Total synthesis and evaluation of [18F]MHMZ

Article abstract of DOI:10.1016/j.bmcl.2007.12.054

Radiochemical labeling of MDL 105725 using the secondary labeling precursor 2-[¹⁸F]fluoroethyltosylate ([¹⁸F]FETos) was carried out in yields of ～90% synthesizing [¹⁸F]MHMZ in a specific activity of ～50 MBq/nmol with a sta

Full text of DOI:10.1016/j.bmcl.2007.12.054

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 1515–1519
Total synthesis and evaluation of [18F]MHMZ
Matthias M. Herth,a,*, Fabian Debus,b, Markus Piel,a Mikael Palner,c
Gitte M. Knudsen,c Hartmut Luddensb and Frank Ro¨scha
¨
aInstitute of Nuclear Chemistry, University of Mainz, Fritz-Strassmann-Weg 2, 55128 Mainz, Germany
bDepartment of Psychiatry, Clinical Research Group, Untere Zahlbacher Straße 8, 55131 Mainz, Germany
cCenter for Integrated Molecular Brain Imaging, Rigshospitalet, Blegdamsvej 9, DK-2100 Copenhagen Ø, Denmark
Received 5 November 2007; revised 27 November 2007; accepted 20 December 2007
Available online 25 December 2007
Abstract—Radiochemical labeling of MDL 105725 using the secondary labeling precursor 2-[18F]fluoroethyltosylate ([18F]FETos)
was carried out in yields of ꢀ90% synthesizing [18F]MHMZ in a specific activity of ꢀ50 MBq/nmol with a starting activity of
ꢀ3 GBq. Overall radiochemical yield including [18F]FETos synthon synthesis, [18F]fluoroalkylation and preparing the injectable
[18F]MHMZ solution was 42% within a synthesis time of ꢀ100 min. The novel compound showed excellent specific binding to
the 5-HT2A receptor (Ki = 9.0 nM) in vitro and promising in vivo characteristics.
ꢀ 2007 Elsevier Ltd. All rights reserved.
Serotonergic 5-HT2A receptors are of central interest in
the pathophysiology of schizophrenia and other dis-
eases, including Alzheimer’s disease and personality dis-
orders.1 The serotonergic system is also implicated in
sleep, aging, and pain.2 In vivo studies of 5-HT2A recep-
tor occupancy would provide a significant advance in
the understanding of the mentioned disorders and con-
ditions. Positron emission tomography (PET) is an
appropriate tool to measure in vivo directly, non-inva-
sively, and repetitively the binding potential of radio
tracers for neuroreceptors.
Ki = 0.13 nM4; (R)-MDL 100907: Ki = 0.57 nM6}. Affin-
ities are more than 100-fold higher for other receptors
such as 5-HT2C, a1, D1, and D2. Nevertheless, it was pro-
posed that the selectivity of [11C]MDL 100907 for 5-
HT2A receptor is slightly higher than the selectivity for
this receptor of [18F]altanserin.8 Both tracers show in
in vitro and in in vivo experiments, high affinity, selectiv-
ity, and a good ratio of specific to non-specific binding
for 5-HT2A receptors.3,7 The advantage of [18F]altanserin
over [11C]MDL 100907 is the possibility to perform equi-
librium scans lasting several hours and to transport the
tracer to other facilities based on the 110 min half-life
of [18F]fluorine. A drawback of [18F]altanserin is its rapid
and extensive metabolism. Four metabolites are formed
in humans that cross the blood–brain-barrier,7 whereas
metabolites of [11C]MDL 100907 do not enter the brain
to any larger extent.9
A number of neurotransmitter analogs labeled with b+-
emitter containing radioligands were synthesized as
radiopharmaceuticals for the imaging of the 5-HT2A
receptor. To date, in vivo studies have been performed
with several 5-HT2A selective antagonists such as
[11C]MDL 100907,3 [18F]altanserin,4 and [11C]SR
46349B5.
The aim of this study was to develop an 18F-analog of
MDL 100907 (1) combining advantages of both ligands,
the better selectivity of MDL 100907 and the superior
isotopic properties of [18F]fluorine. For this purpose
we decided to replace one of the O-methyl groups by
an O-2-[18F]fluoroethyl moiety resulting in [18F]MHMZ
([18F]FE1-MDL 100907) ((3-[18F]fluoro-ethoxy-2-methoxy-
phenyl)-1-[2-(4-fluoro-phenyl)ethyl-4-piperidine-methanol,
2) (Fig. 1).
Within those ligands, [18F]altanserin and [11C]MDL
100907 represent the radioligands of choice for in vivo
5-HT2A PET imaging because of their high affinity and
selectivity for the 5-HT2A receptor {altanserin:
Keywords:
[
18F]MHMZ; MDL 100907;
Antagonist; Positron emission tomography; Autoradiography.
[
18F]Altanserin; 5-HT2A
;
*
Corresponding author. Tel.: +49 6131 39 25849; e-mail:
herthm@uni-mainz.de
These authors equally contributed to this work.
The methoxy group in the 3-position seemed to be more
suitable for labeling because previous [11C]MDL 100907
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.12.054

1516
M. M. Herth et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1515–1519
Figure 1. Structures of [11C]MDL 100907 (1), [18F]MHMZ (2), and MDL 105725 (3).
studies showed that metabolism predominantly resulted
in the formation of its 3-OH analog MDL 105725 ((3-hy-
droxy-2-methoxy-phenyl)-1-[2-(4-fluorophenyl)ethyl-4-
piperidine-methanol, 3). 18F-Labeling in the 2-position
would therefore lead to extensive formation of the
labeled 3-OH-analog (2-[18F]fluoro-ethoxy-3-methoxy-
phenyl)-1-[2-(4-fluoro-phenyl)ethyl-4-piperidine-meth-
anol that may be expected to cross the blood–brain-
barrier or to be metabolized within the brain and thus
interfere with the interpretation of the labeled tracer
uptake.10,11
Nucleosilꢁ 5 C18; MeCN/H2O 40:60, Rf = 8.68 min).
These results justified further analyses like determina-
tion of the affinity and the route for radioactive syn-
theses, receptor autoradiography, and metabolism
studies.
A radioligand competition binding assay was carried
out with GF-62 cells, a clonal cell line expressing high
amounts (5–7 pmol/mg) of the 5-HT2A receptor, in test
tubes containing [3H]MDL (0.2 nM) and seven differ-
ent concentrations of test compounds (1 lM–1 pM) in
a total of 1mL assay buffer. Ketanserin (1 lM) was
used to determine non-specific binding. The 5-HT2A
binding affinities of the racemic MHMZ and the refer-
ence compounds altanserin and MDL 100907 are
shown in Table 1.
A useful synthetic route to MDL 100907 and its racemic
precursor MDL 105725 has been published by Huang
et al.3 The route depended upon a key transformation
of an ester to a ketone via an amide intermediate
(Fig. 2) and was carried out essentially as published3
with minor modifications.
MHMZ showed a 4.5 times lower affinity as compared
to the parent compound MDL 100907 but still was in
the nanomolar range. The assay was performed n = 4
times.
Finally, MHMZ was synthesized via a fluoroalkylation
of the precursor MDL 105725 in dry DMF by addition
of sodium hydride and 1-bromo-2-fluoroethane (Fig. 3)
in a yield of 40%. A chiral derivatization of the final
product MHMZ was not performed.
[18F]Fluoroalkylation of the precursor MDL 105725
was carried out using [18F]FETos, which was produced
in an automated module.12 Optimization of the reaction
conditions gave radiochemical yields of about 90% at a
reaction temperature of 100 ꢂC in a reaction time of
The purity of MHMZ was examined to be higher than
98% as indicated by HPLC analysis (ET 250/8/4
Figure 2. (a) PBr3, toluene; (b) K2CO3, DMF; (c) Me(MeO)NH HCl, EtMgBr, THF; (d) n-BuLi, THF, TBDPS-guajacol; (e) NaBH4, MeOH; (f)
K2CO3, MeOH, H2O.

M. M. Herth et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1515–1519
1517
Figure 3. Synthesis of MHMZ.
Table 1. Receptor binding data of MDL 100907 derivatives and
altanserin
Compound
Ki (nM)
MHMZ
Altanserin
MDL 100907
9.00 0.10
0.74 0.88
2.10 0.13
10 min using 7 mmol precursor and 7 mmol 5 N NaOH
as a base in dry DMF as a solvent.
The optimization procedure of the radiochemical yield
of [18F]MHMZ is exemplified for the parameter temper-
ature in Figure 4. The final formulation of the injectable
solution including a semipreparative HPLC (ET 250/8/4
Nucleosilꢁ 5 C18; MeCN/H2O 40:60, Rf = 8.68 min)
took no longer than 100 min and provided [18F]MHMZ
(2) with a purity >96% as indicated by analytical HPLC
analyses. The specific activity was determined to be
ꢀ50 MBq/nmol with a starting amount of radioactivity
of 3 GBq of [18F]fluorine.
Figure 5. Images of an autoradiography of [18F]MHMZ binding at
14 lm thick rat brain sections; (A and B) total binding at
a
Autoradiographic images of the 5-HT2A receptor ob-
tained with [18F]MHMZ showed excellent visualization
results in rat brain sections (Fig. 5). Images were in com-
plete agreement with the distribution obtained with
[3H]MDL 10090713 (also Fig. 6B and C). Highest bind-
ing was detected in lamina V of the frontal cortex, the
caudate-putamen, the motor trigeminal nucleus, the fa-
cial nucleus, and the pontine nuclei. Minor binding
was detected in the olfactory system, the mesencepha-
lon, and the hippocampus.
concentration of 5 nM with (A) lateral 0.9 mm and (B) lateral
2.4 mm from bregma. Major binding was detected in lamina V (V)
of the frontal cortex, in the caudate-putamen (CPu), and three regions
of the brain stem, the motor trigeminal nucleus (MoT), facial nucleus
(fn), and the pontine nuclei (pn). Non-specific binding was determined
in the presence of 10 lM ketanserin which led to total inhibition of
[
nmol (at the end of the incubation period).
18F]MHMZ binding (cf. C0 Fig. 6). Specific activity was 1.38 MBq/
Competition autoradiography assays (data not shown)
with 5 nM [18F]MHMZ and 10 lM of fallypride,
WAY 100635, and prazosin showed that [18F]MHMZ
is highly specific for 5-HT2A receptors. Displacement
could only be detected with fallypride. Here, co-incuba-
tion led to a displacement of 30% (n = 4, 6% SEM) of
total binding in the frontal cortex as well as in the cau-
date-putamen, which does not imply that [18F]MHMZ
recognizes D2/D3 receptors but might rather be ex-
plained by the known cross affinity of fallypride to 5-
HT2 receptors.14
Binding parameters of [18F]MHMZ of different regions
of the rat brain obtained with autoradiography assays
at sagittal sections are displayed in Table 2. Binding in
the cerebellum was at the level of non-specific binding
so levels of binding in different brain regions are also gi-
ven relative to that.
A comparison of the binding of [18F]altanserin and
[18F]MHMZ (Fig. 6) displays that [18F]MHMZ is in
Figure 4. [18F]Fluoroalkylation of 7 mmol MDL 105725 at different
reaction temperatures using DMF and 7 mmol 5 N NaOH.

1518
M. M. Herth et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1515–1519
Figure 6. Autoradiographic images of the total binding and non-specific binding, respectively, of (A/A0) [18F]altanserin, (B/B0) [3H]MDL 100907 and
(C/C0) 5 nM [18F]MHMZ at 14 lm rat brain sections. Non-specific binding was determined in the presence of 10 lM ketanserin. Specific activity of
[
18F]MHMZ and [18F]altanserin was ꢀ160 kBq/nmol (at the end of the incubation period). Washing was done 2 · 10 min for (A/A0) in ice-cold
reaction buffer, 2 · 2 min at room temperature with (B/B0) and 3 · 2 min at room temperature (4 min with buffer containing 0.01% Triton X-100).
Reaction buffer was 50 mM Tris buffer, pH 7.4, containing 120 mM NaCl2 and 5 mM KCl.
Table 2. Binding parameters obtained with [18F]MHMZ from binding
experiments at 14 lm sagittal sections of the rat brain
(x = means SEM)
n
pmol/mm3
Region/cerebellum
Frontal cortex
Laminae I–IV
Lamina V
4
4
4
4
23.30 1.69
51.60 5.24
27.27 2.76
16.80 2.33
26.9 0.9
59.5 2.8
31.4 1.3
19.2 1.4
Laminae VIa + VIb
Caudate-putamen
no way inferior to [18F]altanserin in terms of specificity
for 5-HT2A receptors. Figure 6 also shows the complete
agreement of the binding of [3H]MDL 100907 and
[18F]MHMZ.15
The metabolite analyses of rat plasma (Fig. 7) showed
that [18F]MHMZ underwent fast metabolism. Plasma
samples were taken at 5, 10, 30, and 60 min and ana-
lyzed by radio-TLC. One polar metabolite was found
in rat plasma which is not likely to cross the blood–
brain-barrier because of its hydrophilicity. The percent-
age of unmetabolized fractions was 43%, 32%, 16%, and
7% at 5, 10, 30, and 60 min, respectively.
Figure 7. (A) Plasma clearances of [18F]MHMZ at 5, 10, 30, and
60 min (n = 3 per time point; means SD shown). (B) Radioactivity in
TLC plate of plasma samples at 5 min pi is shown. Spots for
[
clearly visible.
18F]MHMZ (T) (Rf = 0.76) and its metabolite (M) (Rf = 0.16) were
First autoradiographic studies showed excellent in vitro
binding with high specificity of [18F]MHMZ for 5-HT2A
receptors and very low non-specific binding.
In conclusion, precursors and reference compounds of
[18F]MHMZ were synthesized in high yields. The new
18F-labeled compound could be obtained as an inject-
able solution in overall radiochemical yields of about
42% within a synthesis time of about 100 min in a purity
of >96% and high specific activities. This is very similar
to the radiosynthesis of [18F]altanserin, which takes 75–
100 min and results in a radiochemical yield between
30% and 50%.4
[18F]MHMZ undergoes fast metabolism resulting in one
very polar active metabolite.
Except from the slightly decreased affinity the reported
in vitro data seem to be comparable with those of
[3H]MDL 100907. Our data suggest that the aim of

M. M. Herth et al. / Bioorg. Med. Chem. Lett. 18 (2008) 1515–1519
1519
developing a novel 18F-analog of MDL 100907 (1) com-
bining the better selectivity of MDL 100907 as com-
pared to altanserin and the superior isotopic properties
for the clinical routine of [18F]fluorine as compared to
[11C]carbon could be achieved.
6. Heinrich, T.; Boetcher, H.; Pruecher, H.; Gottschlich, R.;
Ackermann, K.-A.; van Amsterdam, C. Chem. Med.
Chem. 2006, 1, 245.
7. Tan, P.-Z.; Baldwin, R. M.; Fu, T.; Charney, D. S.; Inis,
R. B. J. Labelled Compd. Radiopharm. 1998, 42, 457.
8. Meltzer, C. C.; Smith, G.; DeKosky, S. T.; Pollock, B. G.;
Mathis, A. M.; Moore, R. Y.; Kupfer, D. J.; Reynolds, C.
F. Neuropsychopharmacology 1998, 18, 407.
All together, new auspicious results concerning the syn-
thesis and of the in vitro studies of [18F]MHMZ justify
further experiments like ex vivo brain regional distribu-
tion and in vivo small animal PET studies to verify the
potential of this new 5-HT2A imaging ligand.
9. Scot, D.; Heath, T. G. J. Pharm. Biomed. Anal. 1998, 17,
17.
10. Ullrich, T.; Ice, K. C. Bioorg. Med. Chem. 2000, 8, 2427.
11. Lundkvist, C.; Halldin, C.; Ginovart, N.; Swahn, C.-G.;
Carr, A. A.; Brunner, F.; Farde, L. Life Sci. 1996, 58, 187.
12. Bauman, A.; Piel, M.; Schirrmacher, R.; Ro¨sch, F.
Tetrahedron Lett. 2003, 44, 9165.
13. Lopez-Gimenez, J. F.; Mengod, D.; Palacios, J. M.;
Vilario, M. T. Naunyn-Schmiedeberg’s Arch. Pharmacol.
1997, 356, 446.
Acknowledgments
The authors thank S. Ho¨hnemann and P. Riss for the
syntheses of [18F]FETos. We also like to thank the
VCI (Verband der chemischen Industrie e.V.) for the
donation of solvents. Financial support by Friedrich–
Naumann–Stiftung, the European Network of Excel-
lence (EMIL), and the Deutsche Forschungs-gemein-
schaft (DFG) is gratefully acknowledged.
14. Stark, D.; Piel, M.; Hubner, H.; Gmeiner, P.; Grunder,
¨
G.; Ro¨sch, F. Biorg. Med. Chem. 2007, 15, 6819.
¨
15. Autoradiography experiments were carried out at room
temperature in reaction buffer (50 mM Tris/HCl buffer,
pH 7.4, containing 120 mM NaCl2, and 5 mM KCl) with
[3H]MDL 100907 and [18F]MHMZ and on ice with
[18F]altanserin. Sections with [18F]MHMZ were washed
2 · 2 min in reaction buffer containing 0,01% Triton X-100
and 1 · 2 min in reaction buffer, shortly dipped into
deionized water, and quickly dried in a stream of cold air.
Sections with [18F]altanserin were washed in pure ice-cold
reaction buffer 2 · 10 min, sections with [3H]MDL 100907
were washed in pure buffer 2 · 2 min. Sections were
exposed to Fuji phosphor screen for 3 h when 18F was
References and notes
1. Scientific articles: Kristiansen, H.; Elfing, B.; Plenge, P.;
Pinborg, L. H.; Gillings, N.; Knudsen, G. M. Synapse
2005, 58, 249.
2. Lemaire, C.; Cantineau, R.; Guillaune, M.; Plenevaux, A.;
Christiaens, L. J. Nucl. Med. 1991, 32, 2266.
3. Huang, Y.; Mahmood, K.; Mathis, C. A. J. Labelled
Compd. Radiopharm. 1999, 42, 949.
4. Hamacher, K.; Coenen, H. H. Appl. Radiat. Isot. 2006, 64,
989.
5. Alexoff, D. L.; Shea, C.; Fowler, J. S.; King, P.; Gatley, S.
J.; Schleyer, D. J.; Wolf, A. P. Nucl. Med. Biol. 1995, 22,
892.
3
used and for 5 days when H was used. Screens were read
out with a Fuji FLA-7000 scanner. For 18F quantification
was done after calibration by a standard curve which was
obtained by a dilution series of the radiotracer. Calibra-
tion was repeated for each fresh radiotracer synthesis.
Calibration for sections with 3H was done with Amersham
microscale standards. Calibration, quantification and data
evaluation was done with Multi Gauge, Fujifilm image
analysis software.