully understood,
but proposals have included unique allosteric effects, slow ligand dissociation and receptor
internalization.5,6
5ꢀHT3 receptors belong to a family of receptors that also includes the γꢀamino butyric acid
(GABAA), glycine and nicotinic acetylcholine (nACh) receptors. This is one of the largest families of
-2ꢀ
ACS Paragon Plus Environment

Page 3 of 43
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
ligandꢀgated ion channels and the receptors are responsible for fast synaptic neurotransmission in the
central (CNS) and peripheral nervous systems (PNS). All of the receptors are composed of five subunits
that surround a central ionꢀconducting channel that spans the cellꢀsurface membrane. Agonist activation
of the receptor opens the central pore and allows ions to pass from one side of the membrane to the other,
causing either depolarization or hyperpolarization. The subunits each contain three distinct structural
regions termed the extracellular, transmembrane and intracellular domains. The extracellular domain is
responsible for binding agonists that consequently leads to receptor activation. At the 5ꢀHT3 receptor,
antagonistic drugs compete with serotonin (5ꢀHT) at this orthosteric binding site and prevent the ion
channel from opening. This action is responsible for alleviating the symptoms of nausea and vomiting.1
Antiemetic activity is believed to involve both peripheral receptors and those in the CNS.7ꢀ9 Several
studies have also suggested that 5ꢀHT3 receptors are involved in neurological disorders, such as
depression, drug abuse, schizophrenia, fibromyalgia, pruritus, bulimia nervosa and pain.1,10ꢀ12
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Probes such as radioligands, antibodies and oligonucleotides have been used to map the distribution
of 5ꢀHT3 receptors. First evidence of 5ꢀHT3 receptor expression in the rat brain came from binding studies
using the selective 5ꢀHT3 receptor antagonist [3H]GR65630.13 These and other studies have consistently
found that the highest levels of 5ꢀHT3 receptors in the CNS are located in the dorsal vagal complex in the
brainstem, and in forebrain regions such as the hippocampus, amygdala, nucleus accumbens, putamen and
caudate. The dorsal vagal complex includes the area postrema, nucleus tractus solitarius and the dorsal
motor nucleus of the vagus nerve, regions that are responsible for the vomiting reflex.8 Electrophysiology
provides support for a predominantly presynaptic localization of these receptors, although some
postsynaptic 5ꢀHT3 receptors are also reported.14ꢀ17 Expression in other regions of the brain is usually low
(see ref 18 for a comprehensive review). In the periphery, 5ꢀHT3 receptors are found in vagal afferent
nerve fibers from the heart and GI tract which are responsible for the Bezold–Jarisch reflex and initiation
of vomiting reflex, respectively.19ꢀ21 5ꢀHT3 receptors have also been found in ganglia of the myenteric
plexus and submucosal plexus of the human intestine, consistent with their role in mediating gut
-3ꢀ
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 4 of 43
1
2
3
4
5
6
7
8
9
contractions.22ꢀ24 Elsewhere, lower densities of 5ꢀHT3 receptor can also be found on immune cells and
tissues such as the salivary glands.25,26 Because of their wide tissue distribution, specific PET probes
might be valuable for elucidating the role of 5ꢀHT3 receptors in associated disorders, although relating
symptoms of such complex neuropsychiatric disorders to single neurobiological processes studied by PET
can be challenging.27
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
To this end, the synthesis of several PET ligands that target the 5ꢀHT3 receptor were reported in the
1990s. For instance, 5ꢀHT3 receptor antagonists [11C]MDL 72222,28,29 [11C]Lꢀ683877,30 [11C]Yꢀ25130,31
[11C]YM06031 and [11C]KF1764332 have been synthesized through Nꢀalkylation with [11C]CH3I and their
11
utility assessed in rodents and monkey. Similarly, the clinicallyꢀrelevant antagonist 2 was Cꢀlabeled at
the bicyclic moiety yielding 9ꢀ(methylꢀ11C)ꢀgranisetron33 but to the best of our knowledge neither in vitro
nor in vivo evaluation of this PET tracer was published. Other groups have radiolabeled the nonꢀselective
5ꢀHT3 receptor agonist Nꢀmethylquipazine (NMQ, 2ꢀ(4ꢀmethylpiperazinꢀ1ꢀyl)quinoline), or partial
agonists S21007 and MR18445, and evaluated the resulting tracers [11C]NMQ,34 [11C]S2100735,36 and
[18F]MR1844537 in rat and monkey. However, all of these 5ꢀHT3 receptor PET probes were unsuitable as
imaging agents, owing to high nonꢀspecific binding28,29,32,34,36,37 or low brain uptake.31 More recently,
Zheng and colleagues have 11Cꢀlabeled a series of benzoxazole partial 5ꢀHT3 receptor agonists38 but no
biological study with these PET tracers has been reported to date. While writing this manuscript,
Mukherjee and colleagues also reported the synthesis of novel 5ꢀHT3 receptor ligand (S)ꢀ[18F]fesetron, but
found that the very low brain uptake in rats impairs its utility as a PET tracer.39 Consequently, in vivo
imaging of 5ꢀHT3 receptors in the CNS and PNS by PET remains to be accomplished.
Here, we report the synthesis of 1ꢀ(methylꢀ11C)ꢀNꢀgranisetron ([11C]2) and 18Fꢀlabeled
fluoropalonosetron ([18F]15) and use these to characterize the distribution of 5ꢀHT3 receptors in vitro and
in vivo in rodents.
-4ꢀ
ACS Paragon Plus Environment

Page 5 of 43
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
-5ꢀ
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 6 of 43
1
2
3
4
Results and Discussion
5
6
7
Synthesis of 5-HT3 Receptor Ligands and Precursors for Radiolabeling
8
9
The precursor 3 used for radiolabeling granisetron and unlabeled reference 2 were synthesized as
shown in Scheme 1. Amide 3 was obtained in good yield following our previously published protocol40
and subsequent selective N1ꢀmethylation was achieved by treatment with dimethyl carbonate and
potassium carbonate in refluxing DMF.41
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Synthesis of nonꢀlabeled 5ꢀHT3 receptor ligands granisetron (2), quinolineꢀ4ꢀcarboxylate 6,
and labeling precursor 3. Reagents and conditions: (a) 5, DCC, HOBt, CH2Cl2/DMF, rt, 73%. (b) K2CO3,
dimethyl carbonate, DMF, reflux, 89%. (c) Ac2O, reflux; NaOH, H2O, reflux. (d) POCl3, reflux, 98%. (e)
SOCl2, reflux; Et3N, EtOH, rt, 89%. (f) 1ꢀmethylpiperazine, 140°C, 69%.
In a separate effort to develop novel 5ꢀHT3 receptor ligands based on NMQ, quinoline carboxylate
6 was discovered to bind with high affinity (Ki = 1.9 ± 1.0 nM) and compete for the same binding site as
-6ꢀ
ACS Paragon Plus Environment

Page 7 of 43
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
[3H]granisetron ([3H]2).42 We therefore decided to use structurally different 6 for blocking studies with
the PET tracers (vide infra). It was synthesized starting from fluoroisatin 7 according to the route
developed by Cappelli and colleagues for similar compounds (Scheme 1).43 Briefly, 7 was ringꢀexpanded
via known Pfitzinger reaction to give oxoquinoline carboxylic acid 8,44 which was in turn chlorinated and
esterified to yield 2ꢀchloroquinoline carboxylate 9. The piperazine substituent was subsequently
introduced via SNAr reaction and final quinoline carboxylate 6 was isolated in fairly good yield.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Palonosetron (1) was synthesized according to the published route,45 with some minor experimental
modifications (Scheme 2).
Scheme 2. Synthesis of reference compounds palonosetron (1), cold fluoropalonosetron 15, and labeling
precursors 14aꢀc. Reagents and conditions: (a) (S)ꢀ3ꢀAminoquinuclidine hydrochloride, EDC, HOBt,
iꢀPr2EtN, DMF, rt, 69%. (b) nꢀBuLi (2 equiv.), THF, ꢀ23°C; DMF, ꢀ23°C; HCl, H2O, rt, 74%. (c) H2 (25
bar), Pd(OH)2/C (20 mol%), HClO4, HOAc, 70°C, 90%, dr 46:54; 4M HCl in dioxane/CH2Cl2;
recrystallization from iꢀPrOH/H2O, 27% (S,S)ꢀ1·HCl; sat. aq. NaHCO3/EtOAc, 96% (d) NIS, CF3SO3H,
rt, 41%. (e) pꢀTsOH·H2O, mCPBA; anisole or mesitylene, CH2Cl2/TFE, rt; anion exchange, 23ꢀ67%. (f)
Method A: 14a, CsF, MeCN, 100°C, 14%. Method B: 14c, Cu(OTf)2, KF, 18ꢀcrownꢀ6, DMF, 60°C, 13%.
-7ꢀ
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 8 of 43
1
2
3
4
5
6
7
8
9
Tetrahydronaphthalene carboxylic acid 10 was first coupled with chiral (S)ꢀ3ꢀaminoquinuclidine mediated
by EDC and HOBt in DMF. This procedure afforded amide 11 in much higher yields than the reported
coupling via acid chloride, presumably due to the low solubility of the used quinuclidine hydrochloride in
apolar organic solvents. Closure of the third ring to give isoquinolinone 12 was achieved by sequential
lateral Cꢀ8 lithiation, formylation with DMF and intramolecular enamine formation. It is important to
monitor the latter acidꢀmediated enamine formation and run it to completion, as too short reaction times
will produce substantial amounts of cyclic hemiaminal (carbinolamine) which can be separated and
dehydrated separately. The double bond character of the amide bond (lactam) in 12 imparts aromatic
character to the ring system and hence the reduction of the enamine double bond requires fairly harsh
conditions. In the event, high pressure of hydrogen in the presence of a palladium catalyst and strong
acidic conditions at high temperature45 reduced the double efficiently (90% yield) but with no
diastereoselectivity (dr 46:54 (S,S)ꢀ1:(R,S)ꢀ1). In order to improve the diastereoselectivity in favor of the
desired (S,S)ꢀ1 isomer we briefly explored a more recent protocol published in a patent46 (H2 (10 bar), 20
mol% Pd(OH)2/C, 58ꢀ62°C, EtOAc) that reported a higher diastereoselectivity (dr 6:4). However, in our
hands, applying this procedure resulted in very slow conversion and no diastereoselectivity (dr 1:1). The
two diastereomers of 1 were transformed into hydrochloride salts and separated by recrystallization which
yielded pure (S,S)ꢀ1 in moderate yield.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For the introduction of the 18Fꢀlabel we envisaged to use nucleophilic substitution of an
unsymmetrical diaryliodonium salt of 1 with [18F]fluoride.47ꢀ52 To this end, we searched for methods to
iodinate palonosetron (1). Direct iodination of nonꢀactivated aromatic rings is far less common than the
corresponding chlorination or bromination. After exploring the few published protocols, the method
described by Olah and coꢀworkers, that utilizes in situ generated superelectrophilic iodine(I)
trifluoromethanesulfonate,53 worked best for us. This gave iodopalonosetron 13 as single regioisomer in
acceptable yield. For the projected formation of diaryliodonium salts from 13 we initially employed the
published oneꢀpot procedures developed by Bielawski and Olofsson,54 which was later refined by Zhu et
-8ꢀ
ACS Paragon Plus Environment

Page 9 of 43
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
al..55 We thought it more prudent to resort to a stepwise addition, as we found that 13 is readily
overoxidized by excess oxidant (mCPBA), most probably to the corresponding Nꢀoxide. In the reaction,
mCPBA and pꢀtoluenesulfonic acid were added to 13 and complete conversion to the corresponding bisꢀ
tosyloxyiodoaryl species S1 (Scheme S1) was closely monitored by TLC, after which the aromatic
compound was added to produce the paraꢀsubstituted diaryliodonium salts 14 (for experimental details
see Methods). Previous model studies indicate that nucleophilic fluorination of unsymmetrical
diaryliodonium salts preferentially leads to 18Fꢀincorporation at the less electronꢀrich aromatic ring.47
What is more, the counteranion of the diaryliodonium salt was found to significantly influence
radiochemical yield (RCY).56,57 Therefore, we chose to synthesize anisylꢀ and mesitylpalonosetron
iodonium salts, in order to direct 18Fꢀfluorination towards the aromatic ring of palonosetron, and prepared
14aꢀc with different counteranions. Isolation of pure and defined diaryliodonium salts is not trivial as
handling of such compounds during workup and column chromatography leads to partial counterion
exchange that is very difficult to analyze. Therefore, we utilized a reversedꢀphase HPLC ionꢀexchange
protocol57 to ensure generation of defined iodonium salts.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
We used diaryliodonium salts 14 to access the standard reference fluoropalonosetron 15 for
pharmacological characterization and confirming the identity of radiolabeled product by HPLC. To this
end, anisylpalonosetron iodonium tosylate 14a was heated in the presence of CsF in MeCN.58 The
fluorination reaction was very slow and did not go to completion after 51 h at 100°C, yet we were able to
isolate pure 15 after preparative HPLC in low yield. Analysis of the reaction mixture by HPLC revealed
that most of the starting diaryliodonium 14a was either fluorinated at the anisole moiety (giving 13) or
thermally decomposed to palonosetron (product ratio 15/13/1 20:39:41). With the aim of improving this
product ratio and overall yield of 15 we also explored a copper(II) triflateꢀcatalyzed fluorination reaction
of 14c with naked fluoride that relies on using the mesityleneꢀring as dispensable aryl moiety.59 Although
we did observe complete disappearance of 14c after 24 h at 60°C and obtained a much improved product
-9ꢀ
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 10 of 43
1
2
3
4
ratio (15/13/1 60:13:27), isolation of pure 15 from this mixture by preparative HPLC proved difficult due
5
6
to the structural similarity of the components.
7
The binding affinity of 15, in competition with [3H]granisetron ([3H]2), was measured in HEK 293
cells stably expressing the human 5ꢀHT3A receptor and gave a Ki value of 0.26 ± 0.05 nM (n = 4), making
15 a promising candidate for PET imaging of 5ꢀHT3 receptors. The parent compound palonosetron (1)
exhibited a similar binding affinity (Ki = 0.21 ± 0.03 nM, n = 5) in the same assay (Figure S1) and as
reported elsewhere,2 indicating that addition of a small fluorine atom did not significantly affect its
binding affinity.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Radiolabeling
The radiosynthesis of 1ꢀ(methylꢀ11C)ꢀNꢀgranisetron ([11C]2) was achieved in a oneꢀstep procedure
by Nꢀmethylation with [11C]CH3I in up to 30% RCY (decay corrected to EOB) (Scheme 3). The crude
product was purified by semiꢀpreparative HPLC and isolated in ≥99% radiochemical and chemical purity.
The specific activity was in the range of 170 ± 26 GBq/ꢁmol with a total activity of 3.1 ± 1.0 GBq at the
end of the synthesis (n = 11). Overall, the radiosynthesis of [11C]2 took approximately 40 min from end of
bombardment. For all in vitro and in vivo experiments [11C]2 was formulated in 5% ethanol in PBS
buffer.
The diaryliodonium salts 14aꢀc were 18Fꢀfluorinated using cesium [18F]fluoride/kryptofix (K222)
complex in DMF in the presence of TEMPO (Scheme 3). The radiolabeling was performed in a
homemade semiꢀautomated, shielded hotꢀcell. Under the same reaction conditions, we found that both
iodoanisole (14a, 14b) and iodomesitylene (14c) leaving groups were displaced by [18F]fluoride in the
presence of Cs2CO3 and TEMPO in DMF at 140°C after 12 min. In our study, the anisyl iodonium salts
(14a and 14b) gave slightly higher RCYs than the mesityl iodonium salt (14c) (RCY 6% vs. 4%). For the
anisyl iodonium salts, the influence of the counteranion on RCY was not significant (data not shown).
-10ꢀ
ACS Paragon Plus Environment

Page 11 of 43
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
Furthermore, presence of trace amounts of water in the solvent (DMF), as reported by others,57,60 did not
increase RCYs. Finally, using diaryliodonium precursor 14a, [18F]fluoropalonosetron ([18F]15) could be
synthesized and formulated for in vivo applications in approximately 6% RCY (decay corrected to EOB, n
> 10) and with specific activities at around 41 GBq/ꢁmol at the end of the synthesis. We think that the
rather low specific activity is probably due to the low radiolabeling yield. The radiochemical purity of
[18F]15 was > 95% as determined by HPLC.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 3. Radiosynthesis of [11C]2 and [18F]15.
When analyzing the [18F]fluoride incorporation efficiency by RPꢀHPLC the influence of the pH of
the mobile phase was very noticeable. For instance, in the case of the 18Fꢀfluorination reaction with
precursor 14b we measured 78% [18F]fluoride incorporation by RPꢀHPLC when using 0.1% aq. TFA in
the eluent (pH = 2). Analysis of the same reaction using 10 mM aq. NH4HCO3 buffer in the eluent (pH =
7.7) gave a significantly lower [18F]fluoride incorporation value (< 3%). Similar observations were
recently reported by Ory and colleagues who systematically studied the retention of [18F]fluoride on
-11ꢀ
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 12 of 43
1
2
3
4
5
6
7
8
9
different RPꢀHPLC columns that were eluted with mobile phases of different compositions and pH.61
They observed increased [18F]fluoride recovery with increasing pH. We reason that hydrogen fluoride
(HF), due to its pKa value of 3.19,62 is almost completely dissociated at pH 5 and therefore the more polar
[18F]fluoride is not retained on the RPꢀcolumn. At pH 2, however, about 90% of [18F]fluoride is present as
[18F]HF which may have stronger interactions with the silica core of RPꢀHPLC columns. Based on these
observations, we would recommend to use an eluent with pH > 5 for silica based C18 RPꢀHPLC columns
for both analytical and semiꢀpreparative RPꢀHPLC of 18Fꢀlabeled compounds.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In vitro Studies with [11C]granisetron and [18F]fluoropalonosetron
The stability of [11C]2 was assessed in rodent and human plasma over a period of 20 min during
which no radioactive degradation products were detected (Figure S2). We determined a logD value of
0.43 ± 0.05 (n = 5) at physiological pH (7.4) for [11C]2 and of 0.93 ± 013 (n = 3) for [18F]15. Both PET
tracers have a basic tertiary amine in their bicyclic moiety and we believe the rather low logD values are
likely to be due to protonation at physiological pH (calculated pKa values for tertiary amines: 9.3 for 2
and 8.7 for 15). The calculated lipophilicity (clogP) of the neutral species was 1.7 and 2.4, respectively.
Radiotracer binding was evaluated by in vitro autoradiography using mouse and rat brain slices as
shown in Figure 1. Binding of [11C]2 was highest at hippocampus, cortical regions and septal area in both
mouse and rat which is consistent with previous studies using other 11Cꢀlabeled 5ꢀHT3 receptor
ligands.28,32,36 Binding of [11C]2 to these brain regions was specific as coꢀincubation with cold competitive
ligands palonosetron (1, Ki = 0.21 ± 0.03 nM) or quinolineꢀ4ꢀcarboxylate 6 (Ki = 1.9 ± 1.0 nM)
completely diminished the radioactive signal (Figure 1A).
The binding of [18F]15 to mouse brain sections resulted in a heterogeneous radioactivity pattern
with signals corresponding to hippocampal and cortical regions, and septal area. Blocking experiments
using mouse brain slices that were coꢀincubated with [18F]15 and excess 5ꢀHT3 receptor competitive
-12ꢀ
ACS Paragon Plus Environment

Page 13 of 43
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
antagonists palonosetron (1) or quinolineꢀ4ꢀcarboxylate 6 resulted in complete loss of all these signals
(Figure 1B, top panels). In rat brain sections incubated with [18F]15, high signal intensity was observed in
hippocampal regions, which was almost completely abolished by blocking conditions with cold 1 or 6
(Figure 1B, bottom panels) indicating specific binding. This is in line with recent findings where ex vivo
microPET images of whole rat brain also showed strong hippocampal staining with the palonosetron
congener (S)ꢀ[18F]fesetron.39 Specific binding to hippocampus was also observed with high affinity 5ꢀHT3
antagonist [3H]GR65630 (Bmax 4.9 ± 0.6 and 1.3 ± 0.3 fmol/mg protein for rat and mouse, respectively).63
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Representative autoradiography images of coronal mouse or rat brain sections after incubation
with (A) [11C]granisetron ([11C]2, 0.6 nM) or (B) [18F]fluoropalonosetron ([18F]15, 0.4 nM) under baseline
(vehicle) or blocking conditions (coꢀincubation with palonosetron (1, 1 ꢁM) or quinolineꢀ4ꢀcarboxylate 6
-13ꢀ
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 14 of 43
1
2
3
4
5
6
7
8
(1 ꢁM)). Autoradiograms are superimposed onto morphological hematoxylinꢀeosin (HE)ꢀstained slices
with a color scale for low (blue) to high (red) binding. Identification of specific brain regions was based
on HE staining and according to the anatomical atlas.64 Scale bars: 2 mm.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
We attribute the different radioactive labeling patterns of [11C]2 and [18F]15 to the difference of
mouse and rat brain sections, as indicated by hematoxylinꢀeosin (HE) staining, as well as their difference
in chemical structures and specific radioactivities (137ꢀ189 GBq/ꢁmol and 41 GBq/ꢁmol, respectively).
The specific labeling of distinct rodent brainꢀregions (e.g. hippocampus, cortical regions, septal area)
encouraged us to evaluate both [11C]2 and [18F]15 as PET tracers in in vivo experiments.
In vivo Studies using [11C]granisetron and [18F]fluoropalonosetron
In vivo brainꢀPET scans were performed in rats to study the uptake of [11C]2 and [18F]15 under
baseline and blocking conditions with palonosetron (1). For both tracers, we found very low and
homogeneous radiotracer accumulation in the brain (Figure 2A). Further, from these PET scans, we
calculated the timeꢀactivity curves (TACs, Figure 2B) for the brain regions that have been previously
identified to contain 5ꢀHT3 receptors.13 Specifically, this earlier study used highꢀaffinity antagonist
[3H]GR65630 on rat brain homogenates which gave the highest specific binding in the entorhinal cortex
and area postrema, the latter of which is known to control vomiting. In the MRIꢀtemplate implemented in
our biomedical imaging software, the area postrema it is part of the medulla oblongata. No specific
binding of either [11C]2 or [18F]15 to the aforementioned brain region was detected (Figure 2B).
Furthermore, while the most intense signal was observed in the pituitary gland, uptake of [11C]2 and
[18F]15 was not blocked by 1 and indicates that the signal was nonꢀspecific. In line with our findings, low
brain uptake and strong nonꢀspecific accumulation of activity in pituitary gland was reported in a recent
SPECT imaging study of the serotoninergic system in rat (targeting 5ꢀHT1A, 5ꢀHT2A and 5ꢀHT4 receptors,
and serotonin transporter SERT).65 The absence of specific binding of our 5ꢀHT3 receptor radioligands
-14ꢀ
ACS Paragon Plus Environment

Page 15 of 43
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
might be due to the low receptor number (Bmax), which was reported to be 1ꢀ11 fmol/mg protein in rat and
mouse brain when using highꢀaffinity 5ꢀHT3 receptor antagonist [3H]GR65630.63,66 Such low Bmax values
make brain 5ꢀHT3 receptors a highly challenging target for PET imaging. Furthermore, the partition
coefficients (logD) of both [11C]2 and [18F]15 are lower than the optimal average lipophilicity value of 2.1
for brain penetrating compounds.67
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A
B
-15ꢀ
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 16 of 43
1
2
Figure 2. (A) PET brain images of Wistar rats using [11C]granisetron ([11C]2, top panels) or
[18F]fluoropalonosetron ([18F]15, bottom panels) under baseline or blocking conditions (100 ꢁg/kg 1).
Images are averaged 0ꢀ60 min postꢀradiotracer injection and superimposed on an MRI template, showing
axial, sagittal and coronal sections (from left to right). SUVmax = 1.5 (PET/CT images) or 0.8 (PET/MRI
images). (B) TACs of selected rat brain regions obtained with [11C]2 (top graphs) or [18F]15 (bottom
graphs) under baseline or blocking conditions. TACs were generated from the same rats as shown in (A).
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5ꢀHT3 receptors are not only expressed in the CNS but also in the periphery, most prominently on
enterochromaffin cells of the GI tract. BodyꢀPET scans of rats using either [11C]2 or [18F]15 revealed
significant uptake in liver and intestines for both tracers (Figure 3). At late time points, uptake in stomach
wall was also observed, in particular for [18F]15. Furthermore, we found marked accumulation of [18F]15
in lungs within the first few minutes after injection. This is similar to other reports where an accumulation
of PET signal in this organ was detected with 11Cꢀlabeled 5ꢀHT3 receptor antagonists [11C]MDL 72222,28
[11C]Yꢀ25130,31 [11C]YM06031 and [11C]KF1764332 in rodents.
-16ꢀ
ACS Paragon Plus Environment

Page 17 of 43
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
Figure 3. Coronal PET images of [11C]granisetron ([11C]2, top panels) or [18F]fluoropalonosetron
([18F]15, bottom panels) in the lungꢀliverꢀintestine region in rat. Time series images averaged over the
indicated time (in min) postꢀinjection (p.i.). Labels: Lu, lung; Li, liver; St, stomach; Int, intestine.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Conclusions
We have successfully synthesized and radiolabeled the clinicallyꢀrelevant 5ꢀHT3 receptor
antagonists granisetron (2) and palonosetron (1). In vitro autoradiography using 1ꢀ(methylꢀ11C)ꢀNꢀ
granisetron ([11C]2) and [18F]fluoropalonosetron ([18F]15) showed specific labeling in rodent brain slices
that were consistent with a previous study that used tritiated 5ꢀHT3 receptor antagonist [3H]GR65630.13
We found that in vivo PET studies in rats gave low and homogenous radiotracer signals of [11C]2 and
[18F]15 in brain. We speculate that this might be due to 1) the low expression of 5ꢀHT3 receptors in brain
(Bmax 1ꢀ11 fmol/mg protein) and 2) the basic tertiary amino groups of [11C]2 and [18F]15 that in turn leads
to 3) low logD7.4 values for both tracers which is at the lower end of the optimal range (0ꢀ3) for passing
through the bloodꢀbrain barrier. Although 1 and 2 are very efficient drugs, we conclude that [11C]2 and
[18F]15 are not suitable radioligands for in vivo PET imaging of 5ꢀHT3 receptors in rodents, and in
general, developing specific, highꢀbrain uptake 5ꢀHT3 receptor PET tracers remains a challenge.
Methods
General Chemistry. All reagents and solvents were purchased from SigmaꢀAldrich Chemie
GmbH (Germany), Merck (Germany), Acros (Belgium), ABCR GmbH (Germany) or Fluka
(Switzerland). All chemicals were used as supplied without further purification, except mCPBA which
was dried at high vacuum for at least 1 h before use. Anhydrous solvents were obtained by filtration
through dry alumina columns under argon pressure. Synthesized compounds were >95% in purity
determined by analytical HPLC method on an Agilent HPLC system. Reactions were monitored by TLC
on 0.2 mm ALUGRAM Xtra SIL G/UV or POLYRAM Alox N/UV plates (MacheryꢀNagel). Flash
-17ꢀ
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 18 of 43
1
2
3
4
5
6
7
8
9
column chromatography was carried out using Fluka analytical 60752 silica gel (pore size 60 Å, 40ꢀ63
ꢁm particle size) or CAMAG 507ꢀC neutral aluminum oxide (pore size 60 Å, 40ꢀ160 ꢁm particle size).
Infrared (IR) spectra were recorded on a Jasco FTꢀIRꢀ460 Plus. NMR spectra were recorded on a Bruker
Avance 300 (1H at 300 MHz, 13C at 75 MHz) and Avance II 400 (1H at 400 MHz, 13C at 100 MHz, 19F at
376 MHz). Chemical shifts (δ) are reported in ppm referenced to the residual solvent peak. Multiplicities:
s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet, br = broad peak. Coupling constants (J) are
reported in Hz. High resolution mass spectrometry (HRMS) was performed on a Thermo Scientific LTQ
Orbitrap XL with a Nanospray Ionsource. For purification of radiolabeled product, a semiꢀpreparative
MerckꢀHitachi L2130 HPLC system equipped with a radiation detector VRM 202 (Veenstra Instrument,
Joure, Netherlands) was used with a Waters Symmetry C8 Prep Column (50 x 7.8 mm, 100 Å, 5 µm) and
a gradient method: 10 mM ammonium bicarbonate buffer (solvent A) and CH3CN (solvent B) 0.0ꢀ1.0 min
5% B; 1.0ꢀ12.0 min, 5ꢀ50% B; 13.0ꢀ20.0 min, 50% B; 20.0ꢀ20.1 min, 50ꢀ5% B; 20.1ꢀ40.0 min, 5% B at a
flow rate of 4 mL/min. For product analysis, an analytical Agilent 1100 series HPLC system, equipped
with UV multiꢀwavelength detector and a GabiStar radiodetector (Raytest) was used with an ACE C18ꢀ
AR column (50 x 4.6 mm, 3 µm) with the following conditions: 0.1% TFA in H2O (solvent A) or 10 mM
ammonium bicarbonate buffer (solvent A), CH3CN (solvent B); 0.0ꢀ8.0 min, 5ꢀ50% B; 8.0ꢀ9.0 min, 50%
B; 9.0ꢀ9.1 min, 50ꢀ5% B; 9.1ꢀ12.0 min, 5% B at a flow rate of 1 mL/min.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For the preparation of reference compound 2 and blocker compound 6 see the Supporting
Information. Labeling precursor 3 was synthesized as published elsewhere.40
Nꢀ((1R,3S,4R)ꢀQuinuclidinꢀ3ꢀyl)ꢀ5,6,7,8ꢀtetrahydronaphthaleneꢀ1ꢀcarboxamide (11). We have
modified the previously reported synthesis45 of this compound. To a suspension of 5,6,7,8ꢀ
tetrahydronaphthaleneꢀ1ꢀcarboxylic acid (10, 3.98 g, 22.6 mmol), (S)ꢀ3ꢀaminoquinuclidine
dihydrochloride (3.00 g, 15 mmol) and HOBt monohydrate (3.46 g, 22.6 mmol) in 35 mL DMF, was
added DIPEA (10.5 mL, 60 mmol) and EDC (5 mL, 28.7 mmol). The reaction mixture was stirred at rt
for 27 h and the reaction quenched by the addition of 100 mL 5M NaOH solution. The water phase was
-18ꢀ
ACS Paragon Plus Environment

Page 19 of 43
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
extracted EtOAc (3 x 150 mL), the combined organic layers were washed with 350 mL of 2M NaOH,
dried over Na2SO4, filtered and concentrated under reduced pressure to afford a yellow oil. EtOAc (70
mL) was added and the organic phase was washed again with 2M NaOH (2 x 70 mL), concentrated and
the resulting residue coꢀevaporated with CH2Cl2 (2 x 20 mL), giving an oily foamy solid (5.33 g). The
crude product was purified by flash chromatography on silica gel, using CH2Cl2/MeOH /Et3N 9:1:0.1, to
afford a white foamy solid (3.64 g). In order to remove Et3Nꢀcontaminations, the compound obtained
after flash chromatography was partitioned between 100 mL CH2Cl2 and 100 mL sat. aq. NaHCO3
solution. The water phase was extracted with CH2Cl2 (2 x 100 mL). The combined organic layers were
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
dried over Na2SO4 and concentrated to yield 11 as a white foamy solid (2.95 g, 69%). H NMR (300
MHz, DMSOꢀd6) δ 8.26 (d, J = 6.9, 1H), 7.17 – 7.01 (m, 3H), 3.98 – 3.81 (m, 1H), 3.08 (ddd, J = 13.4,
9.7, 2.0, 1H), 2.86 – 2.63 (m, 8H), 2.57 (dd, J = 14.2, 5.5, 1H), 1.86 (dd, J = 6.1, 3.0, 1H), 1.79 – 1.66 (m,
5H), 1.63 – 1.50 (m, 2H), 1.36 – 1.27 (m, 1H). HRMS (ESI+) calcd for C18H25N2O [M+H]+ 285.1960,
found 285.1963.
2ꢀ((1R,3S,4R)ꢀQuinuclidinꢀ3ꢀyl)ꢀ2,4,5,6ꢀtetrahydroꢀ1Hꢀbenzo[de]isoquinolinꢀ1ꢀone (12). This
was synthesized according to the original procedure,45 with minor experimental modifications. The
reaction was run in dry glassware under an atmosphere of argon. To a suspension of amide 11 (3.00 g,
10.55 mmol) in 33 mL anhydrous THF at ꢀ23 °C (internal temperature), a solution of nꢀBuLi in hexane
2.5M (10.6 mL, 26.5 mmol) was added slowly in portions, over a time period of 1.5 h. The temperature
was maintained between ꢀ14 °C and ꢀ23 °C during the addition. DMF (1.7 mL, 22.58 mmol) was then
added in portions over a time period of 2 h. 9 mL 6M aq. HCl was added slowly and the resulting mixture
was stirred vigorously for 15 min. The twoꢀphase system was concentrated in vacuo to remove most of
the THF and hexane. After addition of 40 mL 5M NaOH, the water phase was extracted with EtOAc (3 x
120 mL). The combined organic layers were dried over Na2SO4, filtered and concentrated in vacuo to
afford a yellow sticky solid (3.13 g). The crude product was purified by flash chromatography on silica
gel (CH2Cl2/ MeOH/Et3N 9:1:0.1) to give 12 as an off white solid (2.30 g, 74%). IR (neat): 2930, 2868,
-19ꢀ
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 20 of 43
1
2
3
4
5
6
7
8
9
1
1651, 1614, 1594, 1454, 1321, 1262, 1173, 1055, 978, 822, 766 cmꢀ1. H NMR (300 MHz, DMSOꢀd6) δ
8.05 (dd, J = 7.9, 1.1, 1H), 7.48 (dd, J = 7.2, 0.9, 1H), 7.44 (s, 1H), 7.39 (m, 1H), 5.01 – 4.92 (pseudo t,
1H), 3.24 (dd, J = 10.4, 1.7, 1H), 3.15 – 3.02 (m, 2H), 2.93 (t, J = 6.0, 2H), 2.85 – 2.70 (m, 5H), 1.95 –
1.82 (m, 3H), 1.78 – 1.57 (m, 3H), 1.46 – 1.33 (m, 1H). HRMS (ESI+) calcd for C19H23N2O [M+H]+
295.1805, found 295.1796.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(S)ꢀ2ꢀ((1R,3S,4R)ꢀQuinuclidinꢀ3ꢀyl)ꢀ2,3,3a,4,5,6ꢀhexahydroꢀ1Hꢀbenzo[de]isoquinolinꢀ1ꢀone
hydrochloride (palonosetron hydrochloride, 1·HCl). This was synthesized according to the original
procedure,45 with some experimental modifications. The reaction was performed in a batch autoclave with
a vacuum trap inserted as reaction vessel. To a solution of 12 (2.03 g, 6.9 mmol) in 30 mL glacial acetic
acid, was added HClO4 70% (0.5 mL), Pd(OH)2/C 20% (40% wet, 838 mg, 0.71 mmol, 11 n/n%) and the
autoclave was closed. After flushing the batch hydrogenator three to five times with approximately 10 bar
H2, hydrogen pressure was built up to 25 bar and the reaction mixture was stirred at 70 °C for 24 h. The
instrument was allowed to cool down to rt and the pressure released. The reaction mixture was filtered
through a plug of celite (preꢀwashed with MeOH and acetone) and the filter cake washed with MeOH.
The solvents were removed under reduced pressure and the resulting mixture was partitioned between 10
mL CH2Cl2 and 80 mL 5M NaOH. The water phase was extracted with CH2Cl2 (1 x 50 mL and 2 x 20
mL), the combined organic layers dried over Na2SO4, filtered and concentrated to afford the crude free
base as a white foamy solid (1.82 g). To a solution of the crude free base in 4 mL CH2Cl2, was added 4
mL 4M HCl in dioxane. The mixture was stirred for 15 min and concentrated in vacuum to afford the
crude hydrochloride as a white solid (2.22 g). The crude HCl salt was coꢀevaporated once with 10 mL iꢀ
PrOH and then recrystallized from iꢀPrOH/H2O (15 mL/0.46 mL, water was added at reflux). Upon
cooling, the desired (S,S)ꢀdiastereomer crystallized out. The crystals were filtered, washed with cold
mother liquor and 5 mL cold iꢀPrOH to afford 1·HCl as a white solid (427 mg, 19%). A second
recrystallization gave another 178 mg (8%). IR (neat): 2962, 2940, 2923, 2875, 2862, 2461, 1644, 1590,
1408, 1302, 1157, 972, 816, 768 cmꢀ1. 1H NMR (300 MHz, DMSOꢀd6) δ 10.39 (s, 1H), 7.70 (dd, J = 6.9,
-20ꢀ
ACS Paragon Plus Environment

Page 21 of 43
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
1.9, 1H), 7.34 – 7.20 (m, 2H), 4.76 (pseudo t, 1H), 3.77 (dd, J = 11.7, 4.7, 1H), 3.65 (br t, J = 11.7, 1H),
3.47 (dd, J = 13.3, 7.5, 2H), 3.29 – 3.15 (m, 4H), 3.08 – 2.94 (m, 1H), 2.90 – 2.67 (m, 2H), 2.24 – 2.16
(m, 1H), 2.12 – 1.60 (m, 7H), 1.37 – 1.21 (m, 1H). HRMS (ESI+) calcd for C19H25N2O [MꢀHCl+H]+
297.1961, found 297.1967.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(S)ꢀ2ꢀ((1R,3S,4R)ꢀQuinuclidinꢀ3ꢀyl)ꢀ2,3,3a,4,5,6ꢀhexahydroꢀ1Hꢀbenzo[de]isoquinolinꢀ1ꢀone
(palonosetron free base, 1). Palonosetron hydrochloride 1·HCl (196 mg, 0.59 mmol) was partitioned
between 15 mL sat. aq. NaHCO3 solution and 30 mL EtOAc. The water phase was extracted with EtOAc
(4 x 15 mL). The combined organic layers were dried over Na2SO4, filtered and concentrated in vacuum
1
to afford 1 as a foamy sticky solid (168 mg, 96%). H NMR (300 MHz, DMSOꢀd6) δ 7.67 (dd, J = 5.8,
3.1, 1H), 7.29 – 7.21 (m, 2H), 4.66 – 4.56 (pseudo t, 1H), 3.78 (dd, J = 12.0, 4.7, 1H), 3.25 – 3.04 (m,
2H), 3.02 – 2.84 (m, 3H), 2.83 – 2.67 (m, 5H), 2.11 –1.94 (m, 2H), 1.84 – 1.76 (m, 1H), 1.75 – 1.54 (m,
4H), 1.46 – 1.35 (m, 1H), 1.32 –1.24 (m, 1H).
(S)ꢀ8ꢀIodoꢀ2ꢀ((1R,3S,4R)ꢀquinuclidinꢀ3ꢀyl)ꢀ2,3,3a,4,5,6ꢀhexahydroꢀ1Hꢀbenzo[de]isoquinolinꢀ1ꢀ
one (13). To a solution of 1 (63 mg, 0.21 mmol) in trifluoromethanesulfonic acid (0.13 mL, 1.47 mmol),
was added Nꢀiodosuccinimide (287 mg, 1.28 mmol) in portions of 57 mg over a time period of 27 h. After
stirring the reaction mixture for a total of 29 h, the reaction was quenched by addition of 10 mL 2M
NaOH. The water phase was extracted with CH2Cl2 (3 x 10 mL). The combined organic layers were dried
over Na2SO4, filtered and concentrated in vacuum to afford a brownish solid (71 mg). The crude product
was purified by flash chromatography on silica gel (CH2Cl2/MeOH/NH4OH (25% in H2O) 9:1:0.05) to
give iodide 13 as a yellowish solid (37 mg, 41%). IR (neat): 2925, 2862, 1633, 1575, 1429, 1316, 1277,
1170, 1033, 975, 820, 779 cmꢀ1. 1H NMR (300 MHz, DMSOꢀd6) δ 7.94 (d, J = 1.6, 1H), 7.66 (d, J = 1.7,
1H), 4.66 – 4.54 (pseudo t, 1H), 3.78 (dd, J = 12.2, 4.7, 1H), 3.22 – 3.14 (m, 2H), 3.08 – 2.87 (m, 3H),
2.85 – 2.68 (m, 5H), 2.09 – 1.92 (m, 2H), 1.89 – 1.82 (m, 1H), 1.78 – 1.53 (m, 4H), 1.50 – 1.38 (m, 1H),
1.28 – 1.23 (m, 1H). 13C NMR (100 MHz, DMSOꢀd6) δ 163.8, 139.4, 138.7, 137.9, 133.7, 133.1, 92.2,
-21ꢀ
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 22 of 43
1
2
3
4
5
6
50.3, 49.56, 46.7, 46.2, 46.0, 33.9, 27.1, 27.0, 25.5, 24.8, 20.9, 20.7. HRMS (ESI+) calcd for C19H24IN2O
[M+H]+ 423.0928, found 423.0919.
7
8
9
(4ꢀMethoxyphenyl)(palonosetron)iodonium ditosylate (14a). To a solution of iodide 13 (37 mg,
0.088 mmol) in CH2Cl2/2,2,2ꢀtrifluoroethanol 1:1 (0.34 mL), was added pꢀTsOH·H2O (35 mg, 0.18
mmol). After pꢀTsOH·H2O was dissolved, dried mCPBA (commercial grade 77%, 23.5 mg, 0.11 mmol)
was added and complete consumption of starting material monitored by TLC. Anisole (11 mL, 0.10
mmol) was added thereafter, the reaction mixture stirred for 6 h at rt and subsequently concentrated in
vacuum. The crude product was purified by flash chromatography on alumina (neutral), first using
CH2Cl2/MeOH 10:0.9, then CH2Cl2/MeOH 10:1.4. Iodonium 14 was obtained as a white powder with
unknown counterions (26 mg). This material was suspended in 0.5 mL MeCN and pꢀTsOH·H2O (15.2
mg, 0.08 mmol) added while stirring at rt. The mixture was concentrated in vacuum to afford a yellowish
oil (34.5 mg). It was estimated by 1H NMR that the obtained oil contained 63% 14a and 37% pꢀTsOH by
weight (21 mg of 14a, 27% yield). Alternatively, quantitative counterion exchange was accomplished
using a RPꢀHPLC protocol as previously described.57 1H NMR (300 MHz, DMSOꢀd6, hydrogen signals
originating from excess pꢀTsOH were omitted for clarity) δ 9.52 (s, 1H), 8.42 (d, J = 1.6, 1H), 8.20 (d, J =
9.0, 2H), 8.15 (d, J = 1.7, 1H), 7.47 (d, J = 8.0, 4H), 7.11 (d, J = 7.9, 4H), 7.07 (d, J = 9.2, 2H), 4.73
(pseudo t, 1H), 3.79 (s, 3H), 3.76 – 3.62 (m, 2H), 3.54 – 3.41 (m, 2H), 3.33 – 3.18 (m, 4H), 3.14 – 3.04
(m, 1H), 2.94 – 2.82 (m, 1H), 2.80 – 2.71 (m, 1H), 2.29 (s, 6H), 2.22 – 2.17 (m, 1H), 2.05 – 1.86 (m, 5H),
1.82 –1.62 (m, 2H), 1.39 – 1.31 (m, 1H). HRMS (ESI+) calcd for C26H30IN2O2+ [Mꢀ2TsOꢀH]+ 529.1346,
found 529.1352; calcd for C26H31IN2O22+ [Mꢀ2TsO+H]2+ 265.0710, found 265.0715.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(4ꢀMethoxyphenyl)(palonosetron)iodonium ditrifluoroacetate (14b). This was synthesized
following the procedure for 14a using iodide 13 (70 mg, 0.17 mmol) in CH2Cl2/2,2,2ꢀtrifluoroethanol 1:1
(0.68 mL), pꢀTsOH·H2O (126 mg, 0.66 mmol), dried mCPBA (commercial grade 77%, 81 mg, 0.36
mmol) and anisole (72 mL, 0.66 mmol). The crude product was dissolved in HPLC solvent C (0.1% TFA
in water) containing little MeCN and the suspension was filtered. The obtained solution was purified by
-22ꢀ
ACS Paragon Plus Environment

Page 23 of 43
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
prep. RP18ꢀHPLC with UV detection at 254 nm, eluting first 2 min with 100% C, then using a gradient
from 100% C to 55% D (0.1% TFA in MeCN/H2O 4:6) in 35 min, 55% D to 100% D in 10 min. The
product eluted after 25 min at 32% D. The combined fractions were lyophilized to give 14b as an off
9
1
white solid (30 mg, 23%). H NMR (300 MHz, CDCl3) δ 12.75 (s, 1H), 8.35 (s, 1H), 7.90 (d, J = 8.8,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3H), 6.91 (d, J = 8.9, 2H), 4.68 (pseudo t, 1H), 3.82 (s, 3H), 3.89 – 3.75 (m, 1H), 3.68 – 3.56 (m, 3H),
3.38 – 3.21 (m, 4H), 3.13 – 3.01 (m, 1H), 2.95 – 2.70 (m, 2H), 2.36 (s, 1H), 2.20 – 1.88 (m, 6H), 1.79 –
1.63 (m, 1H), 1.32 (dd, J = 22.9, 11.1, 1H). 13C NMR (75 MHz, CDCl3) δ 164.4, 162.8, 141.1, 140.2,
138.0, 137.2, 131.9, 131.5, 117.8, 114.8, 104.4, 55.8, 50.9, 50.2, 49.2, 46.0, 45.9, 35.2, 28.2, 25.5, 25.0,
24.0, 21.4, 19.4. ESIꢀMS m/z 529 [Mꢀ2TFAꢀH]+.
(2,4,6ꢀTrimethylphenyl)(palonosetron)iodonium ditetrafluoroborate (14c). This was
synthesized following the procedure for 14a using iodide 13 (35 mg, 0.082 mmol) in CH2Cl2/2,2,2ꢀ
trifluoroethanol 1:1 (0.34 mL), pꢀTsOH·H2O (62 mg, 0.33 mmol), dried mCPBA (commercial grade
77%, 22 mg, 0.10 mmol) and mesitylene (46 mL, 0.33 mmol). The crude iodonium salt was partitioned
between 2 mL water and 8 mL EtOAc. The water phase was washed with EtOAc (2 x 8 mL) and
subsequently extracted with CH2Cl2 (3 x 8 mL). The combined CH2Cl2 layers were concentrated in vacuo
to give the ditosylate salt as a white solid (49 mg, 67%). 1H NMR (300 MHz, DMSOꢀd6) δ 9.50 (s, 1H),
8.11 (d, J = 1.6, 1H), 7.89 (d, J = 1.5, 1H), 7.46 (d, J = 8.0, 4H), 7.23 (s, 2H), 7.11 (d, J = 8.0, 4H), 4.69
(pseudo t, 1H), 3.81 – 3.70 (m, 1H), 3.71 – 3.57 (m, 1H), 3.51 – 3.39 (m, 2H), 3.29 – 3.17 (m, 4H), 3.13 –
3.02 (m, 1H), 2.95 – 2.81 (m, 1H), 2.79 – 2.69 (m, 1H), 2.59 (s, 6H), 2.30 (s, 3H), 2.28 (s, 6H), 2.20 –
2.15 (m, 1H), 2.06 – 1.85 (m, 5H), 1.80 – 1.62 (m, 2H), 1.33 – 1.26 (m, 1H). HRMS (ESI+) calcd for
C28H34IN2O [Mꢀ2TsOꢀH]+ 541.1710, found 541.1692.
The iodonium ditosylate (49 mg, 0.055 mmol) was partitioned between 6 mL CH2Cl2 and 2 mL aq.
1M NaBF4 solution. The aqueous layer was extracted with CH2Cl2 (2 x 6 mL). The combined organic
layers were extracted with aq. 1M NaBF4 solution (8 mL). The remaining water phase was reꢀextracted
with CH2Cl2 (2 x 10 mL). All organic CH2Cl2 layers were combined and concentrated in vacuo to afford
-23ꢀ
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 24 of 43
1
2
3
4
1
14c as a white solid (51 mg, quant.) Integration of tosylate and mesityl signals in the H NMR are
5
6
7
8
consistent with almost complete anion exchange (92:8 BF4 :TsO–). H NMR (300 MHz, DMSOꢀd6) δ
9.49 (s, 1H), 8.11 (d, J = 1.8, 1H), 7.89 (d, J = 1.7, 1H), 7.23 (s, 2H), 4.70 (pseudo t, 1H), 3.81 – 3.60 (m,
2H), 3.53 – 3.41 (m, 2H), 3.30 – 3.18 (m, 4H), 3.11 – 3.03 (m, 1H), 2.93 – 2.83 (m, 1H), 2.78 – 2.71 (m,
1H), 2.59 (s, 6H), 2.30 (s, 3H), 2.21 – 2.17 (m, 1H), 2.04 – 1.87 (m, 5H), 1.79 –1.63 (m, 2H), 1.33 – 1.27
(m, 1H).
–
1
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(S)ꢀ8ꢀFluoroꢀ2ꢀ((1R,3S,4R)ꢀquinuclidinꢀ3ꢀyl)ꢀ2,3,3a,4,5,6ꢀhexahydroꢀ1Hꢀ
benzo[de]isoquinolinꢀ1ꢀone (fluoropalonosetron, 15). A suspension of anisyl iodonium salt 14a (35
mg, 63 w%, 0.025 mmol) and CsF (11 mg, 0.072 mmol) in 1 mL MeCN (in a capped microwave vial)
was stirred at 100 °C for 2 h. The reaction was transferred to an ultrasonic bath, sonicated for 1 h and
subsequently stored in the freezer for 72 h. 2 mL acetonitrile was added and the reaction stirred at 100 °C
for another 48 h. After allowing the reaction to cool down to rt, the mixture was partitioned between 5 mL
CH2Cl2 and 2 mL 10% aq. NaHSO3 solution. 5 mL sat. aq. NaHCO3 solution was added and the water
phase was extracted with CH2Cl2 (2 x 10 mL). The combined organic layers were dried over Na2SO4,
filtered and concentrated under reduced pressure to afford a yellow solid (12 mg). The yellow solid was
dissolved in MeCN with little HPLC solvent C (0.1% TFA in water) and the suspension was filtered. The
obtained solution was purified by prep. RP18ꢀHPLC with UV detection at 254 nm, using a gradient of
90% C/10% E (0.1% TFA in MeCN/H2O 9:1) to 50% E in 30 min, 50% E to 100% E in 10 min. The
product eluted after 21 min at 38% E. The combined fractions were lyophilized to give 15 as the
trifluoroacetate salt (1.5 mg, 14%).
Fluoropalonosetron 15 was also synthesized using a different method from mesityl iodonium salt
14c. The reaction was set up in a glove box under nitrogen atmosphere using a microwave vial as reaction
vessel. Iodonium salt 14c (48 mg, 0.067 mmol), copper(II)triflate (12 mg, 0.033 mmol), NaF (4 mg, 0.068
mmol) and 18ꢀcrownꢀ6 (7 mg, 0.026 mmol) were dissolved in 0.68 mL DMF. The microwave vial was
capped and removed from the glove box. The reaction mixture was stirred at 60 °C for 21 h. DMF was
-24ꢀ
ACS Paragon Plus Environment

Page 25 of 43
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
removed in vacuo and 5 mL sat. aq. NaHCO3 solution was added. The water phase was extracted with
CH2Cl2 (3 x 10 mL). The combined organic layers were dried over Na2SO4, filtered and concentrated in
vacuo to afford a yellow solid (21 mg). The yellow solid was dissolved in MeCN with little HPLC solvent
C and the suspension was filtered. The obtained solution was purified by prep. RP18ꢀHPLC with UV
detection at 254 nm, using a gradient of 90% C/10% E to 50% E in 30 min, 50% E to 100% E in 10 min.
The product eluted after 21 min at 38% E. The fractions were lyophilized to give 15 as the trifluoroacetate
salt (3.8 mg, 13%). 1H NMR (300 MHz, CDCl3) δ 13.75 (s, 1H), 7.55 (dd, J = 9.0, 2.6, 1H), 6.98 (dd, J =
9.1, 2.5, 1H), 4.87 (pseudo t, 1H), 3.77 – 3.65 (m, 1H), 3.63 – 3.53 (m, 3H), 3.39 – 3.21 (m, 4H), 3.07 –
2.92 (m, 1H), 2.88 – 2.77 (m, 2H), 2.42 – 2.33 (m, 1H), 2.19 – 1.89 (m, 6H), 1.80 – 1.73 (m, 1H), 1.39 –
1.29 (m, 1H). 13C NMR (75 MHz, CDCl3) δ 162.4, 138.0, 137.9, 132.7, 132.7, 130.4, 130.3, 119.7, 119.4,
113.5, 113.2, 49.8, 49.6, 49.5, 46.0, 45.9, 34.9, 28.4, 26.3, 25.7, 24.5, 21.8, 19.9 (carbon directly bonded
to fluorine, or carbonyl carbon not observed). 19F NMR (376 MHz, DMSOꢀd6) δ ꢀ118.87 (t, J = 9.1). ESIꢀ
MS m/z 315 [M+H]+. HRMS (ESI+) calcd for C19H24FN2O [M+H]+ 315.1867, found 315.1855.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Radiochemistry. 11C-radiolabeling of granisetron: [11C]CO2 was produced via the 14N(p,α)11C
nuclear reaction by bombardment of nitrogen gas fortified with 0.5% oxygen using a Cyclone 18/9
cyclotron (18ꢀMeV; IBA, Belgium). After reduction over a supported nickel catalyst to [11C]CH4 and
subsequent gas phase iodination, [11C]CH3I was bubbled through a mixture of desmethyl granisetron
precursor 3 (1.0ꢀ1.5 mg) and aq. KOH (30 µL, 5 M) in DMF (0.5 mL). The mixture was heated to 90 °C
for 3 min. After dilution with 0.1% TFA in water (1.5 mL), the crude product was purified using the
semiꢀpreparative HPLC system (product peak at ca. 10 min, Figure S3A). The collected product was
diluted with water (10 mL), trapped on a Cꢀ18 cartridge (Waters, preconditioned with 5 mL EtOH and 10
mL water), washed with water (5 mL) and eluted with acidified EtOH (0.5 mL, 1% HCl). For formulation
of the final product [11C]granisetron ([11C]2), PBS (9.5 mL, 0.15 M) was added to give an ethanol
concentration of 5%. For quality control, an aliquot of the formulated solution was injected into the
11
analytical HPLC system. The identity of the Cꢀlabeled product was confirmed by comparison with the
-25ꢀ
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 26 of 43
1
2
3
4
5
6
7
8
retention time of its nonradioactive reference compound granisetron (2) and by coꢀinjection with 2
(Figure S4). Specific activity of [11C]granisetron ([11C]2) was calculated by comparison of UV peak
intensity with a calibration curve of the cold reference compound 2.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
18F-radiolabeling of fluoropalonosetron: Noꢀcarrierꢀadded [18F]fluoride was produced via the
18O(p,n)18F nuclear reaction by irradiation of isotopically enriched 18Oꢀwater in a fixedꢀenergy Cyclone
18/9 cyclotron (IBA). [18F]fluoride was trapped on a light QMA cartridge (Waters), which was
preconditioned with 0.5 M K2CO3 (5 mL) and water (5 mL). A 1 mL volume of Kryptofix K222/Cs2CO3
solution (Kryptofix K222; 2.5 mg, Cs2CO3, 2.6 mg in MeCN (0.7 mL)/water (0.3 mL) was used for the
elution of [18F−] from the cartridge. The solvents were evaporated at 100 °C under vacuum in the presence
of slight inflow of nitrogen gas, and azeotropic drying was carried out twice with the addition of 0.8 mL
acetonitrile each time, to afford dry [18F]CsF/K222 complex. A solution of the corresponding iodonium salt
precursor 14 (1ꢀ2 mg in 0.5 mL dry DMF, 1ꢀ2 mg TEMPO) was added to the dried [18F]CsF/K222
complex, and the reaction mixture was heated at 140 °C for 12 min and then 8 mL water was added. The
reaction mixture was passed through a preconditioned Cꢀ18 light cartridge, washed with 5 mL water,
eluted the crude product with 1.5 mL acidified ethanol (0.2 M) and followed by 1.5 mL water.
Purification was performed by semiꢀpreparative HPLC, with [18F]fluoropalonosetron ([18F]15) eluting at
about 22 min (Figure S3B). The collected product was diluted with water (8 mL), trapped on a Cꢀ18
cartridge (Waters, preconditioned with 5 mL EtOH and 10 mL water), washed with water (5 mL) and
eluted with acidified EtOH (1.5 mL, 1% HCl). The ethanol solution was concentrated under vacuum to
ca. 0.1 mL, then PBS (2 mL, 0.15 M) was added to give an ethanol concentration of 5%. For quality
control, an aliquot of the formulated solution was injected into the analytical HPLC system. The identity
of the 18Fꢀlabeled product [18F]15 was confirmed by comparison with the retention time of its
nonradioactive reference compound fluoropalonosetron 15 and by coꢀinjection with 15 (Figure S5).
Specific activity of [18F]fluoropalonosetron ([18F]15) was calculated by comparison of UV peak intensity
with a calibration curve of the cold reference compound 15.
-26ꢀ
ACS Paragon Plus Environment

Page 27 of 43
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
Radioligand Competition Binding. All cell culture reagents were obtained from commercial
sources and were of the highest obtainable grade. Human 5ꢀHT3A receptor subunit (accession number:
P46098) cDNA was obtained from Prof. John Peters (Dundee University, UK) and cloned into pcDNA3.1
(Invitrogen) for expression in HEK 293 cells. HEK 293 cells were maintained on 90 mm tissue culture
plates at 37 °C and 7% CO2 in a humidified atmosphere. They were cultured in DMEM : F12 with Glutaꢀ
MAX I media (DMEM/Nutrient Mix F12 (1:1), Invitrogen, Paisley, UK) containing 10 % fetal calf
serum. For radioligand binding studies, cells in 90 mm dishes were transfected using polyethyleneimine
(PEI; 25 kDa linear powder; Polysciences Inc., Eppelheim, Germany). 30 mL of PEI (1 mg/mL), 5 µg
cDNA and 1 mL DMEM were incubated for 10 min at room temperature, added drop wise to an 80 – 90
% confluent plate, and incubated for 2 – 3 days before harvesting. Transfected HEK 293 cells were
scraped into 1 mL of iceꢀcold HEPES buffer (10 mM, pH 7.4) and frozen. After thawing, they were
washed with HEPES buffer, reꢀsuspended, and 50 mg incubated in 0.5 mL HEPES buffer containing the
5ꢀHT3 receptor antagonist [3H]granisetron ([3H]2). For competition binding, reactions were incubated for
24 h at 4 °C with 0.6 nM [3H]granisetron and differing concentrations of competing ligands palonosetron
(1) and fluoropalonosetron 15. Nonꢀspecific binding was determined using 1 mM quipazine. Reactions
were terminated by vacuum filtration using a Brandel cell harvester onto GF/B filters preꢀsoaked in 0.3%
polyethyleneimine. Radioactivity was determined by scintillation counting using a Beckman BCLS6500
(Fullerton, CA, USA). Individual competition binding experiments were analyzed by iterative curve
fitting using the following equation in Prism v4.03:
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Amax − Amin
y = Amin
1 + 10[ L]−log IC
50
where Amin is the nonꢀspecific binding, Amax is the maximum specific binding, [L] is the concentration of
competing ligand and IC50 is the concentration of competing ligand that blocks half of the specific bound
radioligand. For representative doseꢀinhibition curves see Figure S1. Values were calculated for separate
experiments and reported as mean ± SEM.
-27ꢀ
ACS Paragon Plus Environment

ACS Chemical Neuroscience
Page 28 of 43
1
2
3
4
5
6
7
8
9
Determination of logD7.4 and In Vitro Stability Studies. The partition coefficient D was
determined by the shakeꢀflask method as previously described.68 In brief, nꢀoctanol saturated with
phosphate buffer pH 7.4 (0.5 mL) and phosphate buffer saturated with nꢀoctanol (0.5 mL) were mixed
with either [11C]granisetron ([11C]2, 20 µL) or [18F]fluoropalonosetron ([18F]15, 10 ꢁL). The samples were
shaken for 15 min and then centrifuged at 5000 g for 3 min. Radioactivity in each phase was measured in
a gamma counter (Wizard, PerkinElmer). LogD is expressed as the logarithm of the ratio between the
radioactivity concentrations in the nꢀoctanol and the phosphate buffer phase.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The tracer stability was examined in vitro in rodent plasma (mouse and rat) and human plasma at
time points 5, 10, 15 and 20 min. Plasma (400 µL) was incubated with 10 µL of [11C]granisetron ([11C]2)
solution at 37 °C under shaking. At each time point, an aliquot (100 µL) of the mixture was collected and
the reaction stopped with 100 µL iceꢀcooled MeCN. The samples were centrifuged (3 min, 5000 rpm) and
the supernatant collected. Each sample was analyzed by UPLC (Figure S2) using a similar gradient as for
HPLC.
Animals. Animal care and experiments were conducted in accordance with Swiss Animal Welfare
legislation and approved by the Veterinary Office of the Canton Zürich, Switzerland. Male Wistar rats at
5 weeks of age were purchased from Charles River, Sulzfeld, Germany. Unless otherwise stated, for all in
vivo experiments animals were anaesthetized with isoflurane in oxygen/air.
In Vitro Autoradiography. Radiotracer binding was evaluated by in vitro autoradiography
experiments with 20 ꢁm thick coronal brain cryoꢀsections of a wildꢀtype mouse (C57BL/6, male) and a
rat (Wistar, male) mounted on glass slides (SuperFrostTM Plus, Thermo Scientific, Braunschweig,
Germany). Slices were allowed to equilibrate on ice for 10 min before preꢀincubated in HEPES buffer (30
mM HEPES, 0.5 mM MgCl2, 110 mM NaCl, 5 mM KCl, 3 mM CaCl2, pH 7.4) supplemented with 0.1%
BSA for 10 min at 4°C. Slices were incubated with either 0.6 nM [11C]granisetron ([11C]2) or 0.4 nM
[18F]fluoropalonosetron ([18F]15) in HEPES/0.1% BSA for 15 min or 30 min, respectively, in a
humidified chamber at room temperature. For blocking conditions, 1 ꢁM cold palonosetron (1) or 1 ꢁM
-28ꢀ
ACS Paragon Plus Environment

Page 29 of 43
ACS Chemical Neuroscience
1
2
3
4
5
6
7
8
9
cold quinolineꢀ4ꢀcarboxylate 6 was added to the radiotracer solution. The slices were washed for 4 min in
HEPES/0.1% BSA buffer, twice for 3 min in HEPES buffer and twice for 5 s in distilled water at 4 °C.
The airꢀdried slides were exposed to a BASꢀMS 2025 phosphor imaging plate (Fuji Film, Dielsdorf,
Switzerland) for 15 min, scanned in a BASꢀ5000 bioꢀimaging reader (Fuji Film, Dielsdorf, Switzerland)
and analyzed with the AIDA 4.5 software (Raytest, Sprockhövel, Germany). For histological
characterization of the brain morphology hematoxylin and eosin (HE) (Sigma, St. Louis, MO, USA)
staining was applied according to standard protocols. All sections were covered using Eukitt mounting
medium (Sigma) and digitized in a slide scanner (Pannoramic 250, Sysmex, Horgen, Switzerland).
Specific brain regions were identified according to the anatomical atlas and HE staining.64
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In Vivo PET Imaging and Blocking Studies. PET scans were performed with a calibrated VISTA
eXplore small animal PET/CT scanner (Sedecal, Madrid, Spain). Respiratory rate and temperature were
controlled as described previously.69 In total, 8 Wistar rats (226 ꢀ 360 g) were scanned: three brain scans
with [11C]granisetron ([11C]2) under baseline and one brain scan under blocking conditions, one body
[11C]granisetron scan, one brain scan with [18F]fluoropalonosetron ([18F]15) under baseline and one under
blocking conditions as well as one body [18F]fluoropalonosetron scan. Rats were intravenously injected
into the tail vein with 46ꢀ53 MBq (brain scan) and 17 MBq (body scan) [11C]granisetron ([11C]2) or 30 ꢀ
38 MBq (brain scan) and 17 MBq (body scan) [18F]fluoropalonosetron ([18F]15). Radiotracer injection
and dynamic PET acquisition (list mode) were started simultaneously and PET scan duration was 60 min.
Blocking scans were performed by injecting 100 ꢁL cold palonosetron (1, 100 ꢁg/kg) intravenously
shortly before radiotracer application. At the end of each scan, a CT was recorded for anatomical
orientation. PET data were reconstructed in userꢀdefined time frames (2DꢀOSEM) applying random and
scatter corrections. Timeꢀactivity curves (TACs) of brain regions were generated with the implemented
rat brain region of interest (ROI) template supplied by PMOD biomedical imaging software (3.6, PMOD
Technologies Ltd., Zürich, Switzerland). Standardized uptake values (SUV) were calculated as tissue
activities (Bq/cm3), normalized to the injected dose per body weight (Bq/g).
-29ꢀ
ACS Paragon Plus Environment