Synthesis, radiolabeling and preclinical evaluation of a [11C]GMOM derivative as PET radiotracer for the ion channel of the N-methyl-D-aspartate receptor

Article abstract of DOI:10.1016/j.nucmedbio.2017.05.003

Introduction Presently available PET ligands for the NMDAr ion channel generally suffer from fast metabolism. The purpose of this study was to develop a metabolically more stable ligand for the NMDAr ion channel, taking [¹¹C]GMOM ([¹¹C]1) as the lead compound. Methods [¹¹C]1, its fluoralkyl analogue [¹⁸F]PK209 ([¹⁸F]2) and the newly synthesized fluorovinyloxy analogue [¹¹C]7b were evaluated ex vivo in male Wistar rats for metabolic stability. In addition, [¹¹C]7b was subjected to a biodistribution study and its affinity (K_i) and lipophilicity (logD_7.4) values were determined. Results The addition of a vinyl chain in the fluoromethoxy moiety did not negatively alter the affinity of [¹¹C]7b for the NMDAr, while lipophilicity was increased. Biodistribution studies showed higher uptake of [¹¹C]7b in forebrain regions compared with cerebellum. Pre-treatment with MK-801 decreased the overall brain uptake significantly, but not in a region-specific manner. 45?min after injection 78, 90 and 87% of activity in the brain was due to parent compound for [¹¹C]1, [¹⁸F]2 and [¹¹C]7b, respectively. In plasma, 26–31% of activity was due to parent compound. Conclusion Complete substitution of the alpha-carbon increased lipophilicity to more favorable values. Substitution of one or more hydrogens of the alpha-carbon atom in the methoxy moiety improved metabolic stability. In plasma, more parent compound was found for [¹⁸F]2 and [¹¹C]7b then for [¹¹C]1, although differences were not significant. At 45?min, significantly more parent [¹⁸F]2 and [¹¹C]7b was measured in the brain compared with [¹¹C]1.

Full text of DOI:10.1016/j.nucmedbio.2017.05.003

Nuclear Medicine and Biology 51 (2017) 25–32
Contents lists available at ScienceDirect
Nuclear Medicine and Biology
journal homepage: www.elsevier.com/locate/nucmedbio
Synthesis, radiolabeling and preclinical evaluation of a [11C]GMOM
derivative as PET radiotracer for the ion channel of the
N-methyl-D-aspartate receptor
⁎
Pieter J. Klein , Robert C. Schuit, Athanasios Metaxas, Johannes A.M. Christiaans, Esther Kooijman,
Adriaan A. Lammertsma, Bart N.M. van Berckel, Albert D. Windhorst
Department of Radiology & Nuclear Medicine, Neuroscience Campus Amsterdam, VU University Medical Center, P.O. Box 7057, 1007 MB Amsterdam, The Netherlands
a r t i c l e i n f o
a b s t r a c t
Article history:
Introduction: Presently available PET ligands for the NMDAr ion channel generally suffer from fast metabolism.
Received 6 January 2017
Received in revised form 29 March 2017
Accepted 4 May 2017
The purpose of this study was to develop a metabolically more stable ligand for the NMDAr ion channel, taking
[
11C]GMOM ([11C]1) as the lead compound.
Methods: [11C]1, its fluoralkyl analogue [18F]PK209 ([18F]2) and the newly synthesized fluorovinyloxy analogue
[
11C]7b were evaluated ex vivo in male Wistar rats for metabolic stability. In addition, [11C]7b was subjected to
Keywords:
PET
NMDA
Ion channel
a biodistribution study and its affinity (Ki) and lipophilicity (logD7.4) values were determined.
Results: The addition of a vinyl chain in the fluoromethoxy moiety did not negatively alter the affinity of [11C]7b
for the NMDAr, while lipophilicity was increased. Biodistribution studies showed higher uptake of [11C]7b in
forebrain regions compared with cerebellum. Pre-treatment with MK-801 decreased the overall brain uptake
significantly, but not in a region-specific manner. 45 min after injection 78, 90 and 87% of activity in the brain
was due to parent compound for [11C]1, [18F]2 and [11C]7b, respectively. In plasma, 26–31% of activity was due
to parent compound.
Uncompetitive antagonists
GMOM
[
18F]PK-209
Conclusion: Complete substitution of the alpha-carbon increased lipophilicity to more favorable values. Substitution
of one or more hydrogens of the alpha-carbon atom in the methoxy moiety improved metabolic stability. In plasma,
more parent compound was found for [18F]2 and [11C]7b then for [11C]1, although differences were not significant.
At 45 min, significantly more parent [18F]2 and [11C]7b was measured in the brain compared with [11C]1.
© 2017 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction
synaptic plasticity, learning and memory. Deregulation of NMDArs is
involved in several neurological and neuropsychiatric disorders [1–3].
The N-methyl-D-aspartate receptor (NMDAr) is beside the AMPA
and kainate receptors one of the three members of the ionotropic gluta-
mate receptor family, named after their corresponding agonists. The
NMDArs are formed as tetraheteromers, and contain an obligatory
glycine-binding NR1 subunit, in complex with glutamate-binding
NR2A-D, and occasionally NR3A-B subunits. These subunits together
form a ligand-gated ion channel, within which a magnesium binding
site exists. At resting state, the ion channel is blocked by Mg2+, which
inhibits ion flow (Na+, K+ and Ca2+) through the channel. This Mg2+
block can be removed by depolarization of the plasma membrane, a
physiological process underlying the important role of NMDArs in
Phenylcyclidine (PCP), thienylcyclohexyl piperidine (TCP), ketamine,
memantine and MK-801 are compounds known to bind within the ion-
channel site of NMDArs and block the channel pore. These compounds
were initially synthesized for therapeutic purposes, and later
radiolabeled to develop PET or SPECT radiotracers for the NMDArs [4–7].
Another class of interesting compounds that interacts at the ion chan-
nel binding site are the N,N′-diarylguanidines. To date, several N-(2,5-
disubstited phenyl)-N′-(3-substituted phenyl)-N′-methylguanidines,
such as [11C]GMOM, [11C]CNS-5161, [18F]GE-179 and [18F]PK209 (See
Fig. 1.), have been reported, all with promising characteristics, such as
high affinity, moderate lipophilicity, and selectivity for the NMDA ion
channel over the sigma receptor.
[
11C]GMOM, ([11C]1, Ki: 5.2–21.7 nM vs. [3H]MK-801; log P: 2.34;
logD7.4: 1.72) [8,18] demonstrated regional brain uptake in awake rats
ranging from 0.75 0.13% injected dose per gram (ID/g) in the medulla
and pons to 1.15 0.17% ID/g in the occipital cortex. Pre-treatment with
MK-801 (1 mg·kg−1) significantly reduced [11C]1 uptake in all regions
(24–28%), while D-serine, a glycine site co-agonist, increased uptake in
⁎
Corresponding author at: VU University Medical Center, Department of Radiology and
Nuclear Medicine, Neuroscience Campus Amsterdam, Location Radionuclide Center, De
Boelelaan 1085C, 1081HV, Amsterdam, The Netherlands. Tel.: +31 20 4449101;
fax: +31 20 4449121.
E-mail address: pj.klein@vumc.nl (P.J. Klein).
http://dx.doi.org/10.1016/j.nucmedbio.2017.05.003
0969-8051/© 2017 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

26
P.J. Klein et al. / Nuclear Medicine and Biology 51 (2017) 25–32
all regions (10–24%). Administering the NR2B subunit antagonist RO 25–
6981 reduced uptake of [11C]1 by 24–38% over control. In isoflurane anes-
thetized baboons a fairly uniform distribution volume across the brain
was observed. Pre-treatment with MK-801 (0.5 and 1.0 mg·kg−1) did
not alter regional distribution volumes, indicating a lack of saturable
binding. During scans the amount of [11C]1 in plasma was, on average,
18.6 5.9%, and only polar metabolites were observed [8,9].
Thin-layer chromatography (TLC) was performed on Merck DC-
alufolien, silica gel 60, F254. Flash column chromatography was per-
formed on silica gel 60 Å, 230–400 Mesh. All chemicals were used with-
out further purification, unless stated otherwise. The high-performance
liquid chromatography (HPLC) analysis system consisted of a Jasco PU-
2089 HPLC pump (Jasco Benelux, de Meern, the Netherlands), a
Rheodyne injector with a 20 μL loop (Thermo Fischer Scientific, Breda,
the Netherlands), a Jasco UV-2075 Plus UV detector set at a wavelength
of 254 nm and a Raytest Na(I) radioactivity detector (Raytest,
Straubenhardt, Germany). HPLC data were collected and integrated
with the software package GINA 5.01. For high-resolution mass spec-
trometry (HRMS), a Bruker MicroTOFQ with ESI (electrospray ioniza-
tion) in a positive mode (Billerica, USA) was used. Samples were
injected (10 μL) in a liquid flow of methanol/water (1/1) at a rate of
100 μL·min−1. [11C]CO2 was produced by the 14N(p,α)11C nuclear reac-
tion with an IBA Cyclone 18/9 cyclotron (Louvain-La-Neuve, Belgium).
Radioactivity was measured with a Veenstra VDC-405 dose calibrator
(Comecer, Joure, the Netherlands). Radiochemistry was carried out in
homemade, remotely controlled devices [23].
In human, highest uptake of [11C]GMOM was observed in NMDAr
rich regions, such as hippocampus and thalamus. Lowest values were
found in temporal cortex and cerebellum. [11C]GMOM showed rapid
metabolism, after 20 min. 50% of the radioactivity in the plasma was
due to parent compound. After administrating S-ketamine, a significant
reduction in uptake was observed for the entire brain suggesting specific
binding of [11C]GMOM to NMDAr. [11C]GMOM could be used for imaging
and quantifying the NMDAr [10].
[
11C]CNS-5161, (Ki: 1.87 nM vs. [3H]MK-801) a thiomethyl analogue
of GMOM, with a logD7.4 of 2.68, demonstrated uptake in human brain
tissues with the lowest uptake in cerebellum and the highest in putamen
and thalamus. This compound, however, suffered from fast metabolism
and specific uptake could not be demonstrated [11–15].
In humans, [18F]GE-179 (Ki: 2.35 nM vs. [3H]TCP), a fluoroethyl
analogue of CNS-5161 with a logD7.4 of 2.49, showed rapid plasma
metabolism. After 16 min 50% of the radioactivity in the plasma could
be attributed to radiolabeled metabolites. Nevertheless, the rate of
metabolism was slower than for [11C]CNS-5161. Uptake in gray matter
was relatively homogeneous [16,17].
2.2. Chemistry
2.2.1. N-(3-hydroxyphenyl)cyanamide (5a) and N-(3-hydroxyphenyl)-N-
methylcyanamide (5b)
N-(3-hydroxyphenyl)cyanamide (5a) and N-(3-hydroxyphenyl)-N-
methylcyanamide (5b) were prepared according to published proce-
dures [18].
Recently, [18F]PK209 ([18F]2, Ki: 18 nM vs., [3H]MK-801), a
fluoromethoxy analogue of GMOM with a logD7.4 of 1.45 was reported.
Preclinical evaluation in rhesus monkeys showed retention in NMDAr
rich cortical regions relative to cerebellum. In two out of three subjects,
a reduction in signal relative to baseline was seen after MK-801 pre-
treatment. However, [18F]2 suffered from fast metabolism, as 10 min
post injection only 30 6% of the radioactivity in the plasma was due
to parent compound [18,19].
2.2.2. 1-(2-chloro-5-(methylthio)phenyl)-3-(3-hydroxyphenyl)guanidine
(6a)
A
screw cap reaction vessel was charged with N-(3-
hydroxyphenyl)cyanamide (5a) (0.68 g, 5.08 mmol), 2-chloro-5-
(methylthio)aniline hydrochloride (1.16 g, 5.51 mmol) and 200 μL chlo-
robenzene. Next, this vessel was flushed with nitrogen, closed and
stirred at 165 °C for 12 h. The reaction mixture was cooled down to
20 °C and dissolved in ethyl acetate (25 mL) and washed with 0.1 M
HCl (2 × 25 mL) followed by water (25 mL). The pH of the combined
aqueous layers was adjusted with potassium carbonate to pH ≥ 10 and
extracted with ethyl acetate (2 × 25 mL). The organic layers were
collected and dried over anhydrous MgSO4, filtrated and evaporated
to dryness under reduced pressure. The residue was purified by column
chromatography using ethyl acetate/triethylamine (99:1, v/v) to obtain
the title compound as a white solid (1.16 g, 3.77 mmol, 74%). Rf 0.13
(ethyl acetate/triethylamine 99:1 v/v). 1H NMR (CDCl3) δ 7.23 (d, J =
8.40 Hz, 1H, HAryl), 7.05 (t, J = 8.03 Hz, 1H, HAryl), 6.96 (d, J =
2.21 Hz, 1H, HAryl), 6.84 (dd, J = 8.43 Hz, J = 2.25 Hz, 1H, HAryl), 6.74–
6.72 (m, 1H, HAryl), 6.61–6.57 (m, 2H, HAryl), 5.96 (bs, 4H, 3 x NH,
OH), 2.38 (s, 3H, SCH3); 13C NMR (CDCl3) δ 157.76 (Ar-NH), 150.74
(NCN), 143.98 (Ar-O), 140.83 (Ar-NH), 137.93 (Ar-S), 130.02 (Ar),
129.99 (Ar), 124.60 (Ar-Cl), 122.55 (Ar), 121.87 (Ar), 113.67 (Ar),
111.43 (Ar), 109.69 (Ar), 15.55 (CH3).
All compounds described above consist of the same guanidine core
structure, and show fast peripheral metabolism. Structural alterations
exist in one or more short alkyl chains. It is known that the P450 cyto-
chrome system is responsible for O-, N- and S-demethylation and/or oxi-
dation [20]. A common procedure to prevent metabolism of short alkyl
chains is the replacement of hydrogen by deuterium or fluoride [21].
The CH2F group can be used as a bio-isostere of a methyl group,
preventing metabolic oxidation by its electron withdrawing effect [22].
Since [18F]2 next to [11C]1 still showed relatively fast metabolism in
rhesus monkeys, an obvious choice was to introduce deuterium or more
fluorine atoms in the methoxy moiety. From previous SAR studies it was
known that there is some spatial freedom for alterating the methoxy
moiety without losing affinity toward the NMDAr. The methoxy moiety
can be altered with two or three fluorine atoms, but the resulting
molecules showed less promising characteristics for use as a PET tracer
[18]. A vinylic moiety could be an alternative way to prevent metabolic
oxidation of the methoxy moiety of [18F]2.
The aim of the present study was to develop an analogue of [11C]
GMOM with improved metabolic stability and to evaluate its potential
as an NMDAr ion channel ligand.
2.2.3.
3-(2-chloro-5-(methylthio)phenyl)-1-(3-hydroxyphenyl)-1-
methylguanidine (6b)
3-(2-chloro-5-(methylthio)phenyl)-1-(3-hydroxyphenyl)-1-
methylguanidine (6b) was prepared according to a published procedure
[18].
2. Materials and methods
2.1. General
2.2.4. 3-(2-chloro-5-(methylthio)phenyl)-1-(3-(1-fluorovinyloxy)phenyl)-
guanidine (7a)
Nuclear magnetic resonance (NMR) spectra (1H, 13C, 19F NMR) were
recorded on a Bruker Avance 250 (250.13, 62.90, 235.36 MHz respec-
tively) (Billerica, USA). Chemical shifts of the NMR spectra are reported
in parts per million (ppm) relative to the solvent residual peak. Descrip-
tion of signals: s = singlet, bs = broad singlet, d = doublet, t = triplet,
q = quartet, m = multiplet, dd = double doublet, ddd = doublet of dd,
dt = doublet of triplets, tt = triplet of triplets, dq = double quartet.
A screw cap reaction vessel containing a solution of sodium hydride
(60%) (53 mg, 1.33 mmol) in 1,2-dimethoxyethane (2 mL) and N,N-
dimethylformamide (0.15 mL) was stirred at 0 °C. 1-(2-chloro-5-
(methylthio)phenyl)-3-(3-hydroxyphenyl)guanidine (6a) (194 mg,
0.63 mmol) and sodium iodide (19 mg, 0.13 mmol) was added. After
30 min 1-bromo-1-fluoroethene (150 μL, 2.00 mmol) dissolved in 1,2-

P.J. Klein et al. / Nuclear Medicine and Biology 51 (2017) 25–32
27
dimethoxyethane (0.5 mL) was added. Stirring at 60 °C was continued for
24 h. The reaction mixture was diluted with water (25 mL) and extracted
with dichloromethane (3 × 25 mL). The combined organic fraction was
washed with water (25 mL) and brine (25 mL). The organic layer was col-
lected, dried over anhydrous magnesium sulfate, filtered and evaporated
to dryness under reduced pressure. The crude product was purified over
silica (ethyl acetate/hexanes/triethylamine (33/66/1, v/v/v) to obtain the
title compound as light yellow oil. (41 mg, 0.12 mmol, 18%). Rf 0.24 (ethyl
acetate/hexanes/triethylamine 33/66/1 v/v/v). 1H NMR (CDCl3) δ 7.29–
7.16 (m, 3H, HAryl), 7.02–6.98 (m, 2H, HAryl), 6.88–6.79 (m, 2H, HAryl),
J = 8.20 Hz, 1H, HAryl), 7.18–7.16 (m, 2H, HAryl), 6.94 (d, J = 8.20 Hz,
1H, HAryl), 6.85–6.78 (m, 2H, HAryl), 6.59 (s, 1.7H, fumaric acid), 5.98
(bs, 2H, NH), 4.06 (t, J = 4.18 Hz, CH2a), 3.98 (dd, JHF = 40.57 Hz, J =
4.14 Hz, CH2b), 3.29 (s, 3H, NCH3), 2.43 (s, 3H, SCH3). HRMS calcd
for C17H17ClFN3OS 365.0765, found 366.0832 (M + H). Purity
98.4%, rt. = 5.02 min (Platinum EPS, C18 100A 5 μ, 250 × 4.6 mm;
50 mM NH4H2PO4 pH = 2.5/methanol (30/70, v/v); 1 mL∙min−1).
2.3. Radiochemistry
5.22 (bs, 3H, NH), 4.06 (t, JHF = 3.62 Hz, 1H, FCCHa), 3.81 (dd, JHF
=
2.3.1. [11C]1 and [18F]2
38.58 Hz, JHH = 3.79 Hz, FCCHb), 2.44 (s, 3H, SCH3); 13C NMR (CDCl3) δ
160.13 (d, JCF = 274.83 Hz, CF), 154.86 (Ar-NH), 149.13 (NCN), 148.94
(Ar-O), 143.15 (Ar-NH), 138.28 (Ar-S), 130.46 (Ar), 130.13 (Ar), 123.87
(Ar-Cl), 121.97 (Ar), 121.87 (Ar), 118.31 (Ar), 112.74 (Ar), 111.77 (Ar),
70.55 (d, JCF = 30.79 Hz, FC = CH2), 15.90 (SCH3); 19F NMR (CDCl3) δ
−81.28 (dd, JHF = 38.64, 3.22 Hz, 1F, CF).
[
11C]1 and [18F]2 were prepared according to published procedures
[10,18].
2.3.2. 3-(2-chloro-5-(methylthio)phenyl)-1-(3-(1-fluorovinyloxy)phenyl)-
1-[11C]methylguanidine ([11C]7b)
[
11C]CO2 was trapped into a solution of LiAlH4 in THF (0.1 mL) at
room temperature by a helium flow of 10 mL·min−1. The solution was
heated to 130 °C and the helium flow was increased to 100 mL·min−1
to evaporate the THF. After 3 min the helium flow was adjusted to
10 mL·min−1, HI (55% solution, 0.2 mL) was added and [11C]CH3I was
trapped into a second reaction vial containing precursor 7a (0.5 mg,
1.40 μmol), 5 M NaOH (5 μL) and DMSO (300 μL). After complete
trapping of [11C]CH3I the reaction mixture was heated at 60 °C for
2.2.5. 3-(2-chloro-5-(methylthio)phenyl)-1-(3-(1-fluorovinyloxy)phenyl)-
1-methylguanidine (7b)
A screw cap reaction vessel containing a solution of sodium hydride
(60%) (41 mg, 1.03 mmol) in 1,2-dimethoxyethane (2 mL) and N,N-
dimethylformamide (0.15 mL) was stirred at 0 °C. 3-(2-chloro-5-
(methylthio)phenyl)-1-(3-hydroxyphenyl)-1-methylguanidine (6b)
(154 mg, 0.48 mmol) and sodium iodide (16 mg, 0.11 mmol) was
added. After 30 min 1-bromo-1-fluoroethene (150 μL, 2.00 mmol) dis-
solved in 1,2-dimethoxyethane (0.5 mL) was added. Stirring at 60 °C
was continued for 24 h. The reaction mixture was diluted with water
(25 mL) and extracted with dichloromethane (3 × 20 mL). The com-
bined organic fraction was washed with water (25 mL) and brine
(25 mL). The organic layer was collected, dried over anhydrous magne-
sium sulfate, filtered and evaporated to dryness under reduced pres-
sure. The crude product was purified over silica (ethyl acetate/
hexanes/triethylamine (25/75/1, v/v/v) to obtain the title compound
as a yellow oil. (131 mg, 0.36 mmol, 75%). Rf 0.21 (dichloromethane/
methanol 95:5 v/v). 1H NMR (CDCl3) δ 7.37 (t, J = 8.05 Hz, 1H, HAryl),
7.26 (d, J = 8.24 Hz, 1H, HAryl), 7.15–7.09 (m, 2H, HAryl), 7.02–6.97 (m,
1H, HAryl), 6.89 (d, J = 2.28 Hz, 1H, HAryl), 6.81 (dd, J = 8.34, J =
2.32 Hz, 1H, HAryl), 4.16–4.13 (m, 1H, FCHa), 3.95 (bs, 2H, NH), 3.90
(dd, JHF = 38.20, 3.83 Hz, 1H, FCHb), 3.41 (s, 3H, NCH3), 2.45 (s, 3H,
SCH3); 13C NMR (CDCl3) δ 161.79 (Ar-NH), 157.39 (NCN) 152.79 (d,
JCF = 269.93 Hz, CF), 147.29 (Ar-O), 146.10 (Ar-NH), 137.85 (Ar-S),
130.74 (Ar), 130.17 (Ar), 124.38 (Ar-Cl), 122.71 (Ar), 122.37 (Ar),
121.44 (Ar), 116.13 (Ar), 115.43 (Ar), 71.54 (d, JCF = 30.66 Hz, FC =
CH2), 38.91 (NCH3), 16.01 (SCH3); 19F NMR (CDCl3) δ −81.41 (dd,
JHF = 38.18, 3.08 Hz, 1F, CF). The free base (56 mg) was converted
into its fumaric acid salt (39 mg, 0.08 mmol, 52%). 1H NMR (DMSO-
d6) δ 1H NMR (DMSO-d6) δ 7.39 (t, J = 8.34 Hz, 1H, HAryl), 7.28 (d,
3
min. The reaction mixture was quenched with a 10 mM
ammoniumhydrogencarbonate (pH = 7.4, 0.8 mL) solution before
loading it onto preparative HPLC (Platinum EPS, C18 100A, 5 μ;
150 × 10.0 mm; 10 mM ammoniumhydrogencarbonate pH = 7.4/
acetonitrile 25/75 v/v; 5 mL·min−1; 254 nm). The product eluted at
7 min, and was collected and diluted with water (40 mL). The solution
was concentrated on a tC18plus Seppak, rinsed with water (20 mL),
subsequently eluted with ethanol (96%, 1.5 mL) and diluted with a solu-
tion of 7.11 mM NaH2PO4 in 0.9% NaCl (w/v in water), pH 5.2 (13.5 mL)
to give a final solution of 9.6% ethanol in phosphate buffer containing
[
11C]6. The (radio)chemical purity was verified using analytical radio
HPLC (Waters μBondapak, C18 125A, 10 μm, 250 × 4.6 mm; 10 mM
ammoniumhydrogencarbonate pH = 7.4/acetonitrile 30/70 v/v;
1.5 mL·min−1, 254 nm). The specific activity was calculated against a
calibration curve of 7b using the same HPLC system.
2.3.3. Determination of LogDoct,7.4
The distribution of the radiolabeled compounds between 1-octanol
and 0.2 M phosphate buffer (pH = 7.4) was measured in triplicate at
room temperature. Briefly, 1 mL of a 1–5 MBq·mL−1 solution of the
radiolabeled compound in 0.2 M phosphate buffer (pH 7.4) was vigor-
ously mixed with 1 mL of 1-octanol for 1 min at room temperature
using a vortex. After a settling period of 30 min, five samples of 100 μL
were taken from both layers. To determine recovery, 5 samples of
Scheme 1. Reagents and conditions: (i) cyanogen bromide, diethyl ether, 0 °C → rt., 17 h; (ii) methyliodide, potassium carbonate, N,N-dimethylformamide, rt., 3 h; (iii), 2-chloro-5-
(methylthio)aniline hydrochloride, chlorobenzene, 165 °C, 6 h; (iv) sodium hydride, sodium iodide, 1-bromo-1-fluoroethene, 1,2-dimethoxyethane/N,N-dimethylformamide, 60 °C, 24 h.

28
P.J. Klein et al. / Nuclear Medicine and Biology 51 (2017) 25–32
Table 1
Table 2
Affinity (nM) for the ion channel of the NMDAr.
Measured LogD7.4
.
Ki (nM)a
Compound
LogDoct,7.4
a
Compound
1b
2b
7b
21.7
18.4
18.6
7.5
8.1
7.1
[
[
[
11C]1b
18F]2b
11C]7b
1.72
1.45
2.62
0.01
0.02
0.04
a
a
Ki measured against 5 nM [3H]MK-801 and presented as mean
SD of 4 independent
N = 3, presented as mean
Values taken from Klein et al. [18].
SD.
b
determinations, each conducted in triplicate.
b
Ki values taken from Klein et al. [18].
4 °C for 5 min (Hettich Universal 16, Depex B.V., The Netherlands).
Plasma was separated from blood cells, 1 mL was transferred into
an Eppendorf, and acetonitrile (1 mL) was added to precipitate the
proteins. After centrifuging for 1 min at 4000 rpm, the supernatant
was analyzed using HPLC; Dionex Ultimate 3000 HPLC system (Dionex,
Breda, the Netherlands), equipped with a Phenomenex Gemini C8,
250 × 10.0 mm column (Phenomenex, Utrecht, the Netherlands), gradient,
0.1 M ammonium acetate/acetonitrile, 0–3 min 10% acetonitrile, 11 min
80% acetonitrile, 14 min 10% acetonitrile, 4.0 mL·min−1. The HPLC system
was controlled using the software package Chromeleon 6.80 (Dionex,
Breda, the Netherlands). Fractions of 30 s were collected and counted
using a Perking Elmer 2480 single well gammacounter (Groningen, the
Netherlands). Graphs were reconstructed using Excel2007® (Microsoft,
Redmond, WA, USA).
100 μL were taken from the 1–5 MBq·mL−1 solution. All samples were
counted for radioactivity (LKB Wallac 1282 Compugamma CS, Turku,
Finland). The LogDoct,7.4 value was calculated according to LogDoct,7.4
=
10Log(Aoct/Abuffer), where Aoct and Abuffer represent average radioactivity
of 5 1-octanol and 5 buffer samples, respectively.
2.4. Pharmacology
2.4.1. Animals
Adult male Wistar rats (Harlan, Horst, The Netherlands) weighing
200 to 224 g at arrival were housed in groups of five to six animals
per cage with ad libitum access to food (Teklad Global 16% Protein
Rodent Diet, Harlan, Madison, WI, USA) and tap water. Rats were kept
in an animal room with a constant temperature of ~21 °C and in a
12 h light/dark cycle in which lights were switched on at 8:00 am.
Animal procedures were performed in accordance with Dutch laws on
animal experimentation and after approval by the local animal ethics
committee, study number DEC_PET12–01.
Brain tissue was homogenized with an ultrasonic homogenizer
(Braunsonic 1510, Kronberg, Germany) in water, under ice cooling,
and subsequently centrifuged at 4000 rpm and 4 °C for 5 min. Separated
supernatants were treated as above. The recovery of the metabolite
analysis was validated to be above 95%.
3. Results and discussion
2.4.2. Membrane preparation, competition binding assays and
data analysis
3.1. Chemistry
Membrane preparation, competition binding assays and data analysis
were carried out according to published procedures [18].
Starting from 4, the synthesis of guanidines 6a–b, summarized in
Scheme 1, was performed according to methods described previously
and with comparable yields [18]. Alkylation of 6a–b was carried out
by treating the phenol moiety with 1-bromo-1-fluoroethene following
a method adapted from Antonsen et al. [24]. At room temperature 1-
bromo-1-fluoroethene is gaseous, therefore, it was condensed to 0 °C
before mixing it with 1,2-dimethoxyethane and subsequent addition
to the reaction mixture. After purification, 7a and 7b were obtained in
18 and 75% yields, respectively.
2.4.3. Biodistribution and blocking studies
The biodistribution of [11C]1, [18F]2 and [11C]7b was determined in
male Wistar rats without anesthesia. Rats (N = 4 per time point)
received a dose of either 100 MBq [11C]1, [11C]7b or 40 MBq [18F]2 in
the tail vein and were decapitated at 5, 15, 30 or 60 min after injection.
Blood, heart, lung, liver, kidney and bone (skull) were removed. In addi-
tion, several brain areas were dissected: hippocampus, striatum, hypo-
thalamic region, cerebellum, cerebral cortex and rest brain. These
tissues were counted for radioactivity with a gamma counter (LKB
Wallac, Turku, Finland) and weighed. Results were expressed as percent
injected dose per gram (%ID/g). For blocking studies, rats received a sub-
cutaneous dose of 0.6 mg·kg−1 MK-801(salt weight, intraperitoneal),
15 min prior injection of the radiotracer.
3.2. Affinity of 7b
Table 1 shows the affinities of 1, 2 and 7b (tested as fumaric salts) in
nM for the ion channel of the NMDAr against 5 nM of [3H]MK-801. Re-
markably, affinities of all compounds were in the same range. This again
indicates that small structural alterations in the methoxy moiety are
tolerated by the NMDAr.
The affinity of 1 was previously reported as 5.2 and 21.7 nM respec-
tively [8,18]. The discrepancy in affinity between these two studies is
most likely due to differences in the method of determining the affinity
toward the NMDAr. In the present study compounds 1, 2 and 7b were
tested under the same conditions.
2.4.4. Metabolite analysis
The metabolic profile of [11C]1, [18F]2 and [11C]7b was determined in
male Wistar rats without anesthesia. Rats (N = 3 per time point) re-
ceived a dose of either 100 MBq [11C]1, [11C]7b or 50 MBq [18F]2 in
the tail vein and were decapitated at either 15 or 45 min after injection.
A blood sample was obtained and the brain was isolated. All blood sam-
ples were collected in heparin tubes and centrifuged at 4.000 rpm and
Scheme 2. Reagents and conditions: (i) [11C]CH3I, NaOH, DMSO, 60 °C, 3 min.

P.J. Klein et al. / Nuclear Medicine and Biology 51 (2017) 25–32
29
Fig. 1. [11C]GMOM and its fluorinated analogue [18F]PK209.
higher uptake in prefrontal cortex and the lowest in cerebellum. Organ
uptake of [11C]7b was highest in the lungs followed by kidneys and
liver.
3.3. Radiochemistry
Compound 7a was radiolabeled with [11C]CH3I on the N-methyl
position of the guanidine next to the vinyloxyphenyl moiety at 60 °C
in 3 min using NaOH as base in DMSO. [11C]7b (Scheme 2) was pre-
pared in a one-pot synthetic procedure in 20 6% (decay-corrected,
calculated from [11C]CH3I) yield, providing 4.1 1.4 GBq (N = 10) of
formulated product at the end of synthesis with a radiochemical purity
of N99% and with a specific radioactivity of 112 27 GBq/μmol (N = 4).
During synthesis of [11C]7b a side product, 7a [11C]methylated at the
other guanidine N, was formed. This side product was easily removed
using preparative HPLC. Using these reaction conditions a sufficient
amount of [11C]7b was obtained to perform the preclinical experiments.
Further optimization was not performed.
3.6. Blocking studies of [11C]7b
MK-801 (0.6 mg·kg−1) was used to determine the specificity of up-
take of ligand [11C]7b. The blocking dose was chosen to be two-to-
three times the in vivo ED50 of MK-801 for NMDArs [26,27]. MK-801
was administered 15 min prior to injection of the radiotracer, as previous
studies have shown that Tmax of MK-801 is obtained 10 to 30 min after
administration [27]. Fig. 3 shows the tracer biodistribution at 15 min
with and without MK-801 pre-treatment. Total brain uptake at 15 min
was 0.80 0.05 before and 0.66 0.03%ID/g (P b 0.05, unpaired t test
In contrast to [11C]7b, ligands [11C]1 and [18F]2 were radiolabeled
at the phenolic moiety. Currently there is no method known for
radiolabeling the fluoro-vinyloxy moiety with either carbon-11 or
fluorine-18. Both [11C]1 and [18F]2 could be radiolabeled with carbon-
11 in the same position as in [11C]7b, but it was decided to keep both
ligands original as published before. Compound 1 and 2 were radiolabeled
as described previously in sufficient yields and high radiochemical purity
and specific activity [10,18].
with Welch's correction, N = 4, mean
SEM) after pre-treatment.
This overall decrease represents an 18% reduction from control, which
is well within the range of maximum specific accumulation that can be
achieved for ion-channel blockers using this type of in vivo assay
[26,29]. Individual t-tests, conducted for each brain area separately,
showed no region-specific differences in the uptake of [11C]7b after
MK-801 administration.
3.7. Metabolite analysis of [11C]1, [18F]2 and [11C]7b
3.4. Lipophilicity of [11C]7b
The lipophilicity of compound [11C]7b is shown in Table 2 and was
determined as the LogD7.4 in octanol/buffer. Table 2 compares the
LogD7.4 values of 1, 2 and 7b. The lipophilicity of ligand 7b is well within
the borders of 2.0–3.5 stated by Pike indicating an optimal passive brain
entry in vivo [25].
When using PET as quantitative method to image the human brain, it
is important to consider metabolism of a radioligand in vivo, as it is not
possible to distinguish between signals from radiolabeled metabolites
and parent compound. Firstly, radiolabeled metabolites should not
cross the blood–brain barrier as it may hamper interpretation of mea-
sured signals from the brain. Secondly, the speed of metabolism is an im-
portant factor. The lower the amount of parent compound, the lower the
signal from the brain, leading to a decreased noise/signal ratio.
3.5. Ex vivo biodistribution of [11C]7b
Ex vivo biodistribution and regional brain uptake of [11C]7b were
determined in non-anesthetized male Wistar rats, to avoid any con-
founding effect of anesthesia on NMDAr binding.
Metabolites were determined in non-anesthetized male Wistar rats.
Fig. 4 summarizes parent compound and labeled metabolite fractions
for [11C]1, [18F]2 and [11C]7b in plasma and brain. See Supporting infor-
mation for representative chromatograms. The first radioactivity peak
observed with HPLC was identified as the polar fraction, most likely
representing single carbon molecules such as formaldehyde, formic
acid or carbon dioxide. The last peak was identified as the parent
Total brain uptake of [11C]7b at 5, 15, 30 and 60 min after injection
was 0.76 0.08, 0.70 0.05, 0.40 0.04 and 0.28 0.01%ID/g, respec-
tively. Fig. 2 shows regional and total brain uptake as well as organ up-
take. [11C]7b showed uptake across all dissected brain areas, with initial
Fig. 2. Mean ( SEM) uptake of [11C]7b in CNS (left) and various organs (right) expressed in %ID/g (N = 4 for each time point).

30
P.J. Klein et al. / Nuclear Medicine and Biology 51 (2017) 25–32
Fig. 3. Uptake of [11C]7b before and after pre-treatment with MK-801(0.6 mg·kg−1) in CNS (left) and various organs (right).
compound. Apart from these polar metabolite and parent fractions,
multiple non-polar peaks were observed, i.e. 2, 3 and 2 peaks in plasma,
and 3, 3 and 2 peaks in brain for [11C]1, [18F]2 and [11C]7b, respectively.
For each compound, the retention times of metabolites and parent
compound observed in plasma were identical to the retention times
observed in brain.
The amount of parent compound in plasma and brain at 15 and
45 min is summarized in Fig. 5. In plasma, the parent fraction was 42–
47% at 15 min and 26–31% at 45 min for all three radiotracers. Although
no significant difference between the radiotracers could be observed, a
slight trend between the radiotracers is visible. This trend is possibly
related to the degree of substitution of hydrogen atoms in the methoxy
moiety.
After 15 min, for all three ligands, 91–94% of radioactivity in the
brain was due to parent compound. After 45 min the amount of parent
compound of [11C]1 was reduced to 78%. For both [18F]2 (90%) and [11C]
7b (87%) the amount of parent compound was significantly higher than
for [11C]1.
Multiple chemical moieties of [11C]1, [18F]2 and [11C]7b are prone to
metabolism, e.g. by oxidation or demethylation. In a recent [18F]GE-179
study it is shown that one of the metabolites was likely to be a sulfoxide
derivative [17]. The amount of metabolites formed by sulfur oxidation
Fig. 4. Distribution of radiolabeled polar metabolites, radiolabeled non-polar radiolabeled metabolites and parent compound fractions in male Wistar rats of [11C]1, [18F]2 and [11C]7b in
plasma (left side) and brain (right side) at 15 and 45 min (N = 3 SD).

P.J. Klein et al. / Nuclear Medicine and Biology 51 (2017) 25–32
31
Fig. 5. Parent fraction of [11C]1, [18F]2 and [11C]7b in plasma (left side) and brain (right side) of male Wistar rats at 15 and 45 min (** P b 0.01, *** P b 0.001).
could be assumed to be the same for all three compounds. Previously, a
structure–activity relationship of CNS-5161 and its sulfoxide derivatives
has been reported [28]. This study showed a 455 fold decrease in affinity
toward the NMDAr when the sulfur was oxidized. Assuming that the af-
finity will drop dramatically when the sulfur is oxidized, this
radiolabeled metabolite would not interfere with the binding of [11C]
1, [18F]2 and [11C]7b, respectively.
N-demethylation is primarily caused by the cytochrome P450
enzyme system [30]. Ligand [11C]7b, which is radiolabeled at the guani-
dine N-methyl, shows only 1% of polar metabolites after 45 min. Likely,
this process does not contribute much to the amount of radiolabeled
metabolites measured in the brain. In addition, previous structure–
activity relationship studies have shown that omitting the N-methyl in
similar compounds decreased their affinity toward the NMDAr by 6 to
20 fold [18].
References
[1] Madden DR. The structure and function of glutamate receptor ion channels. Nat Rev
Neurosci 2002;3:91–101.
[2] Danysz W, Parsons CG. Alzheimer's disease, β-amyloid, glutamate, NMDA re-
ceptors and memantine – searching for the connections. Br J Pharmacol 2012;
167:324–52.
[3] Lakhan SE, Caro M, Hadzimichalis N. NMDA receptor activity in neuropsychiatric dis-
orders. Front Mol Psychiatry 2013;4:52.
[4] Waterhouse RN. Imaging the PCP site of the NMDA ion channel. Nucl Med Biol 2003;
30:869–78.
[5] Sobrio F, Gilbert G, Perrio C, Barré L, Debruyne D. PET and SPECT imaging of the
NMDA receptor system: an overview of radiotracer development. Mini Rev Med
Chem 2010;10:870–86.
[6] Majo VJ, Prabhakaran J, Mann JJ, Kumar JSD. PET and SPECT tracers for glutamate
receptors. Drug Discov Today 2013;18:176–84.
[7] Sobrio F. Radiosynthesis of carbon-11 and fluorine-18 labelled radiotracers to image
the ionotropic and metabotropic glutamate receptors. J Label Compd Radiopharm
2013;56:180–6.
[8] Waterhouse RN, Dumont F, Sultana A, Simpson N, Laruelle M. Synthesis of [11C] N-(2-
chloro-5-thiomethylphenyl)-N’-(3-methoxyphenyl)-N’-methylguanidine ([11C]GMOM):
a candidate PET tracer for imaging the PCP site of the NMDA ion channel. J Label
Compd Radiopharm 2002;45:955–64.
It is known that metabolically labile sites can be made less vulnera-
ble to metabolism by substitution with fluorine [31]. Compared with
[
11C]1, [18F]2 indeed showed a slightly lower rate of metabolism, but
[9] Waterhouse RN, Slifstein M, Dumont F, Zhao J, Chang RC, Sudo Y, et al. In vivo
evaluation of [11C]N-(2-chloro-5-thiomethylphenyl)-N’-(3-methoxy-phenyl)-N’-
methylguanidine ([11C]GMOM) as a potential PET radiotracer for the PCP/
NMDA receptor. Nucl Med Biol 2004;31:939–48.
[10] van der Doef TF, Golla SSV, Klein PJ, Oropeze-Seguias GM, Schuit RC, Metaxas A, et al.
Quantification of the novel N-methyl-D-aspartate receptor ligand [11C]GMOM in
man. J Cereb Blood Flow Metab 2016;36:1111–21.
[11] Dumont F, Sultana A, Waterhouse RN. Synthesis and In Vitro Evaluation of N,N’-
Diphenyl and N-Naphthyl-N’-phenylguanidines as N-Methyl-D-aspartate Receptor
Ion-Channel Ligands. Bioorg Med Chem 2002;12:1583–6.
[12] Hammers A, Asselin MC, Brooks DJ, Luthra SK, Hume SP, Thompson, et al. Correlation
of memory function with binding of [C-11]CNS 5161, a novel putative NMDA ion
channel PET ligand. NeuroImage 2004;22(Suppl. 2):T54.
at both 15 and 45 min no significant difference was observed between
the parent fractions of [11C]1 and [18F]2 in plasma.
By further substitution of the fluoromethoxy moiety with a vinyl
group ([11C]7b), it was anticipated that the rate of metabolism would
be further decreased, as this should prevent oxidation by the cyto-
chrome P450 system. Again, a slight decrease in rate of metabolism
could be observed, but again the parent fraction in plasma of [11C]7b
was not significantly higher than that of [18F]2 at either 15 or 45 min.
To summarize, at 15 min the fraction of intact parent compound of
[
11C]1, [18F]2 and [11C]7b in the brain was in the same range (more
[13] Asselin MC, Hammers A, Turton D, Osman S, Koepp M, Brooks D. Initial kinetic analyses
of the in vivo binding of the putative NMDA receptor ligand [C-11]CNS 5161 in
humans. NeuroImage 2004;22(Suppl. 2):T137.
than 90%). At 45 min, the parent fraction still was 90% for [18F]2, followed
by [11C]7b (87%) and [11C]1 (78%).
Based on results presented in this study, both [18F]2 and [11C]7b
might be better PET tracers for imaging the NMDAr than [11C]1. Both
radiotracers showed a slightly improved affinity to the ion channel of
the NMDAr and a small increase in metabolic stability. Compound
[14] Zhao Y, Robins E, Turton D, Brady F, Luthra SK, Arstad E. Synthesis and characterization of
N-(2-chloro-5- methylthiophenyl)-N’-(3-methylthiophenyl)-N’-[11C]methylguanidine
[
11C]CNS 5161, a candidate PET tracer for functional imaging of NMDA receptors.
J Label Compd Radiopharm 2006;49:163–70.
[15] Ahmed I, Bose SK, Pavese N, Ramlackhansingh A, Turkheimer F, Hotton G, et al.
Glutamate NMDA receptor dysregulation in Parkinson's disease with dyskinesias.
Brain 2011;134:979–86.
[16] Robins EG, Zhao Y, Khan I, Wilson A, Luthra SK, Arstad E. Synthesis and in vitro
evaluation of 18F-labelled S-fluoroalkyl diarylguanidines: novel high-affinity
NMDA receptor antagonists for imaging with PET. Bioorg Med Chem Lett 2010;
20:1749–51.
[17] McGinnity CJ, Hammers A, Riano Barros DA, Luthra SK, Jones PA, Trigg W, et al. Initial
evaluation of 18F-GE-179, a putative PET tracer for activated N-methyl D-aspartate
receptors. J Nucl Med 2014;55:423–30.
[18] Klein PJ, Christiaans JAM, Metaxas A, Schuit RC, Lammertsma AA, van Berckel BNM,
et al. Synthesis, structure activity relationship, radiolabeling and preclinical evaluation
of high affinity ligands for the ion channel of the N-methyl-D-aspartate receptor as
potential imaging probes for positron emission tomography. Bioorg Med Chem
2015;23:1189–206.
[
11C]7b has a favorable lipophilicity over compounds [11C]1 and [18F]
2. In addition, [11C]7b showed reduced uptake following pre-
treatment with the ion channel blocker MK-801. On the other hand
compound [18F]2 showed less radioactive polar metabolites in the
brain. However, further studies are needed in non-human primates,
but preferably in humans, as anesthesia could interfere with NMDAr
binding [32].
Supplementary data to this article can be found online at http://dx.doi.
org/10.1016/j.nucmedbio.2017.05.003.
[19] Golla SSV, Klein PJ, Bakker J, Schuit RC, Christiaans JAM, van Geest L, et al. Preclinical
evaluation of [18F]PK-209, a new PET ligand for imaging the ion-channel site of
NMDA receptors. Nucl Med Biol 2015;42:205–12.
Acknowledgments
The authors would like to thank Rolph van Kooij for his assistance in
the radiosynthesis and the BV Cyclotron VU for providing [11C]CO2.
This research was performed within the framework of CTMM, the
Center for Translational Molecular Medicine (www.ctmm.nl), project
LeARN (grant 02N-101).
[20] Meunier B, de Visser SP, Shaik S. Mechanism of oxidation reactions catalyzed by
cytochrome P450 enzymes. Chem Rev 2004;104:3947–80.
[21] Gant TG. Using deuterium in drug discovery: leaving the label in the drug. J Med
Chem 2014;57:3595–611.
[22] Fujiwara Y, Dixon JA, O'Hara F, Funder ED, Dixon DD, Rodriguez RA, et al. Practical and
innate carbon-hydrogen functionalization of heterocycles. Nature 2012;492:95–9.

32
P.J. Klein et al. / Nuclear Medicine and Biology 51 (2017) 25–32
[23] Windhorst AD, ter Linden T, de Nooij A, Buijs FL, van Rooij LF, Hersheid JDM. A com-
plete, multipurpose, low cost, fully automated and GMP compliant radiosynthesis
system. J Label Compd Radiopharm 2001;44:S1052–4.
[24] Antonsen Ø, Benneche T, Undheim K. Syntheses of sulfinylmeyhyl ethers and
converesion of these into halomethyl and acyloxymethyl ethers. Acta Chem Scand
B 1988;42:515–23.
[28] Padmanabhan S, Perlman ME, Zhang L, Moore D, Zhou D, Fisher JB, et al. Identification
and characterization of a potential ischemia-selective N-methyl-D-aspartate (NMDA)
receptor ion-channel blocker, CNS 5788. Bioorg Med Chem Lett 2001;11:501–4.
[29] Price GW, Ahier RG, Middlemiss DN, Singh L, Tricklebank MD, Wong EHF. In vivo
labelling of the NMDA receptor channel complex by [3H]MK-801. Eur J Pharmacol
1988;158:279–82.
[25] Pike VW. PET radiotracers: crossing the blood-brain barrier and surviving metabo-
[30] Boersma MG, Cnubben NHP, van Berkel WJH, Blom M, Vervoort J, Rietjens IMCM.
Role of cytochromes P-450 and flavin-containing monooxygenase in the biotrans-
formation of 4-fluoro-N-methylaniline. Drug Metab Dispos 1993;21:218–35.
[31] Purser S, Moore PR, Swallow S, Gouverneur V. Fluorine in medicinal chemistry.
Chem Soc Rev 2008;37:320–30.
lism. Trends Pharmacol Sci 2009;30:431–40.
[26] Murray F, Kennedy J, Hutson PH, Elliot J, Huscroft I, Mohnen K, et al. Modulation of
[3H]MK-801 binding to NMDA receptors in vivo and in vitro. Eur J Pharmacol 2000;
397:263–70.
[27] Wegener N, Nagel J, Gross R, Chambon C, Greco S, Pietraszek M, et al. Neurosci Lett
[32] Petrenko AB, Yamakura T, Sakimura K, Baba H. Defining the role of NMDA receptors
2011;503:68–72.
in anesthesia: are we there yet? Eur J Pharm 2014;723:29–37.